Improving GAC Filter Operations at SCWA Joseph Roccaro Suffolk County Water Authority NYSAWWA Tifft Symposium Sept. 18, 2014
GAC Pilot Phase 1 Results 1,1-DCA: Effect of Virgin GAC Type 1.6 1.4 1.2 1.0 C/C 0 0.8 0.6 0.4 0.2 0.0 0 3,000 6,000 9,000 12,000 15,000 Bed Volumes AC (EBCT = 11 min) DC L CS
Outline GAC Characteristics Specifications Tracking GAC Performance Recent SCWA Work Pilot Testing Contract Specifications
What is activated Carbon? Crude form of Graphite Random or Amorphous structure Highly porous Wide range of pore sizes Visible and molecular cracks and crevices 45 µm dust particle Cluster of 100 molecules Courtesy of Calgon Carbon Corp.
Low Density, Soft, Weakly Adsorbing World of Raw Materials for AC Vegetable Waste Rice HullsCorn Cobs Biomass Fermentation Residues Wood Chips and Sawdust Pine Oak Walnut Teak Ebony Pits, Stones, and Kernels Cherry pits Peach pits Palm Kernels Olive Stones Nut Shells Almond Pecan Walnut Coconut Babassu Coals Coalification Series Peat Lignite Bituminous Anthracite Courtesy of Calgon Carbon Corp. High Density, Hard, Strongly Adsorbing
Starting Materials Critical Properties Ash level and constituents Density Hardness Inherent Transport porosity (permeability) Building Blocks for Adsorption Structure
Description of the Two Types of Activated Carbon Pore Structures Transport Pore Structures Carbon Highways Larger Pores which never adsorb Act as diffusion paths to transport adsorbates 25% of GAC particle Dictate adsorption kinetics Adsorption Pore Structures The finest pores in the carbon structure. Carbon Parking spaces Have adsorption capabilities 40% of GAC particle Define adsorption thermodynamics Transport Pore:Adsorption Pore Varies w/ GAC type; activation process Find the correct parking lot; parking spaces
Carbon Molecular Structure Coal based carbon Aliphatic dislocation of platelet Graphite Platelet 100 angstroms Inter bounding of plates Skeleton 35% Adsorption pores 40% Courtesy of Calgon Carbon Corp. Transport pores 25%
Carbon Molecular Structure Coconut based Carbon Aliphatic dislocation of platelet Graphite Platelet 100 angstroms Inter bounding of plates Skeleton 40% Adsorption pores 45% Courtesy of Calgon Carbon Corp. Transport pores 15%
Carbon Molecular Structure Wood based Carbon Aliphatic dislocation of platelet Graphite Platelet 100 angstroms Inter bounding of plates Skeleton 20% Adsorption pores 35% Courtesy of Calgon Carbon Corp. Transport pores 45%
Order of Attack on Molecular Structure Aliphatic dislocation of platelet 4 2 5 Graphite Platelet 4 2 2 5 2 3 3 5 4 1 1 4 100 angstroms 1 4 3 3 4 4 1 2 3 3 2 5 5 5 4 2 4 4 5 4 Inter bounding of plates 1. Edge dislocations 2. Edges of small plates 3. Internal plate dislocations 4. Edges of large plates 5. Edges inter bonded with other plates Courtesy of Calgon Carbon Corp.
Molecular Structure of 0% Burn off Activated Carbon Calcined CMS Carbon Aliphatic dislocation of platelet AD 0.80g/cc, Iodine No. <200 Adsorption Pore Volume 0.15 cc/g Abrasion No. 98, Ash 3.0% Graphite Platelet 100 angstroms Inter bounding of plates Courtesy of Calgon Carbon Corp.
Molecular Structure of 20% Burn off Activated Carbon AFC-2204 Carbon Aliphatic dislocation of platelet AD 0.64g/cc, Iodine No. 600 Adsorption Pore Volume 0.28 cc/g Abrasion No. 90, Ash 3.75% Graphite Platelet 100 angstroms Inter bounding of plates Courtesy of Calgon Carbon Corp.
Molecular Structure of 20% Burn off Activated Carbon AFC-2204 Carbon Aliphatic dislocation of platelet AD 0.64g/cc, Iodine No. 600 Adsorption Pore Volume 0.28 cc/g Abrasion No. 90, Ash 3.75% Graphite Platelet 100 angstroms Inter bounding of plates Courtesy of Calgon Carbon Corp.
