MIT Clean Water 4 All, Inc. Final Master of Engineering Group - PowerPoint PPT Presentation

MIT Clean Water 4 All, Inc. Final Master of Engineering Group Presentation – Ghana Team May 30 th , 2008 Cash Fitzpatrick Vanessa Green Izumi Kikkawa Tamar Losleben Andrew Swanton

MIT Clean Water 4 All, Inc. Presentation Outline Ghana: Background and Logistics � Horizontal Roughing Filtration: Tamar Losleben � Household Filtration (Biosand Filter) : Izumi Kikkawa � Chlorine Products: Cash Fitzpatrick � HWTS Consumer Choice Study: Vanessa Green � Ceramic Pot (Kosim) filter + Chlorine Disinfection with Aquatabs: Andrew Swanton �

MIT Clean Water 4 All, Inc. Background Percentage of Households by region (Drinking w ater as biggest problem) 25.0 20.0 15.0 % 10.0 5.0 0.0 a n i t t t s t o r s a l n e a o f r a e a V h E c W h h c t s r A r A o A e r e N p g r e p n p t p o U a U r e B r G (National Statistical Services Survey -CWIQ 2003) Local Perception: Lack of Large Percentage of Water Clean Drinking Water is a Source is Dugouts Major Problem

MIT Clean Water 4 All, Inc. Dugouts Dungu Dam St. Mary’s Dam E-Coli, Total Coliform, and Turbidity of Raw Water Samples from Selected Dugouts During the Rainy Season in Tamale and Savelugu Districts Source: Foran, 2007

MIT Clean Water 4 All, Inc. Presentation Outline Ghana: Background and Logistics � Horizontal Roughing Filtration: Tamar Losleben � Household Filtration: Izumi Kikkawa � Chlorine Products: Cash Fitzpatrick � HWTS Consumer Choice Study: Vanessa Green � Ceramic Pot (Kosim) filter + Chlorine Disinfection with Aquatabs: Andrew Swanton �

Pilot Study of Horizontal Roughing Filtration in Northern Ghana as a Pretreatment Method for Highly Turbid Water Tamar Rachelle Losleben

MIT Clean Water 4 All, Inc. Objectives • Characterize dugout particle sizes and distribution – Turbidity, settling stability, filtrability, sequential filtration, solids settleability • Pilot test horizontal roughing filter (HRF) – Particle size characterization, turbidity, flow rate, microbial contamination

MIT Clean Water 4 All, Inc. Photo Credit: Ghanasco Dam Murcott 08

MIT Clean Water 4 All, Inc. Gbrumani Dam

MIT Clean Water 4 All, Inc. Kpanvo Dam Photo Credit: Doyle 07

MIT Clean Water 4 All, Inc. Kunyevilla Dam

MIT Clean Water 4 All, Inc. Settling Test of 4 Dam Waters 200 17-Jan 6:10 PM 180 Ghanasco Dam 160 21-Jan 12:25 PM Kunyevilla Dam 140 Turbidity (NTU) 120 17-Jan 10:30 AM Kpanvo Dam 100 80 21-Jan 12:25 PM Gbrumani Dam 60 21-Jan 12:25 PM 40 Gbrumani Dam Hand 20 Pump 0 0 5 0 0 0 0 0 0 0 0 0 1 2 6 9 2 4 8 4 2 0 4 9 0 1 2 4 1 1 3 Time (min)

MIT Clean Water 4 All, Inc. Slow sand ? filtration (SSF) Dugout Pretreatment Maximum raw water turbidity: 99-99.99% (Wegelin, 1996; Galvis 1993) removal of microorganisms 20-50 NTU (Wegelin, 1996) Raw Dugout Samples in Tamale and Savelugu Districts (Foran, 2007) Dry Season Rainy Season 779 438 Average E.Coli (CFU/100 mL) 26,357 12,797 Average Total Coliform (CFU/100 mL) Average Turbidity 248 NTU 931 NTU

