Groundwater arsenic filter based on composite iron matrix

Groundwater arsenic filter based on composite iron matrix: Performance and large scale deployment Abul Hussam Department of Chemistry & Biochemistry George Mason University, Fairfax, Virginia, USA Abul K. M. Munir Environment Initiative Sono Diagnostics Inc. Kushtia, BANGLADESH 1

OUTLINE • Occurrence and Toxicity • Groundwater Arsenic Distribution and Speciation • Filter Development and Measurements • Filter Quality Comparison by ETVAM • Filter Distribution and Future Outlook 2

Tube-wells in Bangladesh 4000 ppb As (Total) Bheramara, Kushtia 170 ppb As(Total) Sadar Kushtia Arsenic free well Sadar, Kushtia 3

Distribution of Inorganic Arsenic Species Based on p. K H 2 As. O 4 - HAs. O 42 - H 3 As. O 3 HAs. O 42 - H 2 As. O 4 H 2 As. O 3 - Left: Theoretical distribution of H 3 As. O 4 species Right: As(III) and As(V) distribution based on 50 ug each 4

Sono Water Filter- Model 45 -25 (25 L Unit) Composite Fe-Matrix 5

Simple and Effective Arsenic Filter Based on Fe-Composite SONO BUCKET FILTERS Flow: 20 -50 Liters per hour As(Total) < 10 ppb (CL 95%) As(III) < 2 ppb (CL 99. 9%) Life: 5 Years Minimum Maintenance: Very low Cost: US $35. 00 - $40. 00 Waste: Completely SONO TWIN FILTER SONO BUCKET FILTER nontoxic Arsenic filters were developed by a team led by Prof. Hussam and Dr. Munir of Sono/MSUK of Kus Government approved 6 and

RAPID ASSESSMENT: BAMWSP/DFID/WATER AID Phase II Report, March 2001, WS Atkins BDL-13 -28 ppb ; 34 L/day; Tk 300 3 -38 -220 ppb; 201 L/day; Tk 12000 Sono 3 -K CIM T – Ion Exchange, USA 7

Active Material: Composite Iron Matrix (CIM) ORG CAST IRON PROC CAST IRON CIM 8

Active Material Composite Iron Matrix (CIM) · Uniform, porous, lighter, and less fines · Preformed Hydrous Ferric Oxide (HFO) · Active-surface with large surface area · Insitu continuous formation of HFO during filtration · Continuously increased capacity · Removes many other toxic inorganic and organics (lit evidence) · Extremely low disposal hazard · Indigenous, inexpensive ($35. 00 -45. 00 ), and long lasting(>5 yrs) Charcoal: Cooking Wood · Removes organic compounds - Pesticide residues Inactive Materials Sand: Processed River Sand Brick Chips · Flow stabilizer · Disperser · Coarse particulate filter · Mechanical stability 9

SONO Production Facility: Kushtia 10

Arsenic Removal Performance of Four Filters 1 2 3 4 Blue: As in Tubewell water, Pink: Filtered water. Right graph shows progres decrease in arsenic. Initial [Fe] total = 6. 46 mg/L (#4) 11

Performance of Blank Filters • • Figure shows blank filters maximum contaminant limit (MCL) breakthrough occurs the first day even at 88 L. Groundwater composition: As(III): 300 ug/L, As(total) 996 ug/L, and Fe(II) 1. 0 mg/L. The filtered water properties: Sand: Temperature 27. 3 C, p. H 7. 6 0. 1, TDS 210 6 (us/cm), Eh 158 6 m. V vs. NHE. Sand+Brick+Charcoal: Temperature 25. 7 C, p. H 7. 9 0. 1, TDS 208 12 (us/cm), Eh 148 6 m. V vs. NHE. Filtered water has no detectable total Fe (III + II). 12

