How are My Filters Doing? Filter Profiling Reveals All Laurel Passantino and Jacqueline Rhoades Water Treatment Seminar AZ Water February 18, 2014 Statements of fact and opinion expressed are those of the author(s) and presenter(s). AZ Water Association, AZAWWA, and AZWEA assume no responsibility for the content, nor do they represent official policy of the Association. Motivation Why profile filters? Monitoring and Evaluation What should be watched? What tools and techniques are available? Assess Effectiveness Solution Implementation What are some common fixes? Long-Term Operation What ongoing activities are needed? Case Studies 1
Motivation: Why Profile Your Filters? Sometimes, you can just tell Reasons to Profile Filters Diagnose obvious failure (Filter Forensics) Plan for process changes Create baseline for routine operations Optimize performance and cost Evaluate end-of-life conditions 2
Motivation Why profile filters? Monitoring and Evaluation What should be watched? What tools and techniques are available? Assess Effectiveness Solution Implementation What are some common fixes? Long-Term Operation What ongoing activities are needed? Case Studies What Data Should be Captured? Water quality parameters Individual filter effluent turbidity Combined filter effluent turbidity Operational parameters Filter runtime Headloss development profile Ripening time Backwash parameters Backwash duration Backwash volume 3
Turbidity (NTU) 3/14 3/21 3/28 4/4 4/11 4/18 2/18/2014 Understanding Long-term Trends is Critical! 2.5 2 1.5 Turbidity Limit for Unfiltered Supply 1 0.5 0 Qualitative Observations Also Important Normal Backwash Even Surface after Draining vs. Backwash with Hot-Spots vs. Surface Potholes 4
Visual Inspection Can Be the First Line of Defense Boiling during wash Uneven wash distribution Uneven overflow into troughs Cratering Mudballs Cracking at surface Separation at walls Filter media in troughs Cratering suggests possible damage to underdrain. Uniform surface Washwater troughs level Filter Inspection has Two Components Physical Characteristics Media depth Grain size distribution L/D Media support (underdrain, gravel) Operational Parameters Media Expansion Solids retention analysis Backwash Profile 5
Be Prepared! Have a health and safety plan Confined Space Fall prevention Lockout-tagout Drain the filter Work in teams Review available resources AWWA B100 Filter Maintenance and Operations Guidance Manual (AWWARF 90908) J Know Your Tools! Maneuvering - Ladder to enter/exit, ¾ inch plywood to stand on, carabiners, and rope for transferring samples and equipment to the filter deck Measurement - Shovel, level, 3/8 inch steel rod, tape measure Documentation - camera, clipboard, datasheets, object for scale in pictures Filter Coring - 1.5 inch electrical conduit, 5 foot length, baggies Bed Expansion - One-inch interval tubes or cups Analytical - Turbidimeter, glassware, balance, sample bottles, baggies 6
Steps to Performing a Filter Inspection Visual Observation Observe surface or air wash effectiveness View surface for boils or hot spots Look for uneven wash areas or uneven troughs Media Depth Core Sampling Sieve Analysis/ Media Assessment Solids Retention Analysis Backwash Turbidity Profile Quantify media loss by measuring the depth to gravel. Understand stratification by knowing the depth of layer of filter media, including the mixed zone. Take samples at 0-2 inches, 2-6, 6-12, 12-18, 18-24 Sample before and after backwash Collect multiple cores for a representative matrix Send composite sample to lab. Compare size distribution, effective size, uniformity coefficient to design specs. Calculate L/D Used to evaluate effectiveness of backwash Can show too little or too much backwash Keep historical records Collect samples every minute of backwash Analyze turbidity and plot over time Helps prevent excessive washing Determine Media Depth Using Probing and Core Sampling Check to see if troughs are level, then measure distance from trough to bed and check for mounding Use steel rod to probe media, or dig into it to determine depth to gravel Use core sampling tool and baggies to obtain samples at various depths If filter is dual or mixed bed, note depth of each strata, and depth of mixed interface Collect multiple cores to develop a representative matrix of the filter bed 7
Send Core Samples for Sieve Analysis Know the original specifications Understand the key parameters Effective size - particle size opening that will just pass 10% (by dry weight) of a representative sample of filter material (D 10 ) Uniformity Coefficient - calculated ratio of the particle size opening that will just pass 60% (by dry weight) of a representative sample of filter material divided by the size opening that will just pass 10% (by dry weight) of the same sample (D 60 /D 10 ) Media Depth - measured by probing or coring the filter bed L/D ratio - divide media depth (L) by the nominal media diameter (D) to understand the particle barrier (>1100 for low NTU production) Interpret Sieve Analysis Results Effective Size is smaller than specified Larger grains may have broken down due to abrasion during backwashing over time Effective Size is larger than specified Smaller grain size media has been lost over time Uniformity Coefficient is a measure of the grain size distribution The closer to one, the narrower the grain size distribution If larger than specified, a wider grain size distribution is present. This could lead to more intermixing between the sand and the anthracite, resulting in too large of a mixed zone, and decreased filter performance 8
