WRF 4509 Webcast Metals Accumulation and Release within the Distribution System: Evaluation of Mechanisms and Mitigation February 4, 2016

Transcription

1 No part of this presentation may be copied, reproduced, or otherwise utilized without permission. WRF 4509 Webcast Metals Accumulation and Release within the Distribution System: Evaluation of Mechanisms and Mitigation February 4, 2016

2 Useful Links Project Establishing Site-Specific Flushing Velocities Project Investigation of Pipe Cleaning Methods Project Assessment of Inorganics Accumulation in Drinking Water System Scales and Sediments Webcasts on Demand Water Research Foundation. ALL RIGHTS RESERVED.

3 Introduction Melinda Friedman, P.E. President Confluence Engineering Group, LLC

4 Presentation Overview Project Background Overview of PCMC Water System and Research Need Results of Monitoring and Temporal Analysis Main Cleaning Effectiveness and Costs Industry Recommendations Conclusions

5 Project Team Confluence Engineering Group, LLC: Melinda Friedman, Andrew Hill, Stephen Booth, Michael Hallett Utah State University: Laurie McNeill, Joan McLean, David Stevens, Darwin Sorensen, Tiana Hammer, William Kent Park City Municipal Corporation: Michelle De Haan, Kyle MacArthur, Ken Mitchell WRF Project Manager: Jian Zhang PAC Members: Anne Camper, John Consolvo, Dan Giammar, Brian Lakin Utah Water Research Lab: provided additional funding

6 Trace Inorganic Contaminant (TIC) Accumulation and Release with Distribution Systems Non-conservative water quality behavior with distribution systems: in out Source: Hill and Friedman et al., 2010 Journal AWWA

7 Progression of Accumulation and Release Leading to Possible Customer Exposure Example TICs of Concern Arsenic Lead Cadmium Thallium Copper Aluminum Chromium Antimony Plus others

8 Overview of Park City Tailored Collaboration Proposal (WRF 4509) Objectives Develop guidance for system-specific evaluation of vulnerability to metals accumulation and release Enhance industry understanding of accumulation/release mechanisms Evaluate effectiveness and costs of mitigation strategies for the Park City system Technical Approach Field sampling Laboratory studies Main cleaning trials

9 PCMC is a Very Complex System For Its Size 8,500 base population 8 existing sources Shoulder season demand swings Elevation range = 6,500 10, pressure zones 26 pump stations 19 storage reservoirs with capacity of 14,650,000 gallons 120 miles of pipeline

10 Naturally Occurring Metals in Water Sources Judge Tunnel Antimony Cadmium Lead Mercury Zinc Flows: 600-2,500 gpm Spiro Tunnel Antimony Arsenic Cadmium Selenium Thallium Zinc Flows: 2,000-8,900 gpm Meets all MCLs through treatment and blending

11 Research Need Fall 2007 & 2010 customer complaints of discolored water Samples exceeded MCLs in distribution system Arsenic, thallium, manganese and mercury Lead in 2007 event Since the MCL exceedances were in distribution system, not at the entry points, no violations Treated by Park City as public health concern Intensive sampling Public information campaign, bottled water Worked in collaboration with Utah DDW 11

12 WQ Monitoring and Lab Experiments Laurie McNeill Professor Utah State University

13 Distribution System Monitoring How do you know if you are heading towards a colored water event, or worse, elevated TIC concentrations without visual evidence of an event?

14 Routine Monitoring Monthly sampling at 14 DS sites for 1 year 6 pump stations 2 PRVs 5 premise taps 1 tank 8 sources sampled twice

15 Routine Monitoring Field parameters: ph, temp, DO, EC, ORP, Cl 2 residual, turbidity Samples analyzed at USU for: Total and dissolved TICs DOC and TOC Major anions and cations Alkalinity TSS, TDS, VSS

16 Routine Monitoring

17 Particulate arsenic (µg/l) Routine Monitoring: Arsenic Tunnel source on off on Tunnel source off

18 Temporal Analysis of Monitoring Data Determine if TIC releases occurred TIC release = 100% increase beyond expected concentration Account for changes in site-specific source blend 1,040 TIC samples analyzed Are TIC releases related to chemistry shifts and/or Fe/Mn co-occurrence?

