Loading...
The URL can be used to link to this page
Your browser does not support the video tag.
Home
My WebLink
About
5016; Desalination - Poseidon Resources Corporation; Poseidon Resources Corporation Corrosion Pilot Study; 2006-05-02
1 Poseidon Resources Corporation Corrosion Pilot Study FINAL REPORT May 2, 2006 McGuire/Malcolm Pirnie 1919 Santa Monica Blvd. Suite 200 Santa Monica, CA 90404-1955 Study Conducted by: Michael J. McGuire, Ken Reich, Nicole K. Blute, and Nicole J. West 2 Table of Contents TABLE OF TABLES ......................................................................................................................4 TABLE OF FIGURES.....................................................................................................................4 EXECUTIVE SUMMARY.............................................................................................................. 8 INTRODUCTION............................................................................................................................ 11 Proposed Ocean Desalination Plant in Carlsbad.................................................................... 11 Integration of Desalinated Seawater........................................................................................ 11 PROJECT DESCRIPTION ..........................................................................................................12 Purpose and Objectives of the Corrosion Study........................................................ 12 Poseidon Pilot Desalination Treatment Facility ........................................................ 12 Poseidon Corrosion Pilot Facility ................................................................................ 15 Overview......................................................................................................................................15 RO Permeate Source Water Conditioning...........................................................................19 Copper Pipe/ Brass Water Meter Flow-Through Systems............................................... 21 Mortar-Lined Pipe Recirculation Systems..........................................................................24 Distribution System Valve Recirculation Systems...........................................................28 EXPERIMENTAL PLAN ............................................................................................................33 Source Waters ............................................................................................................... 33 System Operations........................................................................................................ 33 Sample Collection and Analysis .................................................................................. 36 Sample Locations......................................................................................................................36 Analyses and Monitoring Frequency.................................................................................... 37 Revised Monitoring Frequency..............................................................................................39 Analytical Methods and Reporting Limits...........................................................................39 Sample Containers, Preservation, and Holding Times ....................................................40 QA/QC Samples.........................................................................................................................42 Visual Observations of Internal Pipe Surfaces..................................................................48 WATER QUALITY DATA RESULTS.......................................................................................49 Data Reduction..............................................................................................................49 Water Quality Calculations for Copper Pipes and Brass Meters...................................49 Water Quality Calculations for Distribution System Valves...........................................49 Water Quality Calculations for Mortar-Lined Pipes..........................................................49 RO Permeate Conditioning...........................................................................................50 Source Water Comparison: MWD vs. Conditioned RO Permeate........................... 52 Copper Pipes and Brass Water Meters........................................................................ 61 Distribution System Valves ......................................................................................... 67 Mortar-Lined Pipes ....................................................................................................... 73 3 DISCUSSION OF CORROSION RESULTS ............................................................................80 Comparison to Water Quality Standards ...................................................................80 Copper Pipes and Brass Water Meters................................................................................80 Distribution System Valves....................................................................................................85 Mortar-Lined Pipes...................................................................................................................90 Statistical Analyses......................................................................................................94 Visual Observation of Internal Pipe Surfaces........................................................... 96 SUMMARY AND CONCLUSIONS ...........................................................................................101 REFERENCES.............................................................................................................................103 4 Table of Tables Table 1: Final RO Permeate Water Quality ..............................................................................................35 Table 2: Sampling Location Key ................................................................................................................36 Table 3: Sampling Locations and Frequencies.......................................................................................38 Table 4: Analytical Methods and Data Reporting Limits.....................................................................40 Table 5: Sample Analysis Methods, Preservation, and Holding Times .............................................41 Table 6: RPD for Duplicates........................................................................................................................43 Table 7: Pipe Sections Sampled for Visual Inspection.........................................................................48 Table 8: Hydroxide Alkalinity Added by RO Conditioning Procedure ...............................................51 Table 9: Statistical Significance of Difference as Determined by Two-Way ANOVA with Replication ( = 0.05) and Average Analytical Results................................................................94 Table of Figures Figure 1: Schematic of Poseidon Pilot Desalination Plant....................................................................14 Figure 2: Poseidon Resources Pilot Facility.............................................................................................15 Figure 3: Plan View of the Corrosion Investigation Pilot Facility.......................................................16 Figure 4: Pipe Test System Layout............................................................................................................18 Figure 5: Corrosion Pilot Plant During Construction.............................................................................19 Figure 6: Schematic of RO Permeate Conditioning Process..............................................................20 Figure 7: Chemical Addition Location – Empty Bag Filters..................................................................21 Figure 8: Schematic of Consumer Copper Plumbing Flow-Through Systems ...............................22 Figure 9: Construction Details of Consumer Plumbing Flow-Through Systems............................23 Figure 10: Consumer Plumbing Flow-through System.........................................................................23 Figure 11: Consumer Plumbing Brass Meters..........................................................................................24 Figure 12: Schematic of Mortar-lined Pipe Loops..................................................................................26 Figure 13: Construction Details of Mortar-lined Pipe Loops ...............................................................27 Figure 14: Mortar-lined Pipe Loops...........................................................................................................28 Figure 15: Schematic of Distribution Valve Loops.................................................................................30 Figure 16: Distribution Valve Loop Construction Details......................................................................31 Figure 17: Distribution Valve Loops ...........................................................................................................31 Figure 18: Exterior of the Corrosion Pilot Facility.................................................................................32 Figure 19: Recirculation Test Systems Under Canopy .........................................................................32 Figure 20: RPD for Calcium ........................................................................................................................43 Figure 21: RPD for Aluminum......................................................................................................................44 Figure 22: RPD for Iron................................................................................................................................44 5 Figure 23: RPD for Zinc................................................................................................................................45 Figure 24: RPD for Lead..............................................................................................................................45 Figure 25: RPD for Copper..........................................................................................................................46 Figure 26: RPD for TDS................................................................................................................................46 Figure 27: RPD for Sulfate..........................................................................................................................47 Figure 28: RPD for Chloride........................................................................................................................47 Figure 29: Alkalinity Increase During Permeate Disinfection and pH Adjustment........................51 Figure 30: pH of Source Waters ................................................................................................................52 Figure 31: Alkalinity of Source Waters .....................................................................................................53 Figure 32: Calcium of Source Waters.......................................................................................................54 Figure 33: TDS of Source Waters..............................................................................................................55 Figure 34: Chloride in Source Waters......................................................................................................56 Figure 35: Sulfate in Source Waters.........................................................................................................56 Figure 36: Dissolved Oxygen in Source Waters.....................................................................................57 Figure 37: Temperature of Source Waters Added to the Pipe Loops..............................................58 Figure 38: Temperature of Water in the Test Systems After 10 hr in the Copper Pipe or One Week in the Mortar-Lined Pipe Loops and Valve Loops...............................................58 Figure 39: Langelier Index of Source Waters.........................................................................................59 Figure 40: Calcium Carbonate Precipitation Potential (CCPP) of Source Waters........................60 Figure 41: Total Chlorine Concentrations in Influent Water to Copper Pipe Test Systems........62 Figure 42: Total Chlorine Concentrations in Water Sampled from Copper Pipe Test Systems After 10 hr Stagnation Periods ..................................................................................................................62 Figure 43: Total Ammonia Concentrations in Influent Water to the Copper Pipe Test Systems..................................................................................................................................................63 Figure 44: Total Ammonia Concentrations in Water Sampled from the Copper Pipes After 10 hr Stagnation Periods.......................................................................................................63 Figure 45: Nitrite Concentrations in Influent Water to Copper Pipe Test Systems .....................64 Figure 46: Nitrite Concentrations in Water Sampled from Copper Pipes After 10 hr Stagnation Periods .......................................................................................................................................64 Figure 47: Change in pH in Water Sampled from the Copper Pipe Test Systems After 10 hr Stagnation Periods ..................................................................................................................65 Figure 48: Change in Turbidity in Water Sampled from the Copper Pipe Test Systems After 10 hr Stagnation Periods .................................................................................................66 Figure 49: Interior of the Distribution System Valves Showing Iron Corrosion............................67 Figure 50: Total Chlorine Concentrations in Influent Water to the Distribution System Valve Loops.....................................................................................................................................68 Figure 51: Total Chlorine Concentrations in Water Sampled from the Distribution System Valve Loops After One-Week Recirculation Periods............................................................................68 6 Figure 52: Total Ammonia Concentrations in Influent Water to the Distribution System Valve Loops.....................................................................................................................................69 Figure 53: Total Ammonia Concentrations in Water Sampled from the Distribution System Valve Loops After One-Week Recirculation Periods............................................................................69 Figure 54: Nitrite Concentrations in Influent Water to the Distribution System Valve Loops....................................................................................................................................................70 Figure 55: Nitrite Concentrations in Water Sampled from the Distribution System Valve Loops After One-Week Recirculation Periods........................................................................................70 Figure 56: Change in pH in Water Sampled from Distribution System Valve Loops After One- Week Recirculation Periods.........................................................................................................................71 Figure 57: Change in Turbidity in Water Sampled from Distribution System Valve Loops After One-Week Recirculation Periods....................................................................................................72 Figure 58: Total Chlorine Concentrations in Influent Water to Mortar-Lined Pipe Loops................................................................................................................................................................74 Figure 59: Total Chlorine Concentrations in Water Sampled from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods........................................................................................74 Figure 60: Total Ammonia Concentrations in Influent Water to Mortar-Lined Pipe Loops .......75 Figure 61: Total Ammonia Concentrations in Water Sampled from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods........................................................................................75 Figure 62: Nitrite Concentrations in Influent Water to Mortar-Lined Pipe Loops ........................76 Figure 63: Nitrite Concentrations in Water Sampled