Authors Home Court Para, Kushtia Years in use: Total water filter (L): 100, 000 As(T) –influent (ppb): Village Home Khordosadhua, Koksha, Kushtia 4. 5 3. 2 125, 000 L 155 ± 7 600 ± 6 As(T) – filtered (ppb): 7 ± 1 8± 1 Fe(T) – influent (ppm): Fe(T) – filtered (ppm): Cost (Taka)/L: 5. 09 ± 4. 85 ± 0. 25 0. 228 ± 0. 04 0. 016 0. 245 ± 0. 0513 0. 024

SONO Filter Performance in the Fiel Parameters Fatic Alampur Kaliskhnp ur Juniadah 2. 32 67, 760 3. 6 104, 960 4. 4 128, 480 2. 5 72, 960 10 14 56 8 As(Total)Input, ppb 32 ± 7 410 ± 15 1139 ± 89 2423 ± 87 As(Total), Filter, ppb <2 8± 2 7± 2 8± 4 Fe(Total) Input, ppm 20. 7± 0. 6 10. 86 ± 0. 56 1. 53 ± 0. 08 0. 6 ± 0. 03 Fe(Total) Filter, ppm 0. 22 ± 0. 02 0. 242 ± 0. 03 0. 25 ± 0. 05 0. 26 ± 0. 03 Years Water Yield (L) Number of Measuremen ts 14

Four Approved Filters: Environmental Technology Verification Program Bangladesh Council of Scientific Research, March 4, 2004 READ-F Bangladesh ALCAN SONO SIDCO Parameter Standard Alumina Comp Fe-Mtrx Cerium IE As(Total), ug/L 50 58 12 (57 24)a 17 15 (6 1)a 16 17 (7 5)a 36 32 (22 4)a As(III), ug/L na 35 12 (36 19)a 8 7 (4 1)a 6 5 (5 2)a 13 16 (6 4)a Fe (total), mg/L 0. 3 (1. 0) 0. 6 0. 2 0. 06 0. 04 0. 16 0. 21 0. 2 0. 1 Phosphate, mg/L 6 0. 5 0. 2 0. 9 0. 6 1. 1 1. 5 100 68 77 12 2 No data 11 2 18 6 No data Sulfate, mg/L Silicate, Si. O 2, mg/L Micro Fe. O Al, mg/L 0. 1 (0. 2) 0. 17 0. 07 0. 11 0. 02 No data Ca, mg/L 75 (200) No data 104 18 No data Mn, mg/L 0. 1 (0. 5) 1. 4 0. 22 0. 12 No data 0. 5 0. 4 No data 50 17 No data 7. 3 0. 17 7. 6 0. 1 7. 4 0. 4 7. 4 0. 3 128 7 17 3 61 6 264 37 Mg, mg/L p. H Flow rate (L/h) 6. 5 – 8. 5 15

ETVAM Manganese (Mn) Data Summary (BCSIR/OCETA Test), March 04, 2004 (Bangladesh limit is 0. 500 mg/L) Filter Technology ALCAN SIDCO SONO READ-F Region Influent Mn mg/L Effluent Mn mg/L Bera 4. 51 0. 26 (n=6) 4. 21 0. 21 (n=12) Hajiganj 0. 09 0. 06 (n=7) 0. 17 0. 29 (n= 13) Manikganj 0. 51 0. 10 (n=6) 0. 44 0. 11 (n=11) Nawabganj 2. 47 0. 24 (n=10) 1. 99 0. 55 (n= 20) Faridpur 0. 15 0. 05 (n=9) 0. 17 0. 03 (n= 14) Bera 1. 07 0. 11 (n= 7) 0. 58 0. 12 (n=15) Hajiganj 0. 07 0. 04 (n=10) 0. 06 0. 03 ( n=19) Manikganj 0. 37 0. 04 (n=6) 0. 33 0. 02 (n= 12) Nawabganj 1. 50 0. 07 (n=7) 1. 06 0. 41 (n=17) Faridpur 0. 60 0. 04 (n=6) 0. 51 0. 07 (n=9) Bera 0. 64 0. 22 (n=10) 0. 09 0. 07 (n=18) Hajiganj 0. 04 0. 03 (n=7) 0. 3 0. 19 (n=14) Manikganj 0. 72 0. 06 (n=12) 0. 26 0. 19 (n=18) Nawabganj 1. 93 0. 21 (n=10) 0. 28 0. 16 (n=20) Faridpur 3 0. 22 0. 07 (n=13) 0. 11 (n=18) NO Data on Mn Analysis 1. denotes one standard deviation; 2. “n” is the number of observations 16 3. In Faridpur (Bhanga) one effluent Mn concentration was 60. 02 mg/L (in day 11) probably this is a wrong entry, we omit this figure in our calculation.