19 Temporal Analysis of Monitoring Data Only 22 releases identified in 1,040 TIC samples As (3), Pb (8), Cu (4), Tl (7) 15 releases were particulate 11 were correlated with Fe/Mn Only 1 particulate release produced visible color despite co-occurring Fe/Mn

20 Temporal Analysis of Monitoring Data Only 22 releases identified in 1,040 TIC samples 15 releases were particulate 7 releases were soluble All 7 releases associated with chemistry shift(s) Parameter ph Cl 2 Residual ORP TDS System Specific Condition 7.5; 8.5; 0.5 or more between individual samples 0.2 mg/l 300 mv 1,000 mg/l; 500 mg/l or more between samples 4 releases were Tl, all associated with low Cl 2

21 Temporal Analysis of Monitoring Data Only 3 excursions (> MCL or AL) out of 1,040 samples (0.3%) 1 particulate Pb 1 particulate Tl 1 soluble Tl In the DS, so not a violation The system is in really good control, despite potential high risk due to legacy accumulation

22 Routine Monitoring Please see the final report for full details on TIC release results Routine snapshot monitoring is: Resource-intensive and impractical for on-going surveillance monitoring But helpful in focusing future efforts Useful for our research purposes

23 Lab Studies Desorption of TICs from DS solids Biofilm grown from DS solids 22 Weeks challenges T = 16 C DOC = 1 mg/l Cl 2 = 0.2 mg/l T = 7, 16, 25 C DOC = 0, 1, 2 mg/l Cl 2 = 0, 0.2, 2 mg/l

24 Lab Studies: Thallium and Manganese Increased chlorine (2 mg/l vs. 0 or 0.2 mg/l) resulted in retention of Tl and Mn in the biofilm matrix and less Tl in column effluent Mechanism: Mn oxidized by Cl 2 Tl sorbs to Mn oxides Tl is highly exchangeable Mn 2+ +Cl 2 MnO 2 (s) Tl Mn oxide Tl Tl Similar results found in desorption study and monitoring data: Cl 2 is critical for stabilizing existing Mn and Tl

25 Main Cleaning Strategies for Removal of Metals-Rich Deposits Andrew Hill, P.E. Project Manager Confluence Engineering Group, LLC

26 Toolbox for Assessment and Control Monitoring System-specific assessment DS bulk water quality monitoring Collection & analysis of deposits Treatment Reduce substrate/metals loading Stabilize finished water chemistry Dist. System O&M Mains cleaning to remove legacy metals Stabilize dist. system water chemistry Controlled blending / source isolation

27 Nature of Legacy Deposits Non-scale-forming pipes Thin, slimy films with cohesive sludge Fe/Mn particulates Co-precipitated trace metals (esp. As, Pb, Tl, Zn) Biofilm Pipe Tap Asbestos Cement Harvested Specimen CML Ductile

28 Main Cleaning Guidance Goal is to select the best main cleaning technique for the job at-hand Site-Specific Factors Pipe Conditions Pipe Material Nature of Deposits Cleaning Objectives Water Quality LOS Asset Management Technique-Specific Factors Performance Contaminant Removal Risks Cost to Conduct Implementation Life-Cycle Costs

29 Field Investigation of Main Cleaning Performance Full-Scale Cleaning Trials UDF at 3 fps and 6 fps Foam swabbing Ice pigging Pipe Specimen Extractions Characterize legacy deposits Before vs. After visual applied at 2 separate sites: AST and UPA - Adjacent Cleaning Loops - Controlled Conditions - Intensive Monitoring Apples-to-Apples Comparison

30 Unidirectional Flushing (UDF) Grab samples for AST Peak TSS = 10 mg/l

31 Foam Swabbing Grab samples for AST Peak TSS = 1,000 mg/l

32 Ice Pigging Grab samples for AST Peak TSS = 900 mg/l

33 Pre-Cleaning vs. Post-Cleaning After UDF at 6 fps After Ice Pigging

34 Total Solids Removed (mg/ft 2 ) Solids Inventory Removed 1,000 AST Site UPA Site % 59% 95% 88% % 7% 7% 12% UDF Ice Pig Swab UDF Ice Pig Swab For UDF: 3 fps 6 fps For Swab: 1 run 2 runs % indicates mass removed compared to 2-pass swabbing

35 Contaminant Removal Trends Composition of solids removed also differs by technique Solids removed by flushing are not simply a subset of the whole Fe and As preferentially removed by UDF Mn, Pb, Tl, TOC (biofilm) less effectively removed by UDF

36 Conceptual Deposit Profile Loose surficial particles Deeper cohesive films UDF Ice Pigging Swabbing

37 So What s a Utility to Do? Should I be swabbing or ice pigging? Is it even worthwhile to flush? And what about costs??