from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods........................................................................................76 Figure 64: Change in pH in Water Sampled from Mortar-Lined Pipe Loops After One-Week Recirculation Periods...............................................................................................................77 Figure 65: Change in Calcium in Water Sampled from Mortar-Lined Pipe Loops After One-Week Recirculation ...................................................................................................................78 Figure 66: Change in Turbidity of Water Sampled from Mortar-Lined Pipe Loops After One-Week Recirculation Periods....................................................................................................79 Figure 67: Copper Leached from Brass Meters After 10 hr Stagnation Periods Compared to California Drinking Water Standards.........................................................................................................81 Figure 68: Lead Leached from Brass Meters After 10 hr Stagnation Periods Compared to California Drinking Water Standards...............................................................................81 Figure 69: Copper Leached from Copper Pipes After 10 hr Stagnation Periods Compared to California Drinking Water Standards........................................................................................................83 Figure 70: Lead Leached from Copper Pipes After 10 hr Stagnation Periods Compared to California Drinking Water Standards..............................................................................84 Figure 71: Zinc Leached from Valve Loops After One-Week Recirculation Periods, Compared to CA Drinking Water Standards...........................................................................................85 Figure 72: Change in Zinc Concentrations in Water from the Valve Loops After One-Week Recirculation Periods...............................................................................................................86 Figure 73: Iron Leached from the Valve Loops After One-Week Recirculation Periods, Compared to California Drinking Water Standards..............................................................87 7 Figure 74: Iron Leached from the Valve Loops After One-Week Recirculation Periods, Compared to California Drinking Water Standards –...........................................................88 Figure 75: Lead Leached from the Valve Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards...............................................................89 Figure 76: Zinc Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards...............................................................90 Figure 77: Change in Zinc Concentrations in Water from Mortar-Lined Pipe Loops After One- Week Recirculation Periods.........................................................................................................................91 Figure 78: Iron Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards................................................................91 Figure 79: Lead Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards...............................................................92 Figure 80: Aluminum Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods, Compared to California Drinking Water Standards....................................93 Figure 81: Change in Aluminum Concentrations in Water from Mortar-Lined Pipe Loops After One-Week Recirculation Periods........................................................................................93 Figure 82: Interior Pipe Surfaces of the MWD1 Copper Pipe – Bottom Photo Shows 20X Magnification.........................................................................................................................................96 Figure 83: Interior Pipe Surfaces of the MWD2 Copper Pipe – Bottom Photo Shows 20X Magnification.........................................................................................................................................97 Figure 84: Interior Pipe Surfaces of the RO1 Copper Pipe – Bottom Photos Shows 20X Magnification of Top Surface (left) and 7X Magnification of Bottom Surface After Cleaning (right)...................................................................................................................................97 Figure 85: Interior Pipe Surfaces of the RO2 Copper Pipe – Bottom Photos Shows 20X Magnification of Top Surface (left) and 7X Magnification of Top Surface After Cleaning (right)...................................................................................................................................98 Figure 86: Interior Pipe Surfaces of the MWD1 Mortar-Lined Pipe – Top Photo Shows Top Pipe Surface; Lower Photo Shows 20X Magnification.........................................................................99 Figure 87: Interior Pipe Surfaces of the RO2 Mortar-Lined Pipe – Top Photo Shows Bottom Pipe Surface; Lower Photos Show 7X (left) and 20X (right) Magnification .................................100 Poseidon Resources Corporation Corrosion Pilot Study Executive Summary 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 8 Executive Summary Poseidon Resources Corporation has proposed an ocean desalination plant in Carlsbad, California that would produce up to 50 million gallons per day of potable drinking water for transmission to the City of Carlsbad water supply system. This seawater desalination process would produce treated water that varies in quality from the City of Carlsbad’s typical imported, treated surface water supply in several respects. The water quality of the reverse osmosis (RO) permeate would generally contain lower total dissolved solids, calcium, alkalinity, and disinfection byproduct formation potential, as well as higher concentrations of bromide, sodium, and chloride ions than the imported water supply. The purpose of the corrosion pilot study was to compare corrosion levels for conditioned desalinated seawater with those of the imported water from the Metropolitan Water District of Southern California (MWD). Two primary scenarios were evaluated over a six-month period for each water source, including: 1) Contact with typical household plumbing materials (copper pipe and brass water meters) using a flow regime representative of household consumption, and 2) Contact with typical distribution main materials (mortar-lined pipe and gate valves) at a representative, constant flow rate. Each of the three test system configurations was comprised of four independent pipe testing units. Two of the four units in each set received disinfected, conditioned RO permeate and the other two received MWD water. Due to the variability of results from corrosion and corrosion pilot testing, the pilot facility included a duplicate of each pipe unit for each water quality. Corrosion-related outcomes assessed in this pilot testing were as follows: 1) Release of metals (iron, zinc, lead, copper, aluminum) from pipes, valves, and meters; 2) Changes in water quality due to pipe corrosion or passivation; and 3) Changes in water aesthetics. The assessment of water quality changes was conducted by comparing observed metal concentrations and other water quality parameters for the desalinated seawater (RO permeate) and local surface water (MWD) circulating in pilot pipe loops and flow-through pipe systems, in addition to comparisons of corrosivity and scaling indices. Disinfected, stabilized seawater RO permeate was produced at the pilot site for comparison with MWD water in corrosion testing. RO permeate was passed through a calcium carbonate (calcite) filter to acquire alkalinity with a target of achieving a positive Langelier Index (LI) and Calcium Carbonate Precipitation Potential (CCPP). Chlorine and ammonia were then added to Poseidon Resources Corporation Corrosion Pilot Study Executive Summary 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 9 the RO permeate to form chloramines, and base was added to achieve a pH of 8.5. The pipe loops were drained and refilled weekly with conditioned RO permeate or MWD water acquired from a potable water line at the Encina power facility in the Carlsbad distribution system. Four types of distribution and household plumbing materials were tested, including new ¾- inch copper pipe commonly used in construction in Carlsbad, harvested home brass water service meters (brass meters) from the Carlsbad system, harvested 6-inch distribution system gate valves from the Carlsbad system, and new 4-inch cement mortar-lined pipe. The copper pipe test system was operated with 11-hour stagnation periods followed by 1 hour of water flow to represent diurnal household use patterns. The copper pipe system was sampled once per week following a 10-hour stagnation period. Water in the mortar-lined pipe and distribution system valve pipe loops was recirculated for one week to achieve flow rates typical in distribution system mains and gate valves. Effluent samples from the mortar-lined pipes and distribution system valves were collected after the water recirculated for one week. Conditioned RO permeate produced in this pilot study achieved the target of producing non- aggressive water with a positive LI and CCPP for comparison with MWD water. In general, RO permeate had a higher pH and higher chloride concentration, but lower alkalinity, TDS, and sulfate concentrations compared to MWD water. In the copper pipe test systems, copper leaching increased over time during the six-month study for both source waters. The MWD source water leached more copper than the conditioned RO permeate, potentially due to higher bicarbonate concentrations; however, the copper concentrations in both source waters were lower than California drinking water standards. Lead leaching into water in contact with the copper pipes decreased over time for both source waters due to passivation of the new lead-tin soldered joints connecting the copper pipe lengths. The RO permeate appeared to leach more lead than the MWD water; however, the lead levels were well below the 15 μg/L Action Level for both source waters and the difference in lead levels, while statistically significant, did not have regulatory or engineering significance. The average concentration of lead was 2.0 μg/L and 2.6 μg/L in the copper pipes supplied with MWD water and RO permeate, respectively. Copper leaching from brass meters into the water samples was stable during the pilot study for both source waters. As in the copper pipes, the RO permeate leached less copper than the MWD water. In contrast, lead leaching into water in contact with the brass meters decreased to a point at which lead leaching had stabilized (i.e., approximately 11 weeks into testing). As with the copper pipes, higher lead leaching was observed for the RO permeate (average of 9.2 μg/L) compared to MWD water (average of 6.6 μg/L). While the differences were statistically significant, the aggressive testing and sampling regime characterized by collection of samples directly from water in contact with the brass meters and freshly soldered joints did not result in lead levels exceeding the lead Action Level of 15 μg/L. Poseidon Resources Corporation Corrosion Pilot Study Executive Summary 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 10 In the mortar-lined pipe loops, initial RO permeate conditioning resulted in a negative LI for a few weeks. Water samples from this period reflected leaching of aluminum from the mortar- lined pipe, which highlights the importance of proper water conditioning. Once stable RO permeate was produced (in terms of corrosion indices, positive LI and CCPP), no significant leaching or statistically significant differences in concentration changes for calcium, aluminum, iron, zinc, or lead were observed for the two source waters. For the distribution system valves, little lead and zinc were leached by either water source. Excluding one valve that released much higher iron concentrations than the other three valves, average iron concentrations leached in both loops were lower than the secondary MCL for iron. There was no significant difference in lead, zinc, or iron leaching between the two source waters. In summary, conditioned RO permeate and MWD source waters in contact with typical household plumbing and distribution system materials did not exceed primary MCL standards, secondary MCL standards, and LCR Action Levels after the initial conditioning period. High levels with respect to regulatory limits were not observed despite the testing of “worst-case” conditions, such as recirculation of water in the mortar-lined pipe and distribution valve loops and sample collection directly from water in contact with brass meters in the copper pipe test systems. Since corrosion differs depending on the type of material, water that may be passivating for one material may be corrosive of another. This pilot study showed that MWD water, while protective of iron valves and mortar-lined pipe, may be more corrosive toward new copper pipe than conditioned RO permeate. The data indicates that the target conditioned water quality parameters in this study, including alkalinity, pH, calcium hardness, and TDS, would provide a water quality that is protective of all of the materials tested. Corrosion pilot testing of both consumer plumbing and distribution system materials provides important pieces of evidence that the proposed introduction of conditioned desalinated seawater supply is not likely to trigger new corrosion problems in the Carlsbad distribution system. Poseidon Resources Corporation Corrosion Pilot Study Introduction 1 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 11 Introduction Proposed Ocean Desalination Plant in Carlsbad Poseidon Resources Corporation’s (Poseidon) proposed ocean desalination plant in Carlsbad, California will take seawater from Cabrillo Power I LLC’s (Cabrillo) Encina Power Station cooling water system, produce up to 50 million gallons per day (MGD) of potable drinking water for transmission to the City of Carlsbad water supply system, and discharge the residual brine back into the power plant cooling water canal for dilution and discharge into the ocean. The desalination plant intake structure is proposed to be located downstream of the steam condensers. The desalination plant would use the heated condenser cooling water as source water. The pretreatment system will be either gravity media filtration or microfiltration (MF). The reverse osmosis (RO) desalination process will be a single-pass design using high-rejection seawater membranes. Product water from the RO process will require chemical conditioning prior to delivery to the distribution system to increase hardness and alkalinity and thereby reduce the water’s corrosion potential. Lime will be used for post-treatment stabilization of the water. The final product water must be disinfected prior to delivery to the distribution system. Chlorine, in the form of sodium hypochlorite, will be added as a disinfectant to meet California DHS water quality standards for potable water disinfection and to control biological growth in the transmission pipeline between the desalination plant and the receiving reservoirs in the distribution system. Ammonia will be added at the plant to convert the post- disinfection free chlorine to chloramines prior to distribution. Integration of Desalinated Seawater This seawater desalination process will produce treated water that varies from the City of Carlsbad’s typical imported, treated surface water supply in several respects. The water quality may be markedly superior to imported water supplies in terms of lower total dissolved solids (TDS) and disinfection byproduct (DBP) concentrations. Because the treated water is derived from seawater and employs an RO membrane treatment process, the treated water will contain higher concentrations of bromide, sodium, and chloride ions than the imported water supply, and will be essentially devoid of alkalinity. While it is accepted that alkalinity will have to be added to render the RO permeate less corrosive, the composition of the treated water will unavoidably differ from the current imported water supply. The purpose of this study was to quantify the potential corrosion-related effects of transitioning the local water supply from the current imported water supply to a supply using conditioned seawater RO permeate. The results were interpreted in the context of 1) satisfying the regulatory requirements of the Lead and Copper Rule (LCR), and 2) comparing potential corrosion outcomes with the imported water supply. Poseidon Resources Corporation Corrosion Pilot Study Project Description 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 12 Project Description Purpose and Objectives of the Corrosion Study The purpose of this corrosion pilot study was to compare corrosion levels of several materials in contact with conditioned, desalinated seawater and corrosion levels associated with imported water from the Metropolitan Water District of Southern California (MWD). The following materials and flow schemes were tested: 1) Household plumbing materials using a flow regime representative of household consumption (constant flow of 0.4 ft/s for approximately 60 minutes followed by a diurnal stagnation period of 11 hours), and 2) Distribution main materials (including distribution system gate valves and cement mortar-lined pipe) at representative, constant flow rates (approximately 0.4-1.0 ft/s). Corrosion-related outcomes assessed included: release of metals (iron, zinc, lead, copper, and aluminum) from pipes, valves, and meters; and changes in water quality due to pipe corrosion or passivation. This study compared corrosion-related outcomes for desalinated water with treated water from MWD in the Carlsbad distribution system. Poseidon Pilot Desalination Treatment Facility Poseidon operates a seawater desalination pilot treatment facility at Cabrillo’s Encina power station in Carlsbad, California. The raw source water for the desalination pilot facility was post-condenser cooling water from the power plant. This once-through power generation station withdraws cooling water from the Pacific Ocean via the Agua Hedionda Lagoon. The pilot desalination treatment system intake withdrew warm water from a small lagoon that was located at the end of the power plant discharge tunnel. The power plant cooling water discharge was typically 5 to 10oF warmer than the ocean seawater. The TDS concentration of the influent seawater varied between 33,000 milligrams per liter (mg/L) and 34,500 mg/L, and averaged 33,500 mg/L. Seawater was conveyed to a feed storage tank, from which it was pumped to the pilot plant pretreatment systems. The two pretreatment systems being tested were the Parkson’s two- stage, continuous backwash sand media filtration system and the Hydranautics HydraSub© immersed MF. The two pretreatment systems were operated independently and each typically produced between 40 and 45 gpm of filtered water. During the course of the corrosion study, both pretreatment systems were in operation at different times. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 13 The pilot RO system consisted of two 4-element pressure vessels in series. The seawater RO membrane elements tested were 8-inch diameter, high salt-rejection units provided by Hydranautics. The RO system was designed to run in a range of 45 to 55 percent recovery and typically operated at 50 percent recovery. The pilot desalination plant was operated continuously since August 2003. The quality of the produced permeate was consistently high - TDS concentrations of 200 and 300 mg/L at an RO system feed pressure between 780 and 900 psi. The two pretreatment systems produced filtered seawater with a silt density index (SDI) of less than 4 and turbidity below 0.1 nephlometric turbidity units (NTU). Throughout most of the corrosion study, a sulfuric acid feed system was operated ahead of the RO membranes to reduce the pH of the RO permeate to near pH 5.5. This lower pH aided in the dissolution of calcium from the calcium carbonate (calcite) filter, which conditioned the RO permeate by adding alkalinity and calcium before the RO permeate was disinfected and subsequently introduced into the various pipe test systems. Figure 1 is a schematic of the Poseidon desalination pilot plant and Figure 2 is a photograph of the RO elements in relation to the corrosion testing facility. The tents in the background cover the corrosion testing facility. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 14Figure 1: Schematic of Poseidon Pilot Desalination Plant Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 15 Figure 2: Poseidon Resources Pilot Facility Poseidon Corrosion Pilot Facility Overview The Poseidon Corrosion Pilot Facility was constructed by McGuire Malcolm Pirnie adjacent to Poseidon’s desalination pilot treatment facility. The plan view of these two facilities is shown in Figure 3. The corrosion pilot facility was configured to fit in a 15-foot x 50-foot pad and was covered by a sunshade. The site had dedicated power, influent water lines, and effluent discharge piping for the pilot pipe-testing configurations. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 16 Figure 3: Plan View of the Corrosion Investigation Pilot Facility TrailerPretreat #2 R.O. Unit Pretreat #1 Park Waste Tank Not to Scale 15’50’ Parking Lot Corrosion Pilot Area N Three different configurations of pipe systems were constructed by McGuire Malcolm Pirnie to compare the corrosivity of RO permeate and the imported MWD water supply on a variety of distribution system and residential plumbing materials: 1) Flow-through systems consisting of new copper pipe and system-harvested residential brass water service meters (brass meters), 2) Recirculation systems of new cement mortar-lined steel pipe, and 3) Recirculation systems containing harvested, cast iron distribution system gate valves. Each of the three test system configurations had four independent pipe testing loops or lengths. Two of the four loops or lengths in each set received conditioned RO permeate and the other two received MWD water. MWD water was acquired from a connection to a potable water line at the Encina power facility. Due to the inherent imprecision of corrosion and corrosion pilot testing, the pilot facility included a duplicate of each pipe loop or test system for each water quality. For the consumer plumbing pipe testing, new copper pipe (twelve 10-foot lengths of ¾-inch type L pipe) and eight harvested ¾-inch cast brass water service meters were provided by the City of Carlsbad. The brass water service meters harvested from homes were approximately Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 17 15 to 20 years’ old and were due for replacement. For the distribution system tests, the City provided eight 10-foot segments of new 4-inch diameter cement mortar-lined steel pipe and eight harvested 6-inch cast iron distribution system gate valves. Poseidon provided calcite filters used to stabilize the RO permeate by adding alkalinity and calcium to the water. Figure 4 shows the layout of the three different kinds of pilot loop types: consumer plumbing (copper pipe and household water meters), distribution main (mortar-lined pipe), and distribution gate valves. Figure 5 is a photograph captured during construction of the test systems, showing an overall view of the corrosion pilot plant. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 18Figure 4: Pipe Test System Layout N2,500 GallonR.O. StorageTank2,500 GallonR.O. StorageTankRO Copper Pipe/Brass Meter Test Systems (30 ft.+)4” Mortar-Lined Pipe (MWD)4” Mortar-Lined Pipe (RO)4” Mortar-Lined Pipe (MWD)4” Mortar-Lined Pipe (RO)200 GallonTanks6” PVC w/ Valves (MWD)6” PVC w/ Valves (RO)6” PVC w/ Valves (MWD)6” PVC w/ Valves (RO)MWD Water Source= Manually Operated Valve= Transfer or Recirculation PumpApproximately 50’Approximately 15’20’ Pipe LengthsChemical AdditionMWD Copper Pipe/Brass Meter Test Systems (30 ft.+)E= Electrically Operated ValveEETo Waste= RO Permeate= MWD WaterNN2,500 GallonR.O. StorageTank2,500 GallonR.O. StorageTankRO Copper Pipe/Brass Meter Test Systems (30 ft.+)4” Mortar-Lined Pipe (MWD)4” Mortar-Lined Pipe (RO)4” Mortar-Lined Pipe (MWD)4” Mortar-Lined Pipe (RO)200 GallonTanks6” PVC w/ Valves (MWD)6” PVC w/ Valves (RO)6” PVC w/ Valves (MWD)6” PVC w/ Valves (RO)MWD Water Source= Manually Operated Valve= Manually Operated Valve= Transfer or Recirculation Pump= Transfer or Recirculation PumpApproximately 50’Approximately 15’20’ Pipe LengthsChemical AdditionMWD Copper Pipe/Brass Meter Test Systems (30 ft.+)E= Electrically Operated ValveEE= Electrically Operated ValveEEEETo Waste= RO Permeate= MWD Water Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 19 Figure 5: Corrosion Pilot Plant During Construction To ensure that sufficient RO permeate was available for continuous testing of the flow- through systems, two 2,500-gallon storage tanks were used. The storage tanks could be isolated from each other and from the corrosion pipe systems. Fresh permeate was conditioned in one of the storage tanks while the other tank fed the flow-through copper pipe test systems. RO Permeate Source Water Conditioning Disinfected, stabilized seawater RO permeate was produced at the pilot site. RO permeate was passed through a calcite filter to pick up alkalinity and calcium, with a target of achieving a positive Langelier Index (LI) and Calcium Carbonate Precipitation Potential (CCPP). Figure 6 presents a schematic of the RO permeate storage and conditioning process. Figure 7 is a photograph of the chemical addition location of parallel empty bag filter housings. Chemicals used to adjust pH and add chloramines were added in a batch mode process to the recirculating 2,500-gallon storage tanks. Volumetric aliquots were added to the storage tanks using in-line empty bag filters shown in Figure 7. The 2,500-gallon tank was recirculated at approximately 50 gpm with water entering at the top of the tank and leaving at the bottom. The inlet and outlet placement was selected to ensure thorough mixing. Initial tests of chlorine mixing revealed that the tank was fully mixed in less than 30 minutes. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 20Figure 6: Schematic of RO Permeate Conditioning Process RO Waterto Copper LoopsRO StorageFor UpstreamPilot Plant(500 Gallons)Calcite FilterCalcite FilterConditionedR.O. WaterHoldingTank #2(2500 gal)ConditionedR.O. WaterHoldingTank #2(2500 gal)ConditionedR.O. WaterHoldingTank #1(2500 gal)ConditionedR.O. WaterHoldingTank #1(2500 gal)ROV1ROV2FV1FV2BF2P2V1P2V3P2V2P2V4DRAINAirReleaseBF1P1V1P1V3P1V2P1V4DRAINAirReleaseTank #2SampleTank #2SampleP1P2Tank #1SampleTank #1SampleSampleInSampleInSampleOutSampleOutRecirculation LineRecirculation LineT1V1T2V1RO Water to 200 Gallon Corrosion Loop TanksTo 1500 Gallon Waste TankWV3V200Existing RO UnitROV2ROV1RO Waterto Copper LoopsSulfuric Acid Feed= Manually Operated Valve= Sample Point= Sample Point= 50 gpm Recirculating/Transfer Pump= Bag Filter for Chemical Addition Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 21 Figure 7: Chemical Addition Location – Empty Bag Filters Copper Pipe/ Brass Water Meter Flow-Through Systems The test of copper pipe and home water service meters used three 10-foot lengths of new Type L ¾-inch copper pipe connected by copper couplings using 50/50 lead/tin solder as shown in the schematic in Figure 8 and construction details in Figure 9. The copper pipe lengths were mounted to a frame on an incline of ¾-inch per foot of pipe. Two brass meters were placed near the downward sloping end of the rise. A sampling point further down slope of the brass meters enabled the collection of water that had only been in contact with the water meters as part of the sampling procedure. The copper pipe and brass meters were part of flow-through systems operating for 1 hour at a flow rate of 0.5 gpm, followed by an 11-hour stagnant period, simulating a diurnal water cycle typical of household use. Water velocities in the copper pipe were approximately 0.4 ft/s. Figure 10 and Figure 11 show the copper pipe flow-through systems and a detail of the brass water meters, respectively. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 22Figure 8: Schematic of Consumer Copper Plumbing Flow-Through Systems ConditionedWaterHoldingTank #2(2500 gal)RO UnitCalciteFilterMWDWaterSourceConditionedWaterHoldingTank #1(2500 gal)ROL2Consumer Plumbing (Flow-through)Copper Pipe, 3/4”, Type “L”,Copper Fittings, 50/50 Lead/Tin Solder, PVC for all Connection PipingETo DistributionLoopsCast BrassWater MeterCast BrassWater Meter30’ (All) ¼” per foot riseCuMWD1Cast BrassWater MeterCast BrassWater MeterCuMWD2Cast BrassWater MeterCast BrassWater Meter30’ (All) ¼” per foot riseCuRO1Cast BrassWater MeterCast BrassWater MeterCuRO2To WasteTo WasteTo WasteTo WasteE= Sample Point= Flow control valve = Manually Operated Valve= Check Valve= Rotameter0.3 to 3.5 gpmMWD1Chemical AdditionChemical Addition= Electrical Valve(Timer controlled)EP1P2ROL1FV1FV2T1V2T1V1T2V2T2V1P2V2P2V4P1V2P1V4T1T2ROV1ROV2= 50 gpm Recirculating/Transfer PumpSulfuric Acid FeedConditionedWaterHoldingTank #2(2500 gal)RO UnitCalciteFilterMWDWaterSourceConditionedWaterHoldingTank #1(2500 gal)ROL2ROL2Consumer Plumbing (Flow-through)Copper Pipe, 3/4”, Type “L”,Copper Fittings, 50/50 Lead/Tin Solder, PVC for all Connection PipingEETo DistributionLoopsCast BrassWater MeterCast BrassWater Meter30’ (All) ¼” per foot riseCuMWD1CuMWD1Cast BrassWater MeterCast BrassWater MeterCuMWD2CuMWD2Cast BrassWater MeterCast BrassWater Meter30’ (All) ¼” per foot rise30’ (All) ¼” per foot riseCuRO1CuRO1Cast BrassWater MeterCast BrassWater MeterCuRO2CuRO2To WasteTo WasteTo WasteTo WasteEE= Sample Point= Sample Point= Flow control valve= Flow control valve = Manually Operated Valve= Check Valve= Check Valve= Rotameter0.3 to 3.5 gpm= Rotameter0.3 to 3.5 gpmMWD1MWD1Chemical AdditionChemical Addition= Electrical Valve(Timer controlled)E= Electrical Valve(Timer controlled)EEP1P2ROL1ROL1FV1FV2T1V2T1V1T2V2T2V1P2V2P2V4P1V2P1V4T1T1T2T2ROV1ROV2= 50 gpm Recirculating/Transfer PumpSulfuric Acid Feed Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 23 Figure 9: Construction Details of Consumer Plumbing Flow-Through Systems Figure 10: Consumer Plumbing Flow-through System ¾” Brass Water Meter ¾” Brass Water Meter ¾” Brass Water Meter (8 meters required) Meters are equipped with ¾” male pipe thread Adapters on both ends ¾” Brass Water Meter ¾” Brass Water Meter ¾” Brass Water Meter 10 ft.10 ft.10 ft. ¾” MPT x ¾” Sweat Copper Male Connector ¾” Sweat x ¾” Sweat Copper Coupling ¾” FPT x ¾” Sweat Copper Female Connector ¾” Copper Pipe,Type “L” (12 each, 10 ft. long) ¾” FPT x ¾” Slip PVC Female Connector Source Water to be tested Effluent to Waste ¾” FPT x ¾” FPT Brass Coupling PVC PipingPVC Piping Note: All joints sweated with 50/50 solder Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 24 Figure 11: Consumer Plumbing Brass Meters The copper pipe and brass meter testing was designed to compare the copper and lead leaching characteristics of the RO permeate and MWD supplies. A sample of the stagnant water in contact with the brass meters was analyzed for copper and lead. Brass meters were used to demonstrate corrosion effects from brass meters as well as brass household plumbing fixtures. The remaining water held in the copper pipe was then sampled for metals and other water quality parameters to determine copper and lead leaching, disinfectant residual, and other changes. This sampling regime guaranteed a “worst-case” determination of copper and lead leaching in the brass meters since water collected was directly in contact with the brass meters during the stagnant period. These results are not directly comparable to actual home sampling under the LCR since LCR water samples would contain water both in contact with and not in contact with brass meters and fixtures. Mortar-Lined Pipe Recirculation Systems The cement mortar-lined steel pipe loops consisted of two 10 foot lengths of new 4-inch cement mortar-lined pipe connected in series, with a PVC line to return the flow to a black 200-gallon polyethylene tank (Figure 12). Construction details of the joined mortar-lined pipe segments are shown in Figure 13. The flow rate of the mortar-lined pipe loop was set at 39 Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 25 gpm (corresponding to a water velocity of approximately 1 foot per second) using a flow control valve located after the sample point at the end of the loop and before the point at which the loop water was discharged into the recirculation tank. Water was recirculated in the loop for one week, sampled, and then replaced with fresh RO permeate or MWD water. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 26Figure 12: Schematic of Mortar-lined Pipe Loops MWD Water SourceTo WasteTo WasteTo WasteTo Waste1 1/2” PVC Return LineMLROD1MLROD1MLROD2MLROD2= Flow control valve= Flow control valveMWD1MWD1= Manually Operated Valve= Sample Point= Sample Point4” Mortar-Lined Pipe, 2-10 Ft. Lengths4” Mortar-Lined Pipe, 2-10 Ft. Lengths1 1/2” PVC Return Line1 1/2” PVC Return LineMLMWD2MLMWD2MLMWD1MLMWD11 1/2” PVC Return LineRemovable Flowmeter(5 to 50 gpm)= 200 Gallon Black HPDE Tank = 50 gpm Recirculating Pump4” Mortar-Lined Pipe, 2-10 Ft. Lengths4” Mortar-Lined Pipe, 2-10 Ft. LengthsConditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 27 Figure 13: Construction Details of Mortar-lined Pipe Loops The purpose of the mortar-lined pipe test was to compare the stability of the new mortar-lined pipe exposed to disinfected, stabilized RO permeate with MWD water. Calcium and aluminum leaching into the water from the mortar lining was quantified. Metal leaching into the water from any exposed metal surfaces was also determined. Figure 14 is a photograph of two of the four cement mortar-lined steel pipe loops. The 200- gallon recirculation tanks are in the background and one of the two 2,500-gallon RO permeate storage tanks can be seen behind the 200-gallon tanks. 4” Steel, Mortar-Lined Pipe (8 each, 10 ft. long) Return Line To Tank Source Water to be tested 10 ft.10 ft. 6” PVC Flange reduced to 1 1/2” PVC loop piping (8 required) 1 1/2” PVC 4” Steel, Mortar-Lined Pipe (8 each, 10 ft. long) Return Line To Tank Source Water to be tested 10 ft.10 ft.10 ft.10 ft. 6” PVC Flange reduced to 1 1/2” PVC loop piping (8 required) 1 1/2” PVC Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 28 Figure 14: Mortar-lined Pipe Loops Distribution System Valve Recirculation Systems Six-inch diameter cast iron distribution system gate valves harvested from the Carlsbad distribution system were also tested in pipe loops that otherwise consisted of non-corrosive PVC pipe. Two valves were placed in each loop, and water was recirculated through the loop at a constant flow of approximately 39 gpm (corresponding to a water velocity of approximately 0.4 ft/s). As in the mortar-lined pipe loops, water in the distribution valve pipe loops was recirculated for one week, sampled, and then replaced with fresh water. The purpose of the valve pipe loop test was to compare the corrosivity of the test waters on the valves and any subsequent release of metals (including iron, zinc, copper, and lead) from Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 29 the valve materials. Additional water quality parameters were tested to characterize the corrosivity of the water. A schematic of the valve loops is presented in Figure 15 and construction details are provided in Figure 16. Figure 17 is a photograph of the valve loop system. Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 30Figure 15: Schematic of Distribution Valve Loops MWD Water SourceTo WasteTo WasteTo WasteTo Waste1 1/2” PVC Return LineVRO1VRO1VRO2VRO2= Flow control valve= Flow control valveMWD1MWD1= Manually Operated Valve= Sample Point= Sample Point1 1/2” PVC Return Line1 1/2” PVC Return LineVMWD2VMWD2VMWD1VMWD11 1/2” PVC Return LineRemovable Flowmeter(5 to 50 gpm)= 200 Gallon Black HPDE Tank = 50 gpm Recirculating PumpConditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 31 Figure 16: Distribution Valve Loop Construction Details 4” Steel, Mortar-Lined Pipe (8 each, 10 ft. long) Return Line To Tank Source Water to be tested 10 ft.10 ft. 6” PVC Flange reduced to 1 1/2” PVC loop piping (8 required) 1 1/2” PVC 4” Steel, Mortar-Lined Pipe (8 each, 10 ft. long) Return Line To Tank Source Water to be tested 10 ft.10 ft.10 ft.10 ft. 6” PVC Flange reduced to 1 1/2” PVC loop piping (8 required) 1 1/2” PVC Figure 17: Distribution Valve Loops Poseidon Resources Corporation Corrosion Pilot Study 2 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 32 Figure 18 shows the exterior of the corrosion pilot plant under the canopy. Figure 19 is a view of the recirculation test systems and on-site analysis space inside the canopy. Figure 18: Exterior of the Corrosion Pilot Facility Figure 19: Recirculation Test Systems Under Canopy Poseidon Resources Corporation Corrosion Pilot Study Experimental Plan 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 33 Experimental Plan Source Waters Disinfected, stabilized seawater RO permeate and the local tap water source were used for the corrosion testing comparison. The local tap water came from MWD water supplied to the Carlsbad distribution system. MWD water was obtained directly from a water line located at the pilot site. This water was treated at the Robert A. Skinner Filtration Plant in Riverside County, disinfected with chloramines, and supplied to the San Diego County Water Authority and the City of Carlsbad. Disinfected, stabilized seawater RO permeate was produced at the pilot site. During a typical once-a-week operational day, one 2,500-gallon tank containing RO permeate was conditioned to be used in the pipe loops that week. The RO permeate was conditioned by 1) adding alkalinity and calcium, 2) disinfecting through the addition of chlorine and ammonia to form chloramines, and 3) adjusting the pH using a strong base. In the first few weeks of pilot testing, the following water quality goals for conditioned RO permeate were found to yield a positive LI and CCPP: pH of 8.5, alkalinity of 45-65 mg/L as CaCO3, hardness of 40-50 mg/L as CaCO3, and TDS of 300-350 mg/L. In addition, a chloramine residual of 2 to 2.5 mg/L with a 5:1 mass ratio of chlorine to ammonia-N was established for the conditioned permeate. To ensure that sufficient RO permeate was available for diurnal flow-through testing of the copper pipe test systems, two 2,500-gallon storage tanks were used, with the newly conditioned tank alternating from week to week. The storage tank supplying the corrosion pipe loops was isolated when the other tank was being conditioned. System Operations One 2,500-gallon tank was filled each week with RO permeate from the seawater desalination pilot facility that was passed through a calcite filter. After the first few weeks of operation in which insufficient calcite was dissolved, it was identified that RO water should be adjusted to a pH of 5.5 to maximize calcite dissolution while minimizing base needed to reach the target pH of 8.5 in the finished water. In addition, a larger particle size of calcite in the calcite filter was necessary to minimize pressure problems and fine particulate carryover into the 2,500-gallon tank. By 7/13/05, the initial operational challenges were overcome. On-site analysis of hardness assisted in the confirmation that positive LI and CCPP values were reached during filling of the 2,500-gallon tank. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 34 After passing through the calcite filter, the RO permeate was disinfected. An unanticipated finding in a previous study investigating DBP formation in disinfected RO permeate was the rapid decay of combined chlorine residual (chlorine plus ammonia) in the disinfected RO permeate compared to chloraminated MWD water (Loveland and Means 2005; Chao et al 1994). An accelerated residual decay had been observed in an RO permeate during bench- scale simulated distribution system (SDS) testing of DBP formation in 2003. When an initial free chlorine residual of 2.5 mg/L was converted to chloramines with ammonium hydroxide, the combined residual decayed from 2.5 mg/L to 1.5 mg/L after only four hours’ contact time. The observed residual decay appeared to be a result of bromamine formation and an accelerated degradation of bromamines. In the same SDS study, it was shown that an initial 4.5 mg/L free chlorine residual in seawater RO permeate (“superchloramination”) when converted to chloramines would yield a stable residual of 2.0 – 2.5 mg/L with a 5:1 Cl2:NH3-N ratio within several of contact time. This residual decay phenomenon was also observed during initial permeate conditioning phase of this study. The total chlorine concentration necessary to yield a stable 2.0 - 2.5 mg/L residual in RO permeate was determined to be approximately 4.0 mg/L. To reach a target chlorine concentration of 4 mg/L, 0.65 L of Clorox™ bleach (containing approximately 6% Cl2) was added to the storage tank and allowed to mix for 1 hour. The chlorine residual in the tank was then measured and 35 ml of 30% liquid ammonium hydroxide (30% w/w, EMD Chemicals) was added to the tank to convert the free chlorine residual to a chloramines residual. The targeted TDS for conditioned permeate was 300 to 350 mg/L. When the conductivity and TDS were high (i.e., above 800 microseimens per centimeter (μS/cm) and 400 mg/L, respectively), the chloramines residual dropped more rapidly and additional chlorine and ammonia were added to increase the chloramines residual. After the 2,500-gallon tank was disinfected, the pH was adjusted by adding liquid sodium hydroxide (50% w/w, FisherBrand) to reach a target of 8.5. The final alkalinity in the conditioned permeate was influenced by the amount of calcite dissolution and the amount of chemicals added for disinfection and pH adjustment of the RO permeate before being introduced into the pipe systems. The targeted alkalinity range was 45-65 mg/L as CaCO3. Following 2 to 2.5 hours of recirculation of the chloraminated water, the conditioned RO permeate was analyzed for pH, total chlorine, and total ammonia. Once a stable chloramine residual was observed, the conditioned water was used to fill the mortar-lined pipe and distribution valve loops. The newly conditioned 2,500-gallon tank was also placed online to feed the copper pipe when the diurnal flow-through periods began. All chemical addition and resulting water quality parameters of the conditioned RO permeate are summarized in Table 1. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 35 Table 1: Final RO Permeate Water Quality Week Chlorine Added Ammonia Added Acid Added Base Added Final pH Final Chlorine Final Ammonia Cl2:NH3-N (L) (ml) (ml) (ml) (units) (mg/l) (mg/l) 1 0.63 33 35 0 8.67 2.2 0.47 4.7 2 0.65 35 0 0 8.82 1.77 0.31 5.7 3 N/A N/A N/A N/A N/A N/A N/A N/A 4 1.41 77 0 0 8.44 2.01 0.52 3.9 5 0.65 35 0 328 8.15 2.1 0.47 4.5 6 1.8 75 0 120 8.61 2.4 0.54 4.4 7 0.65 35 0 100 8.75 1.73 0.41 4.2 8 1.8 79 0 230 8.5 2.36 0.62 3.8 9 0.65 35 0 0 8.5 1.96 0.47 4.2 10 0.79 35 0 20 8.58 1.82 0.38 4.7 11 0.7 35 0 200 8.57 1.88 0.46 4.1 12 0.65 35 0 0 8.52 1.82 0.40 4.6 13 0.65 35 0 50 8.67 1.99 0.41 4.9 14 0.65 35 0 0 8.48 1.72 0.38 4.5 15 0.65 35 0 260 8.49 2.32 0.48 4.8 16 0.65 35 0 33 8.49 2.26 0.50 4.5 17 0.65 35 0 30 8.63 1.83 0.46 4.0 18 0.65 35 0 12 8.49 2.56 0.58 4.4 19 0.65 35 0 92 8.48 2.23 0.50 4.4 20 0.65 35 0 100 8.71 2.4 0.55 4.3 21 0.65 35 0 95 8.54 1.8 0.37 4.8 22 0.65 35 0 45 8.54 2.19 0.52 4.2 23 0.65 35 0 40 8.65 2.52 0.52 4.9 24 0.65 35 0 0 8.55 2.74 0.55 5.0 25 0.65 35 0 125 8.52 3 0.54 5.6 In normal weekly operations, the pipe loop recirculation was stopped after the RO permeate was conditioned and the weekly samples were collected from the pipe loops. The 200-gallon tanks were then drained and refilled with the conditioned RO permeate or fresh MWD water. The 200-gallon tanks for the valve and mortar-lined pipe loops were filled with 100 gallons and 150 gallons of water, respectively. Once filled, the recirculation pumps were restarted. Once all of the pipe loops were filled with or placed online to run using the newly conditioned RO permeate, the RO permeate in the old 2,500-gallon storage tank was drained in preparation for the next week’s fill of fresh RO permeate. One exception occurred in week 3, when RO permeate was not available from the Poseidon RO unit. During that week, the copper pipes were stagnant for a week and the mortar-lined pipes and distribution system valves were run for an extra week using water from the previous week (i.e., water was recirculated for 2 weeks instead of 1 before being replaced). Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 36 Sample Collection and Analysis Sample Locations Sampling locations were selected to reflect various steps of the RO permeate conditioning and disinfection process, MWD influent water, and effluent locations for each of the pipe loops. The specific sample sites and assigned names are listed in Table 2. Table 2: Sampling Location Key Location ID Sampling Location MWD1 MWD influent water T1 RO water in Tank 1 after pH adjustment and disinfection T2 RO water in Tank 2 after pH adjustment and disinfection CuMWD1P Copper pipe loop #1 fed MWD water from pipe CuMWD1M Copper pipe loop #1 fed MWD water from meter CuMWD2P Copper pipe loop #2 fed MWD water from pipe CuMWD2M Copper pipe loop #2 fed MWD water from meter CuRO1P Copper pipe loop #1 fed RO permeate from pipe CuRO1M Copper pipe loop #1 fed RO permeate from meter CuRO2P Copper pipe loop #2 fed RO permeate from pipe CuRO2M Copper pipe loop #2 fed RO permeate from meter MLMWD1E Mortar-lined pipe loop #1 fed MWD water effluent from pipe MLMWD1R Mortar-lined pipe loop #1 fed MWD water after replacement MLMWD2E Mortar-lined pipe loop #1 fed MWD water effluent from pipe MLMWD2R Mortar-lined pipe loop #1 fed MWD water after replacement MLRO1E Mortar-lined pipe loop #1 fed MWD water effluent from pipe MLRO1R Mortar-lined pipe loop #1 fed MWD water after replacement MLRO2E Mortar-lined pipe loop #1 fed MWD water effluent from pipe MLRO2R Mortar-lined pipe loop #1 fed MWD water after replacement VMWD1 Valves in PVC pipe loop #1 fed MWD water VMWD2 Valves in PVC pipe loop #2 fed MWD water VRO1 Valves in PVC pipe loop #1 fed RO permeate VRO2 Valves in PVC pipe loop #2 fed RO permeate Even though there was an 11-hour stagnation period, water samples were collected from the copper pipes and brass meters approximately 10 hours after the pipes were filled to ensure that the samples were obtained before the flow began. The copper pipe system was sampled after 10 hours of stagnation by collecting one small sample (approximately 100 mL) from water in contact with the brass water meters, and then another sample (1 L) from the 30 feet of copper pipe. Prior to sample collection for the first flush brass water meter samples, the sampling port and hose were briefly flushed to purge the volume in the sample tap. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 37 Mortar-lined pipe and distribution system valve pipe loops each had a sampling port on the PVC pipe discharging back into the recirculation tanks. Effluent samples were collected after the water circulated for one week. Sample ports were flushed for at least one minute before samples were collected. Sample contamination was avoided by practicing clean sampling techniques. Samples were collected by directly filling clean sample bottles from the sampling ports without contacting the interior surfaces of the bottles. Sample bottles for field analysis were cleaned with distilled water then rinsed with sample water three times before samples were collected. New bottles were used for each laboratory analysis. Replacement samples of the mortar-lined pipe loops were collected during weeks 1 through 8 to address the concern that the water volume remaining in the pipes after the tanks were drained might change the influent water quality. This concern arose because the water volume left in the 4-inch pipe after gravity draining comprised approximately 10% of the loop volume. When new water was added to the loops, the influent water was mixed with a volume of 10% of the week-old water. Hence, additional samples (“replacement samples”) were collected from the loops after the weekly fresh water refill to determine if this carryover was significant. Note that the “replacement” water in the mortar-lined pipe was mixed at 39 gpm for 10 minutes before samples were collected. For the first eight weeks, replacement samples were used as the baseline water quality for the mortar-lined pipe loops. After week 8 (beginning in week 9, 8/17/05), water quality in the source waters and replacement samples converged and source water quality (MWD1, T1, or T2) was used as the baseline for comparison with the effluent samples. This convergence of source water and replacement water qualities arose from stabilization of the conditioned RO permeate and equilibration of the pipe and valve surfaces during the first eight weeks. Analyses and Monitoring Frequency Samples were collected weekly for field and lab analysis. The chemical and physical analytes that were routinely measured on-site included temperature, pH, turbidity, conductivity, alkalinity, total chlorine, total ammonia, nitrite, and dissolved oxygen. Samples for calcium, metals (iron, zinc, aluminum, lead, and copper), sulfate, chloride, and TDS were sent to the Del Mar Analytical Laboratory (Irvine, CA) for analysis. The frequency of each analysis is listed in Table 3. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 38Table 3: Sampling Locations and Frequencies Sampling Location Temp pH Turb Cond AlkTotal Cl2Total NH3NO2-Dissolved O2Ca2+AlFe, Zn, Pb Pb, Cu TDSSO42-Cl-Macro VisualMWD1 (MWD Water) 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM 2xM 1M 1M 1M -T1 (Conditioned Water) 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM 2xM 1M 1M 1M -T2 (Conditioned Water) 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM 2xM 1M 1M 1M -MLMWD1E (EFFLUENT) 1W1W1W1W1W1W1W1W 1M 1W1W2xM - 1M 1M 1M EndMLMWD2E 1W1W1W1W1W1W1W1W 1M 1W1W2xM - 1M 1M 1M EndMLRO1E 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM - 1M 1M 1M EndMLRO2E 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM - 1M 1M 1M EndMLMWD1R (REPLACEMENT)1W1W1W1W1W1W1W1W 1M 1W1W2xM - 1M 1M 1M EndMLMWD2R 1W1W1W1W1W1W1W1W 1M 1W1W2xM - 1M 1M 1M EndMLRO1R 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM - 1M 1M 1M EndMLRO2R 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W 1W 2xM - 1M 1M 1M EndVMWD1 1W1W1W1W1W1W1W1W 1M 1W - 2xM - 1M 1M 1M EndVMWD2 1W1W1W1W1W1W1W1W 1M 1W - 2xM - 1M 1M 1M EndVRO1 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W - 2xM - 1M 1M 1M EndVRO2 1W 1W 1W 1W 1W 1W 1W 1W 1M 1W - 2xM - 1M 1M 1M EndCuMWD1M (METER) -------- - ---2xM---EndCuMWD2M -------- - ---2xM---EndCuRO1M -------- - ---2xM---EndCuRO2M -------- - ---2xM---EndCuMWD1P (PIPE) 1W 1W 1W 1W 1W 1W 1W 1W 1M 1M - - 2xM 1M 1M 1M EndCuMWD2P 1W1W1W1W1W1W1W1W 1M 1M - - 2xM1M 1M 1M EndCuRO1P 1W 1W 1W 1W 1W 1W 1W 1W 1M 1M - - 2xM 1M 1M 1M EndCuRO2P 1W 1W 1W 1W 1W 1W 1W 1W 1M 1M - - 2xM 1M 1M 1M End Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 39 Revised Monitoring Frequency During the course of the pilot testing, a few changes were made to the sampling frequency to reflect findings of the study. The changes are described below: 1) As of week 10 (8/24/05), metals analysis (Fe, Zn, Pb) of water in contact with the mortar-lined pipes was reduced from two times-per-month to one time-per-month, since no significant leaching of metals was observed from the mortar-lined pipes. 2) After 8 weeks, results showed that the influent samples were similar to the replacement samples; therefore, replacement sample collection was stopped at week 9 (8/17/05). 3) As of week 8 (8/10/05), samples of the source waters for TDS, sulfate, and chloride were collected weekly rather than monthly since variability was observed. Analytical Methods and Reporting Limits Chlorine, ammonia, nitrate, and dissolved oxygen were measured in the field using a Hach DR/890 colorimeter. pH was measured using a Hach SensION™ pH platinum series probe. An Oakton 300 meter and probe was used to measure conductivity. Turbidity analysis was performed using a Hach 2100P turbidimeter. Alkalinity was measured using a Hach AL-DT test kit with digital titrator. The Hach DR/890 colorimeter is factory calibrated for the following analytes: total chlorine, total ammonia, nitrite, and dissolved oxygen. Pre-programmed calibration curves were based on extensive testing by Hach. Initial tests of the colorimeter’s accuracy were performed using standard solutions of ammonia and nitrite. Chlorine was verified by comparison with other online and hand-held colorimetric methods. Samples for TDS, calcium, iron, zinc, lead, copper, aluminum, sulfate, and chloride were delivered to Del Mar Analytical Laboratory each week for analysis. Unfiltered samples were collected for total iron analyses, as particulate ferric iron is rapidly formed in waters with sufficient dissolved oxygen and disinfectant (Lytle et al 2005; McGuire 2005; Benjamin et al 1996) such as those tested in this study. The analytical methods and data reporting limits used in the pilot testing are shown in Table 4. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 40 Table 4: Analytical Methods and Data Reporting Limits Sample Analysis Analytical Method Units Method Detection Limit (MDL) Temperature SM 2550 (Thermometric) ºC N/A pH SM 4500H+-B (Electrometric) pH units N/A Conductivity SM 2510B (Conductance) μS/cm N/A Turbidity SM 2130 B (Nephlometric) NTU 0.02 Total Alkalinity Hach 8203 (Digital Titration) mg/L 10 Chlorine, Total EPA 330.5 (Hach 8167; DPD Method) mg/L 0.02 Ammonia, Total Hach 8155 (Salicylate Method) mg/L 0.02 Nitrite Hach 8507 (Diazotization Method) mg/L 0.005 Dissolved Oxygen Hach 8166 (Hydroquinone Method) mg/L 0.1 Total Dissolved Solids (TDS) EPA 160.1 mg/L 10 Calcium EPA 200.7 (ICP-AES) mg/L 0.040 Iron EPA 200.8 (ICP-MS)* μg/L 8.5 Zinc EPA 200.8 (ICP-MS) μg/L 1 Aluminum EPA 200.8 (ICP-MS) μg/L 1.5 Lead EPA 200.8 (ICP-MS) μg/L 0.040 Copper EPA 200.8 (ICP-MS) μg/L 0.25 Sulfate EPA 300.0 (IC) mg/L 4.5 Chloride EPA 300.0 (IC) mg/L 1.5 *Starting October 19, 2005, iron analyses were performed using EPA method 200.7 with a MDL of 15 micrograms per liter (μg/L.). Sample Containers, Preservation, and Holding Times Required sample sizes, sample bottle materials, preservatives (if any), and maximum holding times for each analytical method are shown in Table 5. Table 5 identifies whether samples had to be analyzed immediately or if they could be preserved and for what length of time. Samples for field analyses were collected in clean 1 L high-density polyethylene (HDPE) bottles. Field measurements were conducted the same day as collected, thereby eliminating the need for preservation of field samples. Samples were not filtered prior to preservation. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 41 Table 5: Sample Analysis Methods, Preservation, and Holding Times Sample Analysis Analytical Method Size of Sample Required Container Material Preservative Maximum Holding Time Temperature SM 2550 30 mL P, G None Analyze Immediately pH SM 4500H+-B 30 mL P, G None Analyze Immediately Conductivity SM 2510B 30 mL P, G Cool, 4°C 28 days Turbidity SM 2130 B 30 mL Hach cell None Analyze Immediately Total Alkalinity Hach 8203 100 mL P, G Cool, 4°C 28 days Chlorine, Total EPA 330.5 10 mL Hach cell None Analyze Immediately Ammonia, Total Hach 8155 10 mL Hach cell Cool, 4°C Analyze Immediately Nitrite Hach 8507 10 mL Hach cell Cool, 4°C Analyze Immediately Dissolved oxygen Hach 8166 25 mL Ampule None Analyze Immediately Total Dissolved Solids (TDS) EPA 160.1 500 mL P None 7 days Calcium EPA 200.7 125 mL‡ P HNO3 to pH<2 6 months Iron EPA 200.8 125 mL‡ P HNO3 to pH<2 6 months Zinc EPA 200.8 125 mL‡ P HNO3 to pH<2 6 months Aluminum EPA 200.8 125 mL‡ P HNO3 to pH<2 6 months Lead EPA 200.8 125 mL‡ P HNO3 to pH<2 6 months Copper EPA 200.8 125 mL‡ P HNO3 to pH<2 6 months Sulfate EPA 300.0 125 mL^ P Cool, 4°C 28 days Chloride EPA 300.0 125 mL^ P Cool, 4°C 28 days P = Plastic (HDPE), G = Glass. ‡,^Combined bottle possible for measurements if additional sample is collected. References: EPA, 1997; Eaton et al., 2001 For samples shipped to the contract laboratory, all samples were collected using clean polyethylene bottles. The contract laboratory provided pre-cleaned bottles for laboratory analysis that included the necessary preservatives. One 125 mL bottle was collected for calcium, aluminum, iron, zinc, copper, and lead; this sample was acidified with nitric acid. One 125 mL bottle for TDS and one 125 mL bottle for sulfate and chloride were also collected without preservative. Combined sample volumes for different analytes were used when possible, based on the contract laboratory’s standard practices. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 42 Chain of custody forms indicating the sample ID, sampling time, preservation, container type, and analyses requested were maintained to ensure that the samples were analyzed within the holding time. QA/QC Samples Quality assurance/ quality control (QA/QC) sampling included sample duplicates, blanks, matrix spikes, and matrix spike duplicates. Field-collected duplicate samples were used to verify consistency in sample collection and handling, in addition to providing replicates of samples for analytical accuracy checks. Sampling at one sample location, which varied each week, provided the sample duplicates. Blanks, matrix spikes, and matrix spike duplicates were performed at Del Mar Analytical Laboratory. Precision of the laboratory analyses was measured using blind field duplicate samples. Relative Percent Differences (RPD) were calculated for the duplicate samples as follows: RPD = [(Sample Result – Duplicate Result) ÷ Sample Result] * 100% Acceptance criteria for the RPD on replicate samples are generally less than 20%. For duplicate samples analyzed during the course of this experiment, more than 93% of all duplicates were within the RPD range of ±20%. Table 6 shows the RPD for each week’s duplicates. The RPD values exceeding the ±20% range were typically near the detection limits of the method (e.g., the -65% RPD for Al was the difference between 4.3 and 2.6 μg/L) or could not be explained (e.g., -58% RPD for Cu was the difference between 41 and 26 μg/L; 34% RPD for Pb was the difference between 6.4 and 4.2 μg/L; 87% RPD for Al was the difference between 12 and 1.6 μg/L). The contract laboratory was contacted about sample data points that appeared to be outliers, and retesting was performed on some of the samples. However, RPD values compared in Table 6 are for field duplicates and do not represent the laboratory QA/QC criteria that would require reanalysis (e.g., matrix spike duplicates). Comparisons of individual duplicate analyses for each analyte are shown in Figures 20 through 28. Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 43 Table 6: RPD for Duplicates Week Ca Al Fe Zn Pb Cu TDS SO42-Cl- 1 20%3% 3 34% -58% 40% 5 0% -2% 0% 19% 21% 0% -2% 0% 0% 6 0% -9% 70%-1% 8 9 0% 4% 4% 13% 2% 0% 0% 10 0% 87% 11 -3% -5% 12 -6% 6% 13 -2% 4% 0% -9% 0% -12% 0% 6% 5% 14 0% 15 0% 0% 16 0% 17 0% -11% -6% 17% 0% 0% -7% -8% 0% 18 -3% 1% 19 -2% 1% 0% 2% 0% 0% -1% 20 -5% 21 0% -5% 22 -5% 24% 23 -5% 3% 24 -5% 25 -6% 0% -3% -4% 6% 7% 26 6% -65% Figure 20: RPD for Calcium 0 10 20 30 40 50 60 70 1 3 5 7 9 11 13 15 17 19 21 23 25 WeekCa (mg/L)Sample Duplicate Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 44 Figure 21: RPD for Aluminum 0 50 100 150 200 250 300 350 400 1 3 5 7 91113151719212325 WeekAl (µg/L)Sample Duplicate Figure 22: RPD for Iron 0 50 100 150 200 250 1 3 5 7 9 1113151719212325 WeekFe (µg/L)Sample Duplicate <8.5 <8.5 Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 45 Figure 23: RPD for Zinc 0 5 10 15 20 25 30 1 3 5 7 9 1113151719212325 WeekZn (µg/L)Sample Duplicate Figure 24: RPD for Lead 0 2 4 6 8 10 12 135791113151719212325 WeekPb (µg/L)Sample Duplicate <0.04 Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 46 Figure 25: RPD for Copper 0 50 100 150 200 250 300 350 1 3 5 7 9 1113151719212325 WeekCu (µg/L)Sample Duplicate <0.25<0.25 Figure 26: RPD for TDS 0 100 200 300 400 500 600 1 3 5 7 9 11 13 15 17 19 21 23 25 WeekTDS (mg/L)Sample Duplicate Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 47 Figure 27: RPD for Sulfate 0 20 40 60 80 100 120 140 160 180 200 1 3 5 7 9 1113151719212325 WeekSO42- (mg/L)Sample Duplicate Figure 28: RPD for Chloride 0 50 100 150 200 250 1 3 5 7 9 1113151719212325 WeekCl- (mg/L)Sample Duplicate Poseidon Resources Corporation Corrosion Pilot Study 3 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 48 Visual Observations of Internal Pipe Surfaces Pipe samples were collected from the corrosion pilot plant at Carlsbad. Two 6-inch long samples were collected from the mortar-lined steel pipe and four samples 12-inches long were collected from the copper pipe lines. Samples were cut out of the pipelines by a two-man crew from the Carlsbad Water Department. Sections of pipe were selected using by measuring the entire length of the pipe, converting that number in feet to millimeters and then using a random number table selecting a random distance in millimeters from one end of each pipe. Table 7 below lists the distances selected from the ends of the pipes for sampling. Table 7: Pipe Sections Sampled for Visual Inspection Pipe Material Water Source End of pipe measured from Distance measured for sample cut, ft Pipe Sample Length, ft Mortar-lined steel MWD1 East 3.6 0.5 Mortar-lined steel RO2 West 6.2 0.5 Copper MWD1 West 22.4 1 Copper MWD2 West 28.3 1 Copper RO1 West 25.5 1 Copper RO2 West 7.6 1 All samples were marked and stored in plastic bags. The direction of flow and the tops and bottoms of the pipe samples were noted on the pipes using a marker. At the corporation yard of Carlsbad Water Department, the crew split the mortar-lined steel pipe samples lengthwise. The mortar linings were dislodged from the supporting steel pipes and were labeled and sent to the Corrosion Testing Laboratories, Inc. (Newark, DE) for visual analysis. Corrosion Testing Laboratories, Inc. split the copper pipe longitudinally and examined the internal surfaces of the pipes using a stereoscope up to 40 times magnification. Poseidon Resources Corporation Corrosion Pilot Study Water Quality Data Results 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 49 Water Quality Data Results Data Reduction Water Quality Calculations for Copper Pipes and Brass Meters Even though there was an 11-hour stagnation period, water samples were collected from the copper pipes and brass meters approximately 10 hours after the pipes were filled to ensure that the samples were obtained before the flow began. The concentration of constituents released to the water as a result of corrosion of the copper pipes and brass meters was calculated as the difference between the effluent water concentration and the source water concentration collected the same day as the effluent. The source water sample collected on the same day represented the closest representation of the water added to the copper pipes 10 hours prior to sampling. Water Quality Calculations for Distribution System Valves Effluent water samples were collected from the distribution system valve pipe loops one week after the pipes were filled. The concentration of constituents added to the water from corrosion of the distribution system valves was calculated as the difference between the effluent water sample and the source water collected the previous week. The source water sample from the previous week was collected while the distribution system valves were being filled. Water Quality Calculations for Mortar-Lined Pipes Replacement water samples of the mortar-lined pipe loops were collected during weeks 1 through 8 to test whether the water volume remaining in the pipes after the tanks were drained (i.e., dead volume) would change the influent water quality when the recirculation tank was refilled with water the next week. Replacement samples were collected from the loops after the weekly fresh water refill until week 9, when it was determined that the water remaining in the dead volume no longer significantly affected the water quality. Therefore, for week 1 through 8, the concentration of constituents added to the water from corrosion in the mortar-lined pipes was calculated as the difference between the effluent water sample and the replacement sample collected the previous week. Beginning on week 9 (8/17/05), corrosion was calculated in the same way as the distribution system valves, with the source water quality from the previous week used as the baseline for comparison with the effluent samples. Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 50 RO Permeate Conditioning The RO pilot plant operators tested alkalinity and hardness using Hach field test kits as they filled the 2,500-gallon storage tanks. The water quality goals for the RO permeate entering the tank, after passing through the calcite filter, were an alkalinity of 45-65 mg/L as CaCO3 and a hardness of 40-50 mg/L as CaCO3. If the measurements fell below these target ranges, calcite was added to the filters to increase the alkalinity and hardness. If alkalinity and hardness did not increase, the pH of the influent water was gradually decreased from 5.5 to 5.2 to dissolve more calcite. If the calcium and alkalinity were too high, the pH of the permeate entering the calcite filters was slowly increased a few tenths of a pH unit. Adding ammonium hydroxide (for chloramine formation) and sodium hydroxide (to reach the target pH of 8.5) increased the alkalinity of the RO permeate. When the conductivity and TDS of the RO permeate were high (i.e., above 800 μS/cm and 400 mg/L, respectively), the chloramines residual dropped rapidly and additional chlorine and ammonium hydroxide were needed to re-adjust the chloramines residual to achieve the target values. During these weeks, at least twice the typical volume of ammonium hydroxide was added. The volume of base required to meet the target pH of 8.5 varied considerably from week to week, as shown in Table 8. The increase in total alkalinity corresponding to the amount of ammonium hydroxide and sodium hydroxide added are depicted in Figure 29. Alkalinities higher than the target of 45-65 mg/L as CaCO3 coincided with re-chloramination events. Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 51 Table 8: Hydroxide Alkalinity Added by RO Conditioning Procedure Starting Alkalinity (mg/L CaCO3) Starting pH Qty of 30% NH4OH added (mL) Qty of 50% NaOH added (mL) Final Alkalinity (mg/L CaCO3) Change in Alkalinity after Base Addition (mg/L) Final pH 1 9.2 33 0 21 N/A 8.67 2 17 8.41 35 0 17 0 8.82 3 N/A N/A N/A N/A N/A N/A N/A 4 60 8.25 77 0 60 0 8.44 5 39 6.45 35 328 69 30 8.15 6 69 7.3 75 120 75 6 8.61 7 50 7.38 35 100 63 13 8.75 8 74 6.95 79 230 102 28 8.5 9 33 7.97 35 0 39 6 8.5 10 56 8.04 35 20 57 1 8.58 11 45 8.97 35 200 66 21 8.57 12 42 7.92 35 0 44 2 8.52 13 47 7.63 35 50 50 3 8.67 14 44 8.16 35 0 50 6 8.48 15 53 6.85 35 260 80 27 8.49 16 54 7.55 35 33 54 0 8.49 17 44 7.86 35 30 56 12 8.63 18 43 7.65 35 12 50 7 8.49 19 47 7.28 35 92 54 7 8.48 20 40 7.23 35 100 53 13 8.71 21 53 7.17 35 95 57 4 8.54 22 54 7.44 35 45 60 6 8.54 23 44 7.34 35 40 50 6 8.65 24 47 7.76 35 0 50 3 8.55 25 47 7.14 35 125 58 11 8.52 Figure 29: Alkalinity Increase During Permeate Disinfection and pH Adjustment 0 50 100 150 200 250 300 350 400 12345678910111213141516171819202122232425 WeekVolume of Hydroxide Added (mL)0 20 40 60 80 100 120 Final Alkalinity (mg/L as CaCO3)Hydroxide Added (mL) Final Alkalinity (mg/L) Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 52 Source Water Comparison: MWD vs. Conditioned RO Permeate During the corrosion pilot study, MWD and conditioned RO permeate source waters were tested for a range of water quality parameters shown in Table 3. Weekly sample results from the source waters are shown in Figures 30 through 40. MWD water pH averaged 8.2 ± 0.4. The average conditioned RO permeate pH of 8.5 ± 0.3 generally achieved the pH target for the RO permeate of 8.5 (Figure 30). Figure 30: pH of Source Waters 7.0 7.5 8.0 8.5 9.06/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05pHMWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 53 Alkalinity for the MWD water ranged from 97 to 115 mg/L as CaCO3 with an average of 105, and RO permeate alkalinity ranged from 17 to 102 with an average of 56 (Figure 31). The range of alkalinity for RO permeate once stable water was produced (i.e., from 7/27/05 forward) was 39 to 102, averaging 59 mg/L as CaCO3. The RO permeate alkalinity met the pilot study target of 45 to 65 mg/L as CaCO3, with a few weeks above the target due to conditioning issues as described previously. Figure 31: Alkalinity of Source Waters 0 20 40 60 80 100 120 1406/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Alkalinity (mg/L as CaCO3)MWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 54 Calcium concentrations in the source waters are shown in Figure 32. Over the entire test, MWD water calcium concentrations averaged 59 mg/L as Ca (ranging from 54 to 64 mg/L). Considering only the stable conditioned RO permeate from 7/27/05 forward, the calcium averaged 20 mg/L and ranged from 14 to 30 mg/L. The target range while conditioning the RO permeate was 16-20 mg/L as Ca (i.e., 40-50 mg/L as CaCO3). Figure 32: Calcium of Source Waters 0 10 20 30 40 50 60 706/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Ca (mg/L)MWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 55 TDS values for MWD water averaged 565 mg/L and ranged from 510 to 680 (Figure 33). This TDS range corresponded to measured conductivities of 796 to 937 μS/cm, yielding an average ratio of TDS-to-conductivity of 0.63. By comparison, conditioned RO permeate TDS ranged from 250 to 500 mg/L and averaged 313 (Figure 33). The RO permeate conductivity range of 501 to 1025 yielded an average TDS-to-conductivity ratio of 0.50. The difference in the TDS-to-conductivity ratio for MWD water versus conditioned RO permeate (0.63 vs. 0.50) arises from the different ions present in the water (i.e., more divalent ions, compared to monovalent ions, are rejected by the RO membranes). To illustrate this difference in the ions present, Figures 34 and 35 show the chloride and sulfate concentrations, respectively, for MWD and conditioned RO permeate. Note that a relatively high sulfate level for RO permeate on 9/28/05 may have arisen from over-addition of sulfuric acid for RO permeate pH adjustment before the calcite filter. Figure 33: TDS of Source Waters 0 200 400 600 800 10006/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05TDS (mg/L)MWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 56 Figure 34: Chloride in Source Waters 0 50 100 150 200 2506/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Cl- (mg/L)MWD Water Conditioned RO Permeate Figure 35: Sulfate in Source Waters 0 50 100 150 200 2506/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05SO42- (mg/L)MWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 57 Dissolved oxygen concentrations in both source waters were high throughout the corrosion pilot study, ranging from 7.4 to 10.3 mg/L for MWD water and 7.3 to 9.6 mg/L in conditioned RO permeate (Figure 36). Figure 36: Dissolved Oxygen in Source Waters 0 2 4 6 8 10 126/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Dissolved Oxygen (mg/L)MWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 58 Temperatures for the source waters added to the pipe loops ranged from 19 to 28ºC in MWD water and 20 to 28ºC in conditioned RO permeate for the 6-month study conducted from June to December at the shaded outdoor facility (Figure 37). Individual pipe loop temperatures were similar after 1 week of recirculation (Figure 38). Figure 37: Temperature of Source Waters Added to the Pipe Loops 10 15 20 25 306/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Temperature (degrees C)MWD Water Conditioned RO Permeate Figure 38: Temperature of Water in the Test Systems After 10 hr in the Copper Pipe or One Week in the Mortar-Lined Pipe Loops and Valve Loops 10 15 20 25 306/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Temperature (degrees C)Mortar Lined MWD1 Mortar Lined MWD2Mortar Lined RO1 Mortar Lined RO2 Valves MWD1 Valves MWD2Valves RO1 Valves RO2 Cu Pipe MWD1 Cu Pipe MWD2Cu Pipe RO1 Cu Pipe RO2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 59 Two corrosivity indices were calculated for the two source waters – LI and CCPP. The LI values for MWD water were consistently positive, ranging from 0.06 to 0.82 with an average of 0.52 (Figure 39). Conditioned RO permeate initially experienced three weeks of negative LI values until 7/27/05, after which the LI values ranged from 0.03 to 0.62, averaging 0.25 (Figure 39). The corresponding CCPP values for MWD water ranged from 0.9 to 9.7, with an average of 6.6 (Figure 40). CCPP values for conditioned RO permeate after 7/27/05 ranged from 0.0 to 6.3, averaging 1.8 (Figure 40). Figure 39: Langelier Index of Source Waters -1.00 -0.50 0.00 0.50 1.006/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Langelier IndexMWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 60 Figure 40: Calcium Carbonate Precipitation Potential (CCPP) of Source Waters -10.00 -7.50 -5.00 -2.50 0.00 2.50 5.00 7.50 10.006/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05CCPP (mg/L) MWD Water Conditioned RO Permeate Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 61 Copper Pipes and Brass Water Meters Water samples from copper pipe and brass water meters were collected an hour before the 11- hour stagnation period ended (i.e., at 10 hours). Residual chloramine concentrations in water collected from the copper pipes after this stagnation period were much lower than the source waters (Figures 41 and 42). Loss of chloramine residual was not due to nitrification, however, because concentrations of ammonia lost (Figures 43 and 44) were not coupled to nitrite produced (Figures 45 and 46). The apparent loss of ammonia is an unexplained phenomenon, but may be related to copper or lead complexation with ammonia. Whether the salicylate total ammonia test quantifies metal-complexed ammonia is not known. In contrast to the copper pipe test systems, the RO permeate mortar-lined pipe loops maintained a stable chloramine residual of at least 0.7 mg/L total Cl2 after recirculating for one week. A possible reason for the more rapid residual loss in the copper pipe test systems was disinfectant demand caused by new copper surfaces. Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 62 Figure 41: Total Chlorine Concentrations in Influent Water to Copper Pipe Test Systems 0.00 0.50 1.00 1.50 2.00 2.50 3.006/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Total Cl2 (mg/L) MWD Influent Conditioned RO Permeate Figure 42: Total Chlorine Concentrations in Water Sampled from Copper Pipe Test Systems After 10 hr Stagnation Periods 0 0.5 1 1.5 2 2.5 36/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Total Cl2 (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 63 Figure 43: Total Ammonia Concentrations in Influent Water to the Copper Pipe Test Systems 0 0.1 0.2 0.3 0.4 0.5 0.66/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Total NH3 as N (mg/L) MWD Influent Conditioned RO Permeate Figure 44: Total Ammonia Concentrations in Water Sampled from the Copper Pipes After 10 hr Stagnation Periods 0 0.1 0.2 0.3 0.4 0.5 0.66/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005 Total NH3 as N (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 64 Figure 45: Nitrite Concentrations in Influent Water to Copper Pipe Test Systems 0.00 0.10 0.20 0.30 0.40 0.50 0.606/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005 NO2 as N (mg/L) MWD Influent Conditioned RO Permeate Figure 46: Nitrite Concentrations in Water Sampled from Copper Pipes After 10 hr Stagnation Periods 0.00 0.10 0.20 0.30 0.40 0.50 0.606/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05 NO2 as N (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 65 As in the source waters, dissolved oxygen concentrations in water samples collected after the 10-hour stagnation period were above 7 mg/L in the copper pipe. pH value changes did not show discernable trends for either water source (Figure 47). Figure 47: Change in pH in Water Sampled from the Copper Pipe Test Systems After 10 hr Stagnation Periods -1.5 -1.0 -0.5 0.0 0.5 1.0 1.56/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05∆ pHMWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 66 Changes in turbidity were low following the initial start-up period (Figure 48). Initial weeks of RO permeate test systems were characterized by a decrease in turbidity values due to the settling of fine calcite initially added to the calcite filters by the RO skid operators. Once the filters were replaced with larger particle size calcite crystals (on 7/13/05) and the 2,500- gallon tank that had received the fine calcite crystals was sufficiently flushed, changes in turbidity from that point forward were low. Figure 48: Change in Turbidity in Water Sampled from the Copper Pipe Test Systems After 10 hr Stagnation Periods -15.0 -10.0 -5.0 0.0 5.0 10.0 15.06/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005∆ Turbidity (NTU)MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 67 Distribution System Valves Recirculating distribution system gate valve loops were tested to compare metal leaching from the valves for the two source waters. The cast iron gate valves harvested from the Carlsbad distribution system were significantly corroded at the onset. Figure 49 shows the interior corrosion of the gate valves. Residual chloramine concentrations dropped within the recirculation week periods to below 0.2 mg/L in all four loops (Figures 50 and 51). Significant nitrification was observed, particularly in the warmer summer months, shown as ammonia loss (Figures 52 and 53) and nitrite production (Figures 54 and 55). High nitrite and low total chlorine concentrations were observed in week 3 (7/6/05) because the water from the previous week was not replaced due to an operational problem with the RO pilot system. Water was recirculated in the loops for 2 weeks before replacement on 7/13/05. The decrease in ammonia concentrations during the pilot testing occurred by an as yet unidentified mechanism, because nitrification during that period could not account for the ammonia loss. This decrease in measured ammonia may have been due to metal complexation of ammonia or reactions of ammonia with the valve surfaces. It is likely that the rapid nitrification observed in the RO permeate loops occurred due to existing ammonia oxidizing bacteria (AOB) biofilms in the corroded surfaces in the iron valves. The source of the AOB was biofilms formed in the chloraminated Carlsbad distribution system. By comparison, nitrification was not observed in the new RO permeate mortar-lined pipe loops, indicating that the nitrification in the distribution valve loops was not caused by AOB in the source water. Figure 49: Interior of the Distribution System Valves Showing Iron Corrosion Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 68 Figure 50: Total Chlorine Concentrations in Influent Water to the Distribution System Valve Loops 0.00 0.50 1.00 1.50 2.00 2.50 3.006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Total Cl2 (mg/L) MWD Influent Conditioned RO Permeate Figure 51: Total Chlorine Concentrations in Water Sampled from the Distribution System Valve Loops After One-Week Recirculation Periods 0.00 0.50 1.00 1.50 2.00 2.50 3.006/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Total Cl2 (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 69 Figure 52: Total Ammonia Concentrations in Influent Water to the Distribution System Valve Loops 0 0.1 0.2 0.3 0.4 0.5 0.66/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005 Total NH3 as N (mg/L) MWD Influent Conditioned RO Permeate Figure 53: Total Ammonia Concentrations in Water Sampled from the Distribution System Valve Loops After One-Week Recirculation Periods 0 0.1 0.2 0.3 0.4 0.5 0.66/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005 Total NH3 as N (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 70 Figure 54: Nitrite Concentrations in Influent Water to the Distribution System Valve Loops 0.00 0.10 0.20 0.30 0.40 0.506/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005NO2 as N (mg/L) MWD Influent Conditioned RO Permeate Figure 55: Nitrite Concentrations in Water Sampled from the Distribution System Valve Loops After One-Week Recirculation Periods 0 0.1 0.2 0.3 0.4 0.56/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05 NO2 as N (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 71 As in the other loops, dissolved oxygen concentrations in the distribution valve loops were above 7 mg/L throughout the study. pH values observed in water samples from the distribution valve loops increased slightly for the MWD water source and decreased slightly for the conditioned RO permeate water source (Figure 56). The pH drop in the RO valve loops could be explained by nitrification; however, the pH increase in the MWD loops could not be explained. Figure 56: Change in pH in Water Sampled from Distribution System Valve Loops After One-Week Recirculation Periods -2 -1.5 -1 -0.5 0 0.5 1 1.5 26/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Δ pHMWD 1 MWD 2 RO 1 RO 2 Water Sources Switched Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 72 Turbidity values were fairly constant in water samples from three of the four distribution valve loops. One of the RO permeate loops (RO1) consistently showed elevated turbidities compared to the others, although the magnitude of the turbidity in RO1 decreased by nearly an order of magnitude by 7/27/05 (Figure 57). On 11/16/05, source waters feeding the RO1 and MWD1 loops were switched. The higher turbidity values were due to the non-representative release of iron from one of the gate valves. Figure 57: Change in Turbidity in Water Sampled from Distribution System Valve Loops After One-Week Recirculation Periods -100 -75 -50 -25 0 25 50 75 1006/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05∆ Turbidity (NTU)MWD 1 MWD 2 RO 1 RO 2 Water Sources Switched Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 73 Mortar-Lined Pipes Water samples from the distribution system cement mortar-lined steel pipe loops were tested to assess calcium, aluminum, and metal leaching for the two source waters. Samples were collected from the mortar-lined pipe loops after one week of recirculation. More significant chloramine decay was observed for the loops containing MWD water, compared with those containing RO permeate. Residual chloramine concentrations in the RO permeate after one- week recirculation periods averaged 1.09 mg/L, compared to an average residual chloramines concentration of 0.26 mg/L in the MWD water (Figures 58 and 59). Low total chlorine, low total ammonia, and high nitrite concentrations were observed in week 3 (7/6/05) because the water from the previous week was not replaced due to an operational problem with the RO pilot system. Water was recirculated in the loops for 2 weeks before replacement on 7/13/05. Nitrification was not observed in the RO permeate-fed mortar-lined pipe loops. In contrast, nitrification occurred rapidly in the mortar-lined pipes fed MWD water, as characterized by a decrease in ammonia (Figures 60 and 61) and increase in nitrite (Figures 62 and 63). Since the mortar-lined pipes were new, no resident AOB biofilms existed on the pipes, indicating that nitrification in the MWD mortar-lined pipes was likely caused by AOB in the MWD water. Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 74 Figure 58: Total Chlorine Concentrations in Influent Water to Mortar-Lined Pipe Loops 0.00 0.50 1.00 1.50 2.00 2.50 3.006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Total Cl2 (mg/L) MWD Influent Conditioned RO Permeate Figure 59: Total Chlorine Concentrations in Water Sampled from the Mortar- Lined Pipe Loops After One-Week Recirculation Periods 0.00 0.50 1.00 1.50 2.00 2.50 3.006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Total Cl2 (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 75 Figure 60: Total Ammonia Concentrations in Influent Water to Mortar-Lined Pipe Loops 0.00 0.10 0.20 0.30 0.40 0.50 0.606/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Total NH3 as N (mg/L) MWD Influent Conditioned RO Permeate Figure 61: Total Ammonia Concentrations in Water Sampled from the Mortar- Lined Pipe Loops After One-Week Recirculation Periods 0.00 0.10 0.20 0.30 0.40 0.50 0.606/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Total NH3 as N (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 76 Figure 62: Nitrite Concentrations in Influent Water to Mortar-Lined Pipe Loops 0.00 0.10 0.20 0.30 0.40 0.50 0.606/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005 NO2 as N (mg/L) MWD Influent Conditioned RO Permeate Figure 63: Nitrite Concentrations in Water Sampled from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods 0.00 0.10 0.20 0.30 0.40 0.50 0.606/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005 NO2 as N (mg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 77 Dissolved oxygen levels in the mortar-lined steel pipe loop samples were above 6.9 mg/L throughout the study. As a result of poor calcium carbonate precipitation control, cement mortar leaching occurred during the initial month and pH increased initially in RO permeate loop water samples (i.e. became more alkaline). After the first month of operation, pH in the MWD-fed loops increased an average of 0.30 during the one-week circulation periods; compared to an average of -0.04 in the RO permeate (Figure 64). The cause for the increase in MWD loop pH was a mystery, as no change or a decrease might have been expected due to nitrification. Figure 64: Change in pH in Water Sampled from Mortar-Lined Pipe Loops After One-Week Recirculation Periods -3 -2 -1 0 1 2 36/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Δ pH MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 78 Initial calcium leaching was observed for the RO permeate due to cement mortar leaching, but no discernable leaching trends were noted after 7/27/05 for either source water (Figure 65). Figure 65: Change in Calcium in Water Sampled from Mortar-Lined Pipe Loops After One-Week Recirculation -30 -20 -10 0 10 20 306/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Δ Ca (mg/L) MWD 1 MWD 2RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 4 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 79 An initial decrease in turbidity was noted for the RO permeate mortar-lined pipe loop samples. As discussed previously, initial weeks of RO permeate conditioning were affected by the introduction of fine calcite crystals from the calcite filter into the RO permeate. The RO permeate turbidity stabilized after the calcite filters were replaced with larger particle size crystals on 7/13/05 and the 2,500-gallon tank that had received the fine calcite crystals was sufficiently flushed. Little change was observed in turbidity in any of the four test loops after 7/27/05, when stable RO permeate was produced (Figure 66). Figure 66: Change in Turbidity of Water Sampled from Mortar-Lined Pipe Loops After One-Week Recirculation Periods -15 -10 -5 0 5 10 156/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Δ Turbidity (NTU)MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study Discussion of Corrosion Results 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 80 Discussion of Corrosion Results Comparison to Water Quality Standards The water quality results for this corrosion project were compared to California DHS drinking water standards for aluminum, copper, zinc, iron, and lead. For compliance, primary and secondary Maximum Contaminant Levels (MCLs) are measured at the water source following treatment. Primary MCLs are not to be exceeded in the water supplied to the public due to health concerns. Secondary MCLs are not to be exceeded in the water supplied to the public because these constituents may adversely affect the taste, odor, or appearance of the drinking water. The primary MCL for aluminum is 1,000 µg/L; secondary MCLs include 1,000 µg/L for copper, 200 µg/L for aluminum, 5,000 µg/L for zinc, and 300 µg/L for iron. The Action Levels for copper and lead are 1,300 μg/L and 15 µg/L, respectively. For compliance purposes, the copper and lead action levels are exceeded if more than 10% of the water samples collected at designated customer taps in a distribution system have a higher concentration than the Action Level. Extrapolation of the corrosion pilot results to predict compliance with the LCR is not possible. The comparison of pilot results with the LCR Action Levels was performed to place the results in the context of allowable limits. The test conditions for measurement of lead and copper concentrations in the pilot unit produced lead and copper concentrations that are higher than would be expected under home sampling conditions. Copper Pipes and Brass Water Meters In samples from the brass meters, copper concentrations were lower than the Secondary MCL and Action Level throughout the course of the study (Figure 67). Lead concentrations in samples from the brass meters were lower than the Action Level (Figure 68). Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 81 Figure 67: Copper Leached from Brass Meters After 10 hr Stagnation Periods Compared to California Drinking Water Standards 0 200 400 600 800 1000 1200 14006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Cu (µg/L)MWD1MWD2RO1RO2 Copper Action Level Copper Secondary MCL Figure 68: Lead Leached from Brass Meters After 10 hr Stagnation Periods Compared to California Drinking Water Standards 0 10 20 30 40 506/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Pb (µg/L)MWD1 MWD2 RO1 RO2 Lead Action Level Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 82 In samples from the copper pipe, copper concentrations were below the Secondary MCL and Action Level throughout the course of the study for both source waters, with the exception of two MWD data points (on 11/9/05) that were outliers (Figure 69). Copper levels, however, increased during the six-month test period, with more significant leaching observed for the MWD supply. Since testing used new copper pipe, the results in this study may be elevated compared to the copper leaching that would have been observed in copper pipe harvested from the distribution system. For both source waters, additional time may have been needed for the new copper pipe to be protectively coated by CaCO3 and other scale-forming compounds. Water quality parameters that impact copper corrosion include bicarbonate, pH, redox, chloride, sulfate, and natural organic matter. For new copper pipe, dissolved inorganic carbon (i.e., bicarbonate at the pH values for these source waters, as measured by alkalinity) in particular can impact copper leaching (Schock 1999). The higher degree of copper leaching in MWD-fed copper loops might be explained by higher bicarbonate concentrations compared to conditioned RO permeate. Edwards et al (1996) found that copper release from pipe was a linear function of alkalinity. In that study (Edwards et al 1996), differences in alkalinity of 50 vs. 100 mg/L as CaCO3 resulted in copper concentrations on par with what was observed in this corrosion pilot study for source waters with approximately the same alkalinity difference (Figure 69). Aqueous copper is primarily complexed by carbonate species at pH values of concern in this study (Schock and Lytle 1995). The difference between MWD and RO permeate copper leaching may also be, in part, due to higher sulfate concentrations and lower pH in the MWD water. Both source waters contained high dissolved oxygen concentrations that were conducive to copper leaching. Formation of cupric ions from the metallic copper surfaces of the new pipe was enabled by the dissolved oxygen present (Schock 1999). The presence of excess ammonia in the chloraminated source waters may also have contributed to the leaching of copper, as ammonia can complex copper and increase aqueous copper concentrations (Schock 1999; AwwaRF 1990; Butler and Ison, 1966). Additional research is needed to fully understand the role of free ammonia in copper release from copper pipe. Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 83 Figure 69: Copper Leached from Copper Pipes After 10 hr Stagnation Periods Compared to California Drinking Water Standards 0 200 400 600 800 1000 1200 14006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Cu (µg/L)MWD 1MWD 2RO 1RO 2 Copper Action Level Copper Secondary MCL Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 84 Lead concentrations in samples from the copper pipes were lower than the Action Level after the RO permeate was adequately conditioned to obtain a positive Langelier Index (Figure 70) and lead leaching stabilized (i.e., after 8/17/05). Conditioning of the lead/tin soldered joints appeared to occur, potentially due to protective scale buildup on the solder or decreased leaching once the more soluble components of the solder were removed. Lead carbonate and lead carbonate hydroxide precipitates are typically responsible for lead passivation of surfaces (Schock 1999). Lead leaching from lead-tin solder is strongly influenced by pH (Schock 1999). According to equilibrium calculations, pH adjustment to the range of 8.0 to 8.5 can reduce lead solubility (as Pb2+) in water (Trussell 1985). Both source waters were typically in this range (Figure 30). Dissolved oxygen in the source waters could have oxidized the metallic lead surfaces to yield oxidized, aqueous lead (Schock et al 1996). As with copper, the oxidized aqueous lead ions may complex with free ammonia; however, the relative importance of this complexation has not been demonstrated (Schock 1999; Trussel 1985). Overall, lead leached into the water from the newly soldered joints of the copper pipe test systems was low once passivation occurred. Figure 70: Lead Leached from Copper Pipes After 10 hr Stagnation Periods Compared to California Drinking Water Standards 0 10 20 30 40 506/22/057/6/057/20/058/3/058/17/058/31/059/14/059/28/0510/12/0510/26/0511/9/0511/23/0512/7/0512/21/05Pb (µg/L)MWD 1 MWD 2 RO 1 RO 2 Lead Action Level Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 85 Distribution System Valves Zinc leached into the water from the distribution system valves were far below the California Secondary MCL for zinc throughout the course of the study (Figure 71). Figure 72 shows the change in zinc concentrations between the influent source water and effluent after one week of recirculation. One MWD source water sample result may have skewed the data on 10/12/05 for both MWD loops, since the MWD source water typically ranged from 3 to 21 μg/L but the source water zinc level was 72 μg/L for that set of samples. Overall, little zinc was leached from the valve loops during the corrosion study. Figure 71: Zinc Leached from Valve Loops After One-Week Recirculation Periods, Compared to CA Drinking Water Standards 0 1,000 2,000 3,000 4,000 5,0006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Zn (µg/L)MWD1 MWD2 RO1 RO2 Zinc Secondary MCL Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 86 Figure 72: Change in Zinc Concentrations in Water from the Valve Loops After One-Week Recirculation Periods -100 -75 -50 -25 0 25 50 75 1006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Δ Zn (µg/L) MWD 1 MWD 2 RO 1 RO 2 Water Sources Switched Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 87 Iron leached from one of the distribution valves supplied RO permeate were above the secondary MCL (Figure 73). During the last month of the study, the water supply to this valve was switched to the MWD water instead of RO permeate, at which time recirculated MWD water demonstrated high iron concentrations as well. RO permeate fed into the former MWD1 loop yielded the low iron results observed for MWD water. These results indicated that the RO1 valves were the source of the high iron levels observed in the RO1 loop and that the RO1 loop iron release did not occur as a result of aggressive source water. Overall, the other three valve loops showed iron releases below or near the secondary MCL. The week-long recirculation of water in the valve loops presents a “worst-case” scenario for metal leaching and buildup in the water far more severe than would occur in a distribution system. Figure 73: Iron Leached from the Valve Loops After One-Week Recirculation Periods, Compared to California Drinking Water Standards 0 2,000 4,000 6,000 8,000 10,0006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Fe (µg/L)MWD1 MWD2 RO1 RO2 Iron Secondary MCL Water Sources Switched Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 88 Figure 74: Iron Leached from the Valve Loops After One-Week Recirculation Periods, Compared to California Drinking Water Standards – Enlarged View of Iron up to 2,000 μg/L 0 500 1,000 1,500 2,0006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Fe (µg/L)MWD1 MWD2 RO1 RO2 Iron Secondary MCL Water Sources Switched Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 89 Lead released from the distribution system valves were below the Action Level, except for one RO loop sample on 7/20/05 (Figure 75). Distribution valve loop RO1 leached much higher concentrations of lead compared to the other three loops, which may be related to the significant iron leaching observed for RO1 as well. The source of lead in the cast iron distribution valves, however, is not known. Figure 75: Lead Leached from the Valve Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards 0 10 20 30 40 506/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Pb (µg/L)MWD1 MWD2 RO1 RO2 Lead Action Level Water Sources Switched Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 90 Mortar-Lined Pipes Zinc and iron were lower than the Secondary MCLs and lead was below the Action Level in samples from the mortar-lined pipe loops throughout the course of the pilot study (Figures 76, 78, and 79). Since the mortar-lined pipe was new, any metal leaching would be expected to originate from exposed flanges areas. No significant zinc leaching occurred, as shown from the change in zinc concentrations depicted in Figure 77. The two MWD outlying data points in Figure 77 arose from what was likely a laboratory error in the MWD source water sample, as discussed previously. Figure 76: Zinc Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards 0 1,000 2,000 3,000 4,000 5,0006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Zn (µg/L)MWD1 MWD2 RO1 RO2 Zinc Secondary MCL Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 91 Figure 77: Change in Zinc Concentrations in Water from Mortar-Lined Pipe Loops After One-Week Recirculation Periods -80 -60 -40 -20 0 20 40 60 806/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Δ Zn (μg/L) MWD 1 MWD 2 RO 1 RO 2 Figure 78: Iron Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards 0 50 100 150 200 250 3006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Fe (µg/L)MWD1 MWD2 RO1 RO2 Iron Secondary MCL Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 92 Figure 79: Lead Leached from the Mortar-Lined Pipe Loops After One-Week Recirculation Periods Compared to California Drinking Water Standards 0 5 10 156/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Pb (µg/L)MWD1 MWD2 RO1 RO2 Lead Action Level Aluminum concentrations in samples from the mortar-lined pipe loops supplied with RO permeate dropped below the Secondary MCL after the RO Permeate conditioning was optimized and a positive LI was consistently produced (i.e., after 7/27/05; Figure 80). Initially significant aluminum concentrations released into the RO permeate (Figure 81) coupled with a strong increase in pH (Figure 64) indicated that the aluminum in the cement mortar lining was leached in the RO permeate loops, most likely due to aggressive water produced in the first few weeks of operation. The initial decrease in aluminum concentrations for the MWD loops (Figure 81) is unexpected and cannot be explained at this time. Cement mortar lining is comprised of a range of calcium silicates and calcium aluminates, ranging from 5 to 36% aluminum (depicted as Al2O3) depending on the type of cement used (AWWA 2002; Trussell 1985). Studies have shown that waters with low alkalinities and negative LIs (i.e., aggressive waters) can leach significant quantities of calcium hydroxide and aluminum from new cement mortar-lined pipe (Douglas et al. 1996; Berend and Trouwborst 1999; AWWA 2002). For new cement mortar-lined pipe, a protective coating is formed when a non-aggressive water containing adequate alkalinity enables the formation of calcium carbonate deposits in the pore structure and on the surface of the cement mortar, thereby limiting the further dissolution of the calcium hydroxide component of the cement mortar (Douglas et al. 1996). As demonstrated in this corrosion pilot study, RO permeate