Possible physicochemical reactions in different parts of the filtration process. CIM- Composite Iron Matrix Top layer: Oxidation of soluble iron Oxidation of ferrous iron Fe(II) + O 2. - + Fe(III)OH 2+ Fe(II) + O 2. - Fe(III) + H 2 O 2 Fe(II) + CO 3. - Fe(III) + HCO 3 - CIM hydrous ferric oxide =Fe. OH + H As. O - =Fe. HAs. O - + H O ( 2 4 4 2 (HFO) 24 Fe(III) complexation and precipitation. K = 10 ) =Fe. OH + HAs. O 42 - =Fe. As. O 42 - + H 2 O ( K = 1029 ). 17

Mechanisms of Arsenic Removal · Corrosion of Fe. O-composite continuously generates hydrated iron oxides with high specific surface area. · Mn (ca. 0. 2%) in the Fe. O-composite acts as a catalyst for rapid conversion As(III) to As(V) · Arsenate is removed by surface complexation with HFO (possibly magnetite and maghemite ( -Fe 2 O 3)). · Removal process is independent of the concentration of As(III) + As(V) input concentration. · Excess Ca 2+, Fe(II)/Fe(III) and other divalent cations enhance adsorption and complex formation through doublelayer charging. · Further, cementation reactions produce a porous high surface area insoluble spent material very similar to natural HFO with a high capacity for arsenic removal. 18

Two Filters: Alumina vs. CIM Influent As(T): 600 -1240 ug/L Influent As(T): 400 -600 ug/L (Nawabgong Data: Arsenic Removal Technology: Environmental Technology Verification Program for Arsenic Mitigation, Bangladesh Council of Scientific and Industrial Research, Dhaka, Bangladesh. 2003 -2004). 19

Filter Problems and Solution Problem Decreased flow < 10 L/hour Cause Clogging of top layer sand by HFO –both buckets Solution Wash or replaces topmost sand layers Increased Poor hygiene or Pour 5 L hot bacterial poor maintenance water every 15 days. contaminati on Visibly Nonuse for a turbid water while (or free flow junction malfunction) Filter and discard first 3 -4 buckets (or return the filter) 20

Drinking water inorganic quality parameters: Comparison of 3 Kolshi and Sono bucket filter with international and national standards. Results include independent test results by WS Atkins. 3 -Kolshi was approved by the Constituent Bangladesh Arsenic Mitigation Water Supply Project, BG for household use. USEPA WHO, Bangladesh 3 -Kolshi Sono Bucket 0. 05 Waterb 0. 003 -0. 018 Filter Waterb 0. 003 -0. 020 0. 3 (1. 0) 0. 08 - 0. 49 0. 2 0. 02 6. 5 -8. 5 7. 74 0. 1 7. 7 0. 1 Sodium - mg/L 200 26. 7 3. 1 19 - 25 Calcium - mg/L 75 (200) 59. 1 7. 5 5 - 87 Copper - mg/L 1. 3 1. 0 - 2. 0 1. 5 0. 005 <0. 002 Manganese - mg/L 0. 05 0. 1 - 0. 5 0. 1 (0. 5) <0. 001 <0. 01 5 3. 0 5 (15) 0. 01 <0. 007 0. 05 -0. 2 0. 1(0. 2) 0. 03 <0. 022 0. 015 0. 01 0. 10 0. 006 <0. 004 Chromium, mg/L 0. 1 0. 05 <0. 002 Cadmium, mg/L 0. 005 0. 003 0. 01 <0. 001 2. 0 0. 7 1. 0 0. 063 0. 01 <0. 082 0. 006 0. 005 <0. 013 0. 07 0. 003 <0. 002 Nickel, mg/L 0. 1 0. 02 <0. 002 Selenium, mg/L 0. 05 0. 01 <0. 012 <0. 002 Silver, mg/L 0. 1 <0. 002 Chloride, mg/L 250 200 (600) 10 - 40 2. 0 - 12. 2 Arsenic (total)mg/L Iron (total) - mg/L p. H Zinc - mg/L Aluminum -mg/L Lead -mg/L Barium, mg/L Antimony, mg/L Molybdenum, mg/L (MCL) 0. 01 Guideline 0. 01 0. 3 21