Molecular Structure of 40% Burn off Activated Carbon F400 Carbon Aliphatic dislocation of platelet AD 0.48g/cc, Iodine No. 1050 Adsorption Pore Volume 0.48 cc/g Abrasion No. 80, Ash 5.0% Graphite Platelet 100 angstroms Inter bounding of plates Courtesy of Calgon Carbon Corp.
What is adsorption? Intermolecular attractions in these smallest pores result in adsorption forces that Cause condensation of adsorbate gases Precipitation of adsorbates from solutions Adsorption forces are a result of: Interactions of the carbon structure with outer bonding electrons of the adsorbate molecules. For similar type and size bonds: The greater the number of bonds per unit volume molecule, the greater the adsorption force.
Relationship between Structure and Adsorption Force Carbon Atom Adsorbate Molecule Carbon Skeletal Structure The Adsorption force present at the adsorption site is the sum of all the individual interactions between carbon atoms and the adsorbate molecule. (10,000,000 X Magnification) Courtesy of Calgon Carbon Corp.
Relationship between Structure and Adsorption Force London Dispersion Force Field for One Graphite Plate Butane Adsorbate Molecule Graphite Plate of Carbon skeletal Structure Adsorption Force Field Strength for Butane Adsorption = 0.5 Kcal/mole Butane adsorbed = 1.0 Kcal/mole Butane adsorbed = 2.0 kcal/mole Butane adsorbed = 4.0 Kcal/mole Butane adsorbed Courtesy of Calgon Carbon Corp.
Relationship between Structure and Adsorption Force London Dispersion Force Field for Two Graphite Plates Adsorption Force Field Strength for Butane Adsorption = 0.5 Kcal/mole Butane adsorbed = 1.0 Kcal/mole Butane adsorbed = 2.0 kcal/mole Butane adsorbed = 4.0 Kcal/mole Butane adsorbed = 8.0 Kcal/mole Butane adsorbed Butane Adsorbate Molecule Graphite Plate of Carbon Skeletal Structure Courtesy of Calgon Carbon Corp.
GAC Pilot Phase 1 Results 1,1-DCA: Effect of Virgin GAC Type 1.6 1.4 1.2 1.0 C/C 0 0.8 0.6 0.4 0.2 0.0 0 3,000 6,000 9,000 12,000 15,000 Bed Volumes AC (EBCT = 11 min) DC L CS
1,1-DCA H Cl H H C C H Cl 1,1 Dichloroethane (1,1 DCA)
Common & Useful GAC Specification Parameters Size Distribution / Physical Attributes Apparent Density Effective Size Uniformity Coefficient (U.C.) Moisture content Ash Content Hardness/Abrasion Adsorptive Performance Tests
Apparent Density Density of a packed bed of GAC particles, filled in a standard cylinder in a manner to minimize voids between particles Provides info on type of carbon and changes undergone during reactivation. Bituminous = 28 41 lb/ft 3 ** Coconut = 28 35 lb/ft 3 ** Lignite = 22 26 lb/ft 3 ** ** - from AWWA B604
Effective Size / U.C. Effective Size (E.S.) Screen Size which holds 90% of the carbon above it (mm) Uniformity Coefficient (U.C.) = Ratio of size opening that will pass 60% of material divided by that opening that will pass 10% of same sample AWWA B604 UC 2.1. Effect bed packing; dp; adsorption kinetics
Abrasion Number Measure of structural strength AWWA B604 measures loss of MPD when subjected to action of stirrer or steel balls Ability to withstand handling; slurrying; hydraulic impacts
Ash Content / Moisture Content Ash Content Helps to identify the type of GAC and indicate quantity of extractable material. Non-useable impurities on GAC; paying for weight Contributes to metals leaching; ph at start-up Indicator of minerals present; often insoluble Moisture Content Needed to calculate dry wt.; cost calculations Paying for weight of water AWWA (<8%)
Adsorptive Performance Tests (Activity Parameter Tests) BET Iodine # Butane # TCN Tannin #
BET Surface Area Measure of overall surface area Larger surface area = potential for greater adsorption capacity Measures the amount of N 2 gas adsorbed by GAC Known surface area occupied by N 2 Uses Brunauer-Emmett-Teller (BET) isotherm eqn. N 2 small M.W. & size allows access to adsorption pores ASTM D 3037