Calculate L/D Original specification for a dual media filter 36 inches of 1 mm anthracite 6 inches of 0.5 mm sand 42 total inches Filter inspection revealed 32 inches of anthracite 6 inches of mixed layer 3 inches of sand 41 total inches Rough L/D calculation (32in X 25.4)/1mm = 813 (6in X 25.4)/0.75 mm = 203 (3in X 25.4)/0.5mm = 152 L/D = 1168 adequate particle barrier for low NTU production Assess Backwash Effectiveness with a Solids Retention Analysis Collect core samples Take samples at 0-2 inches, 2-6 inches, and every 6 inches thereafter for the entire bed depth Use same locations before and after backwash Rinse using AWWA protocol Rinse 50 grams of sample in 100mL of DI water 5 times (500 ml total) and measure turbidity Multiply turbidity by 2 and plot against bed depth Bed depth vs. (NTU/100g media) Add mobile lab photo 9
Turbidity (NTU) 2/18/2014 Assess Bed Cleanliness After Backwash 10-30 NTU: Very clean, not well ripened. Bed is too clean, examine wash rate and length 30-60 NTU: Clean and partially ripened. No need for action 60-120 NTU: Reasonably clean, well ripened. Slightly dirty, reschedule retention analysis soon. 120-300 NTU: Dirty media, well ripened. Evaluate filter wash system procedures. 300-600 NTU: Dirty media with mudballs. Rehab bed! 400 350 300 Use Backwash Turbidity Profile to Determine Backwash Duration 250 200 150 100 50 0 0 5 10 15 Backwash Time (min) AWWA Standard = 10 NTU 10
Motivation Why profile filters? Monitoring and Evaluation What should be watched? What tools and techniques are available? Assess Effectiveness Solution Implementation What are some common fixes? Long-Term Operation What ongoing activities are needed? Case Studies Implement Solutions to Improve Performance Add more media Change media type Pre-treatment process chemistry Modify backwash duration Implement temperature dependent high flow backwash rate (account for water temperature to achieve bed expansion) Address stratification of filter bed through staged backwash programming Good operational practices 11
Motivation Why profile filters? Monitoring and Evaluation What should be watched? What tools and techniques are available? Assess Effectiveness Solution Implementation What are some common fixes? Long-Term Operation What ongoing activities are needed? Case Studies Periodic Evaluations Lead to Long-Term Success Once per quarter (per season) Adjust high flow rate for temperature Check media expansion make adjustments Review unit filter run volume data Once per year Check media depth Core the filter solids retention Review all filter profiles Every three to five years Send media to lab for sieve analysis Add media if necessary BUT know why it s being lost 12
Case Studies 1. Catastrophic failure filter forensics 2. Commissioning a new treatment process 3. Retrofit filters for new water quality goals 4. Routine maintenance 1. Filter Forensics: Underdrain Failure Examining media and backwash procedures helps identify contributing factors Tools & Techniques Visual observation of backwash Visual observation of media surface Media coring Physical characteristics 13
Average Run Time (hrs) 2/18/2014 1. Observations and Conclusions Hot-spots during backwash Uneven surface Mixed sand & anthracite Media Anthracite Sand Filter No. 15 16 18 15 16 18 Design Depth 20 in. 10 in. In-place Depth 12.5 12.5 12.3 10 9.5 13 Design ES 0.45-0.55 0.9-1.1 In-place ES 0.5 0.5 0.5 1.0 1.0 1.0 Design UC <1.4 <1.4 In-place UC 1.33 1.35 1.30 1.33 1.41 1.37 Replace underdrains Replace (not replenish) media Change backwash strategy to ensure stratification 2. Commissioning a New Process Filter run times (terminated by headloss) shorter than anticipated at new WTP Potential factors Coagulation chemistry Abrasion concern with GAC media Backwash practices Tools & Techniques Trending runtime and headloss Visual observation of backwash and media surface Media coring and floc retention analysis Sieve analysis (media abrasion) 18 16 14 12 10 8 6 4 2 0 2/6 2/16 2/26 3/8 3/18 14
Depth from Media Surface (in) 2/18/2014 2. Observations and Conclusions Mat on surface Clumps during & after backwash Particles not penetrating bed Turbidity of Rinse Water (NTU) 0 50 100 150 200 250 0-10 -20-30 Before BW -40 After BW -50-60 -70-80 Modify coagulation chemistry (coagulant aid, filter aid) Standardize backwash protocols No further action required after sieve analysis Dual media filters replaced with GAC 3. Retrofit Filters for New Water Quality Goals Filter runtimes (terminated by turbidity breakthrough) were shorter than anticipated Tools & Techniques Trending runtime and headloss Visual observation of backwash and media surface Media coring and floc retention analysis 15
Filter Runtime (hr) 2/18/2014 3. Observations and Conclusions Add sand to increase L/D Pre- Treatment Modified Pre-Treatment Returned to Original Conditions Pre-Treatment Impacted Filter Runtimes Media ~L/D 36 GAC 760 42 GAC 1070 60 GAC 1260 36 GAC + 6 Sand 1100 4. Routine Maintenance 10 year old media: Is media replacement needed? Tools & Techniques Visual observation of media surface Measurement of media levels Media coring Physical characteristics Visual observation of backwash 16
4. Observations and Conclusions Do not replace, replenish to design levels (add 4 to 6 inches of new anthracite) Carefully monitor backwash practices 10 ANTHRACI TE SAN D GRAVEL ANTHRACI TE SAN D GRAVEL 1 Filter Profiling Reveals All! Many tools are at your disposal Don t wait for a failure Begin to establish filter baselines now There s always room for improvement 17
Questions? Laurel Passantino laurel.passantino@asu.edu Jacqueline Rhoades jrhoades@hazenandsaywer.com 18