MIT Clean Water 4 All, Inc. Horizontal Roughing Filters (HRF) Mafi Kumasi HRF Photo Credit: Christian- Chirifoyili HRF Murtie 07 Particle removal mechanisms in HRF (Wegelin, 1996)

MIT Clean Water 4 All, Inc. Ghanasco Dam Pilot HRF G G granite gravel D local gravel P Photo Credit: Carl Allen broken pottery 08

MIT Clean Water 4 All, Inc. Ghanasco Dam Pilot HRF 91.5 cm 700 L 1 inch PVC pipe 95 Polytank 4 inch PVC cm elbow 1 inch valve (brass gate or PVC ball) 54 3.5 m 2.5 m 1.0 m cm 12 –18 mm 4–8 mm 8 –12 mm 1 inch PVC elbow 4 inch PVC Effluent flows to 4 inch to 1 inch PVC pipe soak-away reducer Cinderblocks and drainage adobe bricks

MIT Clean Water 4 All, Inc. Pilot HRF Settling Test 22-Jan 5:26 PM G Granite Gravel 24-Jan G Granite 250 Gravel 22-Jan 5:26 PM D A Local Gravel 24-Jan D Local 200 Gravel 22-Jan 5:26 PM P Broken Pottery 24-Jan P Broken Turbidity (NTU) Pottery 150 B 16-Jan 5:10 PM G tank 22-Jan 5:26 PM G tank no mix 100 24-Jan G tank no mix 22-Jan 5:26 PM G tank mixed 50 24-Jan G tank mixed E C 16-Jan 5:10 PM P tank 24-Jan P tank no 0 mix 0 15 20 60 90 120 240 480 1440 1920 3000 24-Jan P tank mixed Time (min)

MIT Clean Water 4 All, Inc. Comparison of the Turbidity Reduction Performance of HRF Media Average Average % Average Average additional additional Average Filtration HRF filtration turbidity turbidity % total coefficient , λ effluent rates removed by removed HRF turbidity (ml/min) (min^-1) HRF after by HRF turbidity settling after reduction settling 51 NTU 220 46 TU G 61 % 84 % 0.002 (1.6 m/hr) granite gravel 72 NTU 170 30 TU D local 47 % 76 % 0.0007 (1.3 m/hr) gravel 61 NTU 200 18 TU 55 % P 80 % 0.0006 (1.5 m/hr) broken pottery < 50 NTU 41-270 --- Goal: --- --- --- (0.3-2.0 m/h)

MIT Clean Water 4 All, Inc. Comparing Pilot Ghanasco HRF Filtrability to Mafi Kumasi HRF Filtrability 250 AVERAGE G Granite Gravel Filtered Water Volume (ml) AVERAGE D Local 200 Gravel AVERAGE P Broken Pottery 150 AVERAGE Mix and Settled G tank 100 AVERAGE Mix and Settled P tank 1-Feb Mafi HRF Inflow 50 1-Feb Mafi HRF Outflow 0 1 min 2 min 3 min Filtration Time (min)

Ouagadougou Pilot HRF International Institute for Water and Environmental Engineering Burkina Faso • June 5 - July 28, 2006 • Loumbila Dam (Sylvain, 2006) http://aochycos.ird.ne/HTMLF/ETUDES/HYDRO/LOUMBILA.HTM