Iodine (I 2 ) # Activity Test Measures the amount of 0.02N I 2 solution that will adsorb under specific conditions (ASTMD4607). Results expressed as mg iodine / gm carbon Test inexpensive, fast, reproducible Limitations/problems: Reacts with ash, adsorbates, oxygen on carbon Measures volume present in pores from 10 28 Å May not correlate well with trace contaminant removal
Trace Contaminant Number TCN TCN Acetoxime Relatively new test method Acetoxime very soluble in water Measures smallest pore fraction liquid phase isotherm test; Measure Acetoxime uptake via UV adsorption AWWA B604-12 Few labs perform test TCN-G G = gas = CF4 (tetrafluoromethane). Weakly adsorbed; small MW compound Measure wt
Tracking GAC Performance How to determine how well GAC performed? A few key parameters Bed Volumes (BV) Empty Bed Contact Time (EBCT) Carbon Use Rate (CUR) Definition of breakthrough
GAC Bed Volume Volume occupied by GAC bed Example: Two (2) 20K vessels Assume GAC density = 30 #/ft 3 40,000 # GAC 30 #/ft 3 = 667 ft 3 GAC 667 ft 3 x 7.48 Gal / ft 3 = 9973 Gal. = 1BV BV Treated Gallons treated / 10,000 Gal. Normalizes data w/different Q; vessel size
EBCT & CUR EBCT = Empty Bed Contact Time Volume of GAC Bed Q Expressed in minutes Removal of specific contaminants ƒ(ebct) Recommendations for contaminant removal CUR = Carbon Use Rate Amount of GAC used at contaminant breakthrough expressed as # GAC/MG; # GAC/KGal ; mg/l
Breakthrough in GAC Filter
Down flow Adsorption column Inlet Outlet Headspace Saturated Virgin Mass Transfer Zone Courtesy of Calgon Carbon Corp.
WRF Project #4235 GAC Pilot Study - Phase 1 8 pilot columns 6 dia. GAC bed depth: 8.5 ft (2.6 m) GAC mesh size: 8 x 30 U.S. Standard Mesh 11 min & 22 min EBCT Compared VOC removal: virgin vs. reactivated GAC direct activated vs. re-agglomerated coconut GAC lignite GAC effect of BW; EBCT
GAC Pilot Study (Avg. Influent Concentrations) 1,4 dioxane = 2.3 ug/l 1,1-dichloroethane (1,1-DCA) = 2.2 ug/l 1,2-dichloroethane (1,2-DCA) = 0.9 ug/l 1,1,2-trichlorotrifluoroethane (TCTFA; Freon 113) = 4.2 ug/l 1,1,1-trichloroethane (1,1,1-TCA) = 3.2 ug/l cis-1,2-dichloroethene (cis-1,2-dce) = 1.2 ug/l 1,1-dichloroethene (1,1-DCE) = 2.1 ug/l Carbon tetrachloride (CT) = 0.9 ug/l 1,2,3-trichloropropane (1,2,3-TCP) = 0.8 ug/l Tetrachloroethene (PCE) = 3.2 ug/l Trichloroethene (TCE) = 3.7 ug/l
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA)
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA) 1,2 dichloroethane (1,2-DCA)
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA) 1,1,2-trichlorotrifluoroethane (TCTFA) 1,2 dichloroethane (1,2-DCA)
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA) 1,1,2-trichlorotrifluoroethane (TCTFA) 1,2 dichloroethane (1,2-DCA) 1,1,1-trichloroethane (1,1,1-TCA)
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA) 1,1,2-trichlorotrifluoroethane (TCTFA) 1,2 dichloroethane (1,2-DCA) 1,1,1-trichloroethane (1,1,1-TCA) cis-1,2-dichloroethene (cis-1,2-dce)
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA) 1,1,2-trichlorotrifluoroethane (TCTFA) 1,2 dichloroethane (1,2-DCA) 1,1,1-trichloroethane (1,1,1-TCA) cis-1,2-dichloroethene (cis-1,2-dce) 1,1-dichloroethene (1,1-DCE)
GAC Pilot Phase 1 Results Contaminant Breakthrough Order 1.4 1.2 1.0 0.8 C/C 0 0.6 0.4 0.2 0.0 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 Bed Volumes 1-4 dioxane 1,1-dichloroethane (1,1-DCA) 1,1,2-trichlorotrifluoroethane (TCTFA) 1,2 dichloroethane (1,2-DCA) 1,1,1-trichloroethane (1,1,1-TCA) cis-1,2-dichloroethene (cis-1,2-dce) 1,1-dichloroethene (1,1-DCE) carbon tetrachloride (CT)