source water with a negative LI and low alkalinity produced during the startup phase yielded significant aluminum leaching and pH increases. However, once a positive LI and CCPP were achieved after 7/27/05, aluminum leaching in the mortar-line pipe loops was low. These findings highlight the importance of conditioning RO permeate to prevent mortar-lined pipe Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 93 corrosion, as well as the non-corrosiveness of conditioned RO permeate when target water quality goals are achieved. Figure 80: Aluminum Leached from the Mortar-Lined Pipe Loops After One- Week Recirculation Periods, Compared to California Drinking Water Standards 0 200 400 600 800 10006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Al (µg/L)MWD1 MWD2 RO1 RO2 Aluminum Secondary MCL Aluminum Primary MCL Figure 81: Change in Aluminum Concentrations in Water from Mortar-Lined Pipe Loops After One-Week Recirculation Periods -500 -400 -300 -200 -100 0 100 200 300 400 5006/22/20057/6/20057/20/20058/3/20058/17/20058/31/20059/14/20059/28/200510/12/200510/26/200511/9/200511/23/200512/7/200512/21/2005Δ Al (μg/L) MWD 1 MWD 2 RO 1 RO 2 Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 94 Statistical Analyses The statistical significance of differences between MWD- and RO permeate-fed loops were tested using two-way analysis of variance (ANOVA) with replication. The experimental design incorporating replicate pipe loops for each water supply enabled a much more rigorous statistical testing than if only single loops had been used. For the two-way ANOVA, the two variables were water source and time. The two-way ANOVA with and without replication were conducted using the Microsoft® Office Professional Excel 2003 (SP1) Analysis ToolPak. Table 9 summarizes the results of the tests of statistical significance evaluating the null hypothesis that there is no difference in metal leaching between the source water test systems (MWD and RO). Unless otherwise noted, the data included in the statistical analyses were differences in metal concentrations between source water and water samples collected after planned contact with pipe surfaces. These data included samples collected between 7/27/05 and 12/14/05, i.e., once stable RO permeate with a positive LI and CCPP was produced. Table 9: Statistical Significance of Difference as Determined by Two-Way ANOVA with Replication (α = 0.05) and Average Analytical Results Cu Pb Fe Zn Ca Al Copper Pipe Yes MWD avg. 565 RO avg. 258 Yes≠ MWD avg. 2.0 RO avg. 2.6 Brass Meters Yes MWD avg.109 RO avg. 68 Yes≠ MWD avg. 6.6 RO avg. 9.2 Mortar-Lined Pipe No MWD avg. -0.031 RO avg. 0.032 No MWD avg. -1.7 RO avg. 10 No‡ MWD avg. 1.5 RO avg. 2.2 No MWD avg. 0.9 RO avg. 1.2 No MWD avg. 7.3 RO avg. 9.2 Distribution Valves* No MWD avg. 0 RO avg. 1 No MWD avg. 159 RO avg. 199 Yes‡ MWD avg. 5.1 RO avg. 9.3 * Two-way ANOVA without replication due to high RO1 valve results ‡ Outlier on 10/12/05 removed from analysis ≠ Data set starting on 8/31/05 rather than 7/27/05 For the copper pipe test systems, a statistically significant difference existed between the copper leached from the MWD systems and the RO systems. The MWD-fed pipes, however, leached more copper into the water than the RO permeate-fed pipes. Average copper concentrations in water from both the MWD and RO permeate copper test systems were below that of the Action Level, as shown in Figure 69. A similar pattern, although with lower magnitude of copper concentrations, was observed for samples from the brass meters. A different behavior was observed for the lead concentrations in both the copper pipe samples and brass meter samples. For lead, particularly in the copper pipe, the concentration differences between influent and effluent water continued to decrease to an equilibrium level reached around 8/31/05 (Figure 70). Although the RO permeate had a consistently positive LI Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 95 beginning on 7/27/05, the lead from freshly soldered joints required a few weeks beyond this point to reach the equilibrium leaching level observed during this study. Consequently, the lead data considered in the statistical analyses for the copper pipe and brass meters included data from 8/31/05 to 12/14/05. ANOVA results showed a statistically significant difference between samples from the MWD- and RO-fed copper pipe systems. However, the average lead concentrations leached from the soldered joints on the copper pipes were low for both sources (2.0 μg/L for MWD and 2.6 μg/L for RO permeate) and the levels did not have any regulatory or engineering significance. By comparison, the lead released into water in contact with the brass meters was three to four times greater than for the copper pipe. The average lead concentrations leached from the MWD-fed and RO permeate-fed brass meters were 6.6 and 9.2 μg/L, respectively, and the difference was statistically significant. The data plot as a function of time, however, shows that the levels were consistently below the lead action level after 8/31/05 (Figure 68). The aggressive testing using copper pipes with fresh solder and samples collected from water directly in contact with brass meters produced lead levels below the lead action level of 15 μg/L. For the mortar-lined pipe, no statistically significant differences were observed for lead, iron, zinc, calcium, and aluminum between 7/27/05 and 12/14/05. Averages shown in Table 9 and trends in the plots as a function of time (Figures 76, 78, 79, and 80) highlight that little difference in leached metal concentrations was seen for the two source waters. Statistical analysis of the distribution system valve data showed no difference for lead leaching into water from the valves. Very low lead averages were also observed. Zinc, on the other hand, showed a statistically significant difference between the two source waters. Table 9 notes that data from 10/12/05 was excluded due to an outlier that appeared to be an analytical problem at the laboratory. If that anomalous data point (shown on Figure 77) was included, the statistical analyses showed no significant difference due to the resulting high variability in the data. Excluding the anomalous data on 10/12/05, the RO permeate and MWD average zinc concentration leached were 9.3 μg/L and 5.1 μg/L, respectively. However, the importance of this statistical difference is minor when comparing the results to the zinc secondary standard of 5,000 μg/L. For the distribution system valves loops, water samples from one valve loop consistently had higher iron concentrations than the other three valves. During the last month when the source waters were switched for this loop, it was confirmed that the valve’s release of iron into the water was independent of whether the water source was MWD or RO permeate. If these anomalously high results were included in the statistical analysis, falsely-based significance would be observed between iron leaching from MWD and RO permeate loops. Consequently, a two-way ANOVA without replication was run for iron in the distribution valve loops RO2 and MWD2. Results of this analysis showed no statistical significance between the MWD- and RO permeate-fed loops for iron. Average iron concentrations leached in both loops were below the secondary MCL for iron. Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 96 Visual Observation of Internal Pipe Surfaces Corrosion Testing Laboratories, Inc. concluded that no pitting or corrosion loss of the copper pipe from either MWD or RO water was observed. Green scale formation was seen in all 4 pipe samples (2 MWD and 2 RO). RO permeate appeared to show more significant green scale with an unidentified black compound underlying the green scale. Note that copper leaching into the water was more pronounced for the MWD water; thus, the scale in the pipes receiving RO permeate may have been protective. Figure 82: Interior Pipe Surfaces of the MWD1 Copper Pipe – Bottom Photo Shows 20X Magnification Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 97 Figure 83: Interior Pipe Surfaces of the MWD2 Copper Pipe – Bottom Photo Shows 20X Magnification Figure 84: Interior Pipe Surfaces of the RO1 Copper Pipe – Bottom Photos Shows 20X Magnification of Top Surface (left) and 7X Magnification of Bottom Surface After Cleaning (right) Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 98 Figure 85: Interior Pipe Surfaces of the RO2 Copper Pipe – Bottom Photos Shows 20X Magnification of Top Surface (left) and 7X Magnification of Top Surface After Cleaning (right) Mortar-lined pipe samples from testing loops MWD1 and RO2 were visually inspected. Figure 86 shows that the MWD1 pipe experienced cracking, which was likely caused by the aggressive nature of cutting a subsection of pipe from the pipe loop. The RO2 pipe loop (Figure 87) showed evidence of depressions in the pipe surface, which may have occurred prior to achieving the targeted RO permeate conditioning values of hardness, alkalinity, and Langelier Index. Water quality testing showed evidence of aluminum leaching prior to 7/27/05 (Figures 80-81). Subsequently, little aluminum leaching was observed once the RO permeate was properly conditioned. Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 99 Figure 86: Interior Pipe Surfaces of the MWD1 Mortar-Lined Pipe – Top Photo Shows Top Pipe Surface; Lower Photo Shows 20X Magnification Poseidon Resources Corporation Corrosion Pilot Study 5 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 100 Figure 87: Interior Pipe Surfaces of the RO2 Mortar-Lined Pipe – Top Photo Shows Bottom Pipe Surface; Lower Photos Show 7X (left) and 20X (right) Magnification Poseidon Resources Corporation Corrosion Pilot Study Summary and Conclusions 6 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 101 Summary and Conclusions Corrosion pilot testing of conditioned RO permeate and MWD source waters in consumer plumbing and distribution system materials yielded the following conclusions: • Conditioned RO permeate tested in this pilot study and found to yield a positive LI and CCPP was characterized by the following water quality parameters: pH of 8.5, alkalinity of 45 to 65 mg/L as CaCO3, TDS of 300 to 350 mg/L, calcium hardness of 40 to 50 mg/L as CaCO3. • Copper and lead release into water was tested from new copper pipe joined with lead/tin solder and harvested brass water meters. Samples were collected after 10 hours of an 11 hour stagnation period, which was preceded and followed by an hour of flowing water to simulate diurnal household water use. • Copper leaching increased over time for both water sources during the study. More leaching occurred in test systems receiving MWD source water, potentially due to higher bicarbonate levels that could complex copper. However, the copper concentrations were lower than the Lead and Copper Rule Action Level. • Lead release into the water from the copper pipe solder decreased as a function of time due to passivation and stabilized after approximately 11 weeks. Lead levels were well below the 15 μg/L Action Level for both source waters and the difference in lead levels, while statistically significant, did not have any regulatory or engineering significance. • Lead leached into the water from the brass meters was higher than that released from the copper pipe solder although the brass meters were approximately 15 to 20 years old. Higher lead concentrations were observed for the RO permeate (average of 9.2 μg/L) compared to MWD water (average of 6.6 μg/L). While the differences were statistically significant, the aggressive testing and sampling regime characterized by collection of samples directly from water in contact with the brass meters and freshly soldered joints did not result in lead levels exceeding the lead Action Level of 15 μg/L. • Mortar-lined pipe corrosion was tested using new cement mortar-lined pipe in a recirculating pipe loop in which source water was replaced once per week. • Initial RO permeate conditioning during startup resulted in a negative LI and CCPP, which resulted in aluminum leaching from the mortar-lined pipe. However, once stable RO permeate was produced, no significant leaching or statistically significant Poseidon Resources Corporation Corrosion Pilot Study 6 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 102 • differences in concentration changes for calcium, aluminum, iron, zinc, or lead were observed for the two source waters. • Metals release from harvested cast iron distribution system gate valves was tested using recirculating pipe loops with source water replaced once per week. • Little lead and zinc were leached into the water by either water source in the distribution valve loops. Excluding one valve that released much higher iron concentrations than the other three valves regardless of the source of water, average iron concentrations leached in both loops were below the secondary MCL for iron. No significant difference was observed in lead, zinc, or iron leaching between the two source waters. • In summary, conditioned RO permeate and MWD source waters in contact with typical household plumbing and distribution system materials did not exceed primary MCL standards, secondary MCL standards, and LCR Action Levels after the initial conditioning period, despite the testing of “worst-case” conditions such as recirculation of water in the mortar-lined pipe and distribution valve loops and sample collection directly from water in contact with brass meters in the copper pipe test systems. • Since corrosion differs depending on the type of material, water that may be passivating for one material may be corrosive of another. For example, this pilot study showed that MWD water, while protective of iron valves and mortar-lined pipe, may be more corrosive toward new copper pipe than conditioned RO permeate. The data indicates that the target conditioned water quality parameters in this study, including alkalinity, pH, calcium hardness, and TDS, would provide a water quality that is protective of all of the materials tested. • Corrosion pilot testing of both consumer plumbing and distribution system materials provides important pieces of evidence that the proposed introduction of conditioned desalinated seawater supply is not likely to trigger new corrosion problems in the Carlsbad distribution system. Poseidon Resources Corporation Corrosion Pilot Study References 7 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 103 References American Water Works Association (AWWA) with assistance from Economic and Engineering Services, Inc. 2002. Permeation and Leaching. Distribution System White Paper produced for the USEPA. http://www.epa.gov/safewater/tcr/tcr.html. Benjamin, M.M., Sontheimer, H., and Leroy, P. 1996. Corrosion of Iron and Steel. In: Internal Corrosion of Water Distribution Systems. AwwaRF: Denver, CO. Berend, K and Trouwborst, T. 1999. Cement-mortar pipes as a source of aluminum. Jour. AWWA, 91 (7), 91-100. Chao, P-F., Frey, M., Arias, M., Loveland, J., and Summers, R.S. 2004. The stability of chlorinated-derived residuals in desalinated seawater. Presented at the AWWA Annual Conference and Exposition, Orlando, FL. Douglas, B.D., Merrill, D.T., and Catlin, J.O. 1996. Water quality deterioration from corrosion of cement-mortar linings. Jour. AWWA, 88 (7), 99-107. Edwards, M., Schock, M.R., and Meyer, T.E. 1996. Alkalinity, pH, and copper corrosion by- product release. Jour. AWWA, 88 (3), 81-94. Loveland, J. and Means, E. 2004. Disinfection by-product formation in a simulated distribution system: Blending desalinated seawater from the Poseidon Resources Corporation pilot facility with local drinking water sources. Report prepared for Poseidon Resources Corporation. Loveland, J. and Means, E. 2005. Implications of Bromamine Chemistry for Drinking Water Treatment. Presented at the Water Quality and Technology Conference, Quebec City, Quebec. Lytle, D.A., Sarin, P., and Snoeyink, V.L. 2005. The effect of chloride and orthophosphate on the release of iron from a cast iron pipe section. Journal of Water Supply: Research and Technology – AQUA. 54 (5), 267-281. McGuire, M.J. 2005. Personal Communications. Schock, M.R. 1999. Internal Corrosion and Deposition Control. In: Water Quality and Treatment: A Handbook of Community Water Supplies. McGraw Hill, Inc.: New York. Schock, M.R., Lytle, D. A., and Clement, J.A. 1995. Effect of pH, DIC, orthophosphate, and sulfate on drinking water cuprosolvency. EPA/600/R-95/085. Poseidon Resources Corporation Corrosion Pilot Study 7 5372001 / STM INDEPENDENT ENVIRONMENTAL ENGINEERS, SCIENTISTS AND CONSULTANTS Corrosion Pilot Study 104 Schock, M.R., Wagner, I., and Oliphant, R.J. 1996. Corrosion and Solubility of Lead in Drinking Water. In: Internal Corrosion of Water Distribution Systems. AwwaRF: Denver, CO. Trussell, R.R. 1985. Corrosion. In: Water Treatment Principles and Design. Ed. J.M.M. Montgomery. John Wiley and Sons: New York.