USAGE INFORMATION (2001 -July 2005) • 20, 000 SONOTM filters installed in 14 districts • 60 -180 liters per day consumed for 2 - 4 years • 250, 000 direct beneficiaries • >1 Billion liter of clean water consumed Tube well-head filter • Cost: <0. 0043 Taka / Liter 22

SONO Dual Filters 80 L/ hour 2423 to 8 ppb As(T) 23

SONO Triple Filters Flow: 120 L/hour 24

SONOTM Arsenic Filter at Work : 4000 ppb to 7 ppb As after SONO Filtered 4000 ppb As SONO by SDC/MSUK, Kushtia, Bangladesh Jagassar, Bheramara, Bangladesh 25

PEOPLE Abul K. M. Munir and Zafreen A. Hossain INVOLVED MSUK and SDCI, Kushtia, Bangladesh Abul Barkat Department of Economics, University of Dhaka, Bangladesh Amir H. Khan Department of Chemistry, University of Dhaka, Bangladesh Mohammed Alauddin Chemistry Department, Wagner College, NY, USA Mohammad Habibuddowla Department of Environment, Commonwealth of Kentucky, USA MSUK-SDCI Kushtia Laboratory Dr. Tahmina Begum, Monwar Ahmed, Mir Morshedur Rahman, Mizanur Rahman, Aslam Hossain, Ziaul Haq, Abu Sayeed, and many others. 26

Team Members 27

Organizations Involved • Sono Diagnostic Center EI, Kushtia • MSUK: Manob Sakti Unnayan Kendro, Kushtia • HDRC: Human Development Research Center, Dhaka, Bangladesh • Chemistry Department, George Mason University, USA • Quasem-Nahar Trust Fund, Kushtia 28

Students carrying filtered water home 29

Leaching experiment (EPA-TALP) with spent materials after a year Element Conc. (mg/L) Al at p. H =7 (Sand) 0. 028 at p. H=4 (Sand) 0. 02 at p. H =7 (Sand + Iron) 0. 049 at p. H =4 (Sand + Iron) 0. 014 at p. H =7 (Sand + Iron) Rainwater 0. 02 As <0. 016 0. 025 <0. 016 Ag <0. 002 Ba 0. 002 0. 007 0. 006 0. 009 0. 008 Be <0. 001 Ca 2. 65 8. 39 4. 89 8. 13 4. 00 Co <0. 002 Cr <0. 002 Cu 0. 005 0. 018 0. 007 0. 008 0. 002 Fe 0. 02 0. 031 0. 041 0. 051 0. 016 Mg 0. 2 1. 03 0. 519 0. 831 0. 52 Mn 0. 013 0. 081 0. 005 0. 047 0. 002 Mo <0. 003 Ni 0. 002 0. 009 0. 003 0. 005 0. 004 Pb <0. 004 K 0. 82 1. 26 0. 608 0. 527 0. 62 Na 2. 64 0. 750 3. 35 1. 85 7. 2 Se <0. 012 Sr 0. 006 0. 017 0. 014 0. 02 0. 012 Sn 0. 007 0. 011 0. 003 0. 006 0. 003 TI <0. 067 V <0. 001 Zn 0. 011 0. 038 0. 013 0. 021 0. 024 30