Conclusions GAC Pilot Phase I GAC not good for 1,4-dioxane removal. 1,1-DCA - least adsorbable of VOCs present. 1,2,3-TCP, TCE, and PCE - most adsorbable of VOCs present. No detectable breakthrough after 30,000 BV. GAC type important for VOC removal Direct activated bituminous - least effective Coconut-shell based - most effective Bituminous - Re-agglomerated outperformed direct activated Lignite - slightly less effective than re-agglomerated coal-based GAC on a bed volume basis, but slightly more effective on a mass-based carbon usage rate
Conclusions GAC Pilot Phase 1 (cont d.) Reactivated GAC Acceptable removal of contaminants: Agglomerated reactivated outperformed its virgin counterpart for all adsorbates except 1,1,2-TCTFA & 1,1,1-TCA. Direct activated reactivated GAC outperformed its virgin counterpart for all adsorbates except 1,2-DCE & CT. Metals leaching a potential problem - Agglomerated Bituminous EBCT effect on VOC removal: EBCT from 11 to 22 min. = 13% in CUR Backwashing effect on VOC removal No measurable effect. EPA proposed cvoc Rule a major concern
Metal Leaching from Reactivated GAC Phase 1 GAC Pilot
GAC Pilot Study - Phase 2 Virgin vs. reactivated; coconut GAC GAC bed depth: 4.25 ft GAC mesh size: 8 x 30 U.S. 5.5 min EBCT Compared VOC removal: - virgin vs. reactivated bituminous - coconut GAC: Iodine #: 900 1300 Tested performance predicting surrogate parameters TCN; Dye; TAC TIC
Phase 2 - GAC Performance Coconut # 1 Coconut # 2 Coconut # 3 RBA Coconut # 4 VBA RBD VBD BV to 50% Breakthrough of 1,1-DCA CUR (#GAC/MG) 12500 12000 11500 11400 10000 8200 4500 4500 336 334 362 405 516 625 897 855 I2 # 1130 1190 1432 929 870 832 864 978 TCN # 14.5 15.4 14 12.7 14.9 12.6 6.8 6.7 R = reactivated; B = bituminous; V = virgin; D = direct activated; A = agglomerated
Metal Leaching from Reactivated GAC Phase 2 GAC Pilot B.V. Sb As Mo Ni Se V 14 4.22 6.3 15.3 30 2.56 3.56 1.51 10.2 50 1.28 1.79 2.3 5.78 85 2.91 4.12 115 3.19 3.1 190 2.9 1.38
Conclusions GAC Pilot Phase 2 Reactivated Bituminous GAC Confirmed acceptable removal of contaminants Confirmed metals leaching in agglomerated GAC a potential problem for SCWA. Different sources for spent GAC also had metals carryover. TCN appears to be good predictor of performance Coconut GAC confirmed as effective for VOC removal I# is good predictor of performance
GAC Pilot Study - Phase 3 Looking @ reactivated vs. acid washed reactivated GAC GAC bed depth: 4.25 ft GAC mesh size: 8 x 30 U.S. 5.5 min EBCT Study underway AW Reactivated performing better than non-aw Reactivated
Vanadium (ug/l) Phase 3 Metals Results Vanadium Concentration 18 16 14 12 10 8 Reactivated GAC AW reactivated GAC 6 Common Influent 4 2 0 0 20 40 60 80 100 120 Bed Volumes
SCWA Current GAC Spec s. Contractor: previous experience; mechanical capability GAC: Virgin bituminous & coconut (8 x 30) Iodine # = 1000 (coconut) / 900 (bituminous) Water Soluble Ash 1% (max.) Abrasion # - 75 (min.) Leachable metals after specified rinse: As = 2.5 ppb Sb = 0.4 ppb Others not above background
Best Surrogate for GAC Performance? CS R-AC L AC R-DC DC CUR #/MG 460 590 600 730 930 1900 Iodine # 1174 812 659 847 899 870 TCN # 15.4 15.6 2.8 13.5 7.5 9.1 TAC TIC 2.4 2.36 1.08 2.26 1.42 1.71 Dye # 35.5 41.3 95.6 47.3 41.9 36.3 BET (m 2 /g) 855 550 611 626 716 742 A.D. (g/cc).494.595.386.593.49.507 Adsorption Pore Transport Pore 39% 32% 31% 37% 34% 35% 23% 23% 39% 18% 29% 25%
Improvements to GAC Use @ SCWA 2014: Began full-scale use of coconut GAC Began dividing up zones by contaminants present; not fixed geographic zones 2015: Include TCN # for bituminous GAC in specs.? Leave I2 # for coconut GAC in specs.? Approval for use of Reactivated GAC? Under Regulatory Review Specifics: Pooled use required; WQ Sampling as per DOH Potential 30% cost savings over virgin GAC