MIT Clean Water 4 All, Inc. Comparison of Pilot HRF Performance Blue Nile Ghanasco Dam, Tamale, Ouagadougou Health Project, Northern Ghana , Burkina Sudan Faso (referenced by (Losleben, 2008) Wegelin, 1996) (Sylvain, Media gravel quartz gravel broken granite local broken 1989) gravel burnt gravel pottery bricks G D P Average filtration rate 0.30 1.6 1.3 1.5 1.0 (m/h) Filter length and media 270 cm, 30-50 350 cm, 12-18 400 cm, 15- size (mm) 85 cm, 15-20 250 cm, 8-12 25 85 cm, 5-10 100 cm, 4-8 150 cm, 5-15 Raw water turbidity 40-500 NTU 313 NTU 301 301 NTU 5-50 NTU NTU Prefiltered water turbidity 5-50 NTU 51 NTU 61 NTU 4-19 NTU 72 NTU Faecal coliforms* (/100ml) Raw water > 300 --- 8400 8400 8400 --- Prefiltered water < 25 --- --- 15500 500 --- Mean turbidity reduction 77 % 87 % 84 % 76 % 80 % 32 % * as E.coli

MIT Clean Water 4 All, Inc. HRF Channel Design 1 m Granite 1 m Gravel 44 m 2 m 2.6 m Kunyevilla Channel Total Channel Length 45 m Q = 75,000 L/day Raw λ = 0.13 hr^-1 Dugout Slow Water Sand 700 NTU Filter q = 1.6 m/h 20 NTU 22.5 m 6.4 m 16.1 m (Wegelin, 1996)

MIT Clean Water 4 All, Inc. Presentation Outline Ghana: Background and Logistics � Horizontal Roughing Filtration: Tamar Losleben � Household Filtration (Biosand Filter) : Izumi Kikkawa � Chlorine Products: Cash Fitzpatrick � HWTS Consumer Choice Study: Vanessa Green � Ceramic Pot (Kosim) filter + Chlorine Disinfection with Aquatabs: Andrew Swanton �

MIT Clean Water 4 All, Inc. Background ~Biosand Filter (BSF)~ • Household treatment • Intermittent slow sand filtration • Removes: • >90 % of E.coli bacteria • 100 % of protozoa and helminthes (worms) • 50-90 % of organic and inorganic toxicants • <67 % of iron and manganese • most suspended solids • 270,000 BSFs installed in 25 countries – Disadvantages: Diagram of Biosand Filter – does not suite treatment of high turbid water » Decline in treatment efficiency, frequent clogging and maintenance requirement Turbidity Limit ~50 NTU

MIT Clean Water 4 All, Inc. Local Plastic Design BSF Biolayer: schumutzdecke , biofilm – most purification proceeds here – estimated to be 5-10 cm in depth 1 Modification: Create additional biolayer oxygen diffusion is essential standing water layer should be 5-10 cm 10 cm 5 cm Evaluation: Turbidity 47 cm E.Coli Total Coliform flow rate A, A’ B C 1) B.J.Buzunis, Intermittently Operated Slow Sand Filtration: A New Water Treatment Process , March 1995

MIT Clean Water 4 All, Inc. Results & Discussion -Flow Rate- Design flow rate ~ 20 L/hr 45 Upper limit 30 L/hr 40 Flow Rate [L/hr] Lower limit 5 L/hr 35 BSF A 30 BSF A' 25 BSF B 20 BSF C 15 10 0 4 8 10 12 15 17 19 21 23 25 27 31 33 35 38 42 45 no decline in flow rate Day average flow rate [L/hr] LPD BSF (standard deviation) A (without modification) 32.0 (4.1) A' (without modification) 25.9 (4.9) B (additional 5 cm sand layer) 21.8 (6.0) C (additional 10 cm sand layer) 21.1 (4.3) lower flow rates for BSF B & C No clogging

MIT Clean Water 4 All, Inc. Results & Discussion -Turbidity- 500 400 Turbidity [NTU] Dugout 300 BSF A BSF A' 200 BSF B BSF C 100 0 Variation in operation? 7 9 11 15 17 19 21 23 25 27 31 33 35 42 44 Need for cleaning? Day After day 13 Dugout and BSF Average turbidity [NTU] (standard deviation) Dugout 306 (97) A (without modification) 22 (17) A' (without modification) 20 (14) B (additional 5 cm sand layer) 15 (6.8) C (additional 10 cm sand layer) 14 (1.4)