38 Main Cleaning Desktop Cost Analysis Cost Estimates Support Utility Planning Cleaning project budgeting and resource allocation Technique selection via life-cycle cost comparison Not appropriate to select a technique based solely on cost, given differences in cleaning performance and feasibility For example, if stubborn films are present, frequent UDF is not a substitute for less frequent application of swabbing or ice pigging

39 Three Key Goals for Cost Estimate Development 1. Comprehensive Capital Staff Labor for Planning Crew labor and resources for field work Third-party vendor support 2. Detailed and reliable Clearly-defined conditions and inputs Line-item breakdown 3. Comparative Similar project size, scope, and conditions Normalized to $ per pipe-mile

40 Results from Cost Analysis Disclaimer Cost estimates are sensitive to inputs used. Results are presented for relative comparison purposes. Refer to final report for line-item breakdown and underlying assumptions Breakdown allows utilities to modify inputs/variables and developed tailored costs

41 Summary of Application Costs

42 Life-Cycle Cost Approach Net Present Value approach Equivalent Annual Cost (EAC) Various application frequencies (t = 2-10 yrs)

43 Equivalent Annual Cost ($ per pipe-mile) Equivalent Annual Costs by Application Frequency $4,000 $3,500 2-year Recurrence Interval 5-year Recurrence Interval 10-year Recurrence Interval $3,000 $2,500 $2,000 $1,500 $1,000 $500 $0 UDF Ice Pigging Swab w/ Swab w/ Hydrants Stations UDF Ice Swab w/ Pigging Hydrants Swab w/ Stations UDF Ice Pigging Swab w/ Hydrants Swab w/ Stations Capital Investments Utility Labor Other Utility Resources Third Party/Vendor

44 Need to also Consider Risks and feasibility issues Higher / more challenging with swabbing Availability of in-house resources With vendor support, swabbing may cost more Ability to dispose of cleaning residuals TSS/Metals TDS

45 Main Cleaning Summary Routine preventive main cleaning is critical to reduce metals inventory and exposure Plastic and cement pipe are not clean or risk-free High-velocity UDF has its limitations Manganese films and biofilm are largely untouched Utilities should consider site-specific conditions and objectives to select the most appropriate technique Main cleaning is expensive and labor intensive Prioritize locations and frequency based on risk factors and system-specific monitoring

46 Industry Recommendations

47 Accumulation and Release Control Framework Step 1. Assess Existing System Conditions and Vulnerability Deposit type and prevalence Water chemistry stability Data Collection & System Understanding Step 2. Address Legacy Deposits and Contaminants Step 3. Reduce On-going Loading Remove loose/cohesive deposits Stabilize remaining deposits Main Cleaning and Chemistry/Hydraulics Control Contaminants (regulated metals, microbes) Substrates (iron, manganese, hardness) Nutrients (ammonia, iron, manganese) Treatment/Removal Source: Adapted from Friedman and Hill et al., 2010, WRF #3118

48 Need to Halt the Progression to Avoid Exposure -Aggressive cleaning -Swabbing -Ice pigging -Pigging and relining -Other techniques -Source water treatment -Unidirectional flushing -Chemistry stabilization -Maintain chlorine residual/ oxidizing conditions -Flow direction and velocity control -Avoid conventional flushing

49 Develop a Data-Driven Integrated Main Cleaning Approach If risk factors warrant, use an integrated main cleaning strategy instead of an either/or Risk Factors: Known inventory of legacy TICs History of release On-going loading Chemistry stability High chlorine demand areas Prioritize aggressive cleaning in higher risk areas Continue UDF/Monitoring in lower risk areas

50 Conventional Flushing Not recommended as a routine flushing approach Can t control flow direction or velocities Stirs stuff up Not a main cleaning approach, it is a water moving approach Benefits may be very short lasting If have to use it Keep flow rates low Visual clarity and elapsed time not recommended for conventional flushing termination criteria Use turbidity ( 2NTU) AND chlorine

51 Baseline Monitoring What does routine, baseline monitoring tell us? Can observe differences spatially and seasonally Can observe impacts of source changes MAYBE can pick up release event But cannot assess duration, frequency, or magnitude Cannot assess exposure health effects risks without this information Would need event based or continuous monitoring in high risk areas to better characterize releases Help understand water quality conditions and potential problem areas

52 Investigative Monitoring Suggested Classification Scheme Source Tracking Trace Metals Substrates Chemistry Source Use Records Chloride-to-Sulfate Ratio Conductivity/TDS Arsenic (p/s) Lead (p/s) Copper (p/s) Thallium (p/s) Antimony (p/s) Iron (p/s) Manganese (p/s) Turbidity (indicator) Color (indicator) ph Oxidation-reduction potential (ORP) Disinfectant Residual TDS PO4 Competing ions Alkalinity p = particulate s = soluble Metal speciation very important for understanding nature of release

53 Conclusions (1) Accumulation and release of TICs is a complicated problem involving chemical, physical, and biological mechanisms Stable chemistry and adequate disinfectant residual are critical for minimizing soluble releases Cannot rely on visual clarity alone as indicator of TIC stability Do not tell customers that discolored water is only an aesthetic issue unless you have done the monitoring to prove it

54 Conclusions (2) Unidirectional flushing important, but will not return pipes to clean pipe condition. More aggressive techniques needed. Conventional flushing is not a main cleaning technique and should be used very carefully. We have the tools to prevent accumulation and release but requires source-to-tap effort. Risk-based prioritization approach needed.

55 Thank You Comments or questions, please contact: For more information visit: