An Evaluation of Household Drinking Water Treatment Systems in Peru: The Table Filter and the Safe Water System

Transcription

1 An Evaluation of Household Drinking Water Treatment Systems in Peru: The Table Filter and the Safe Water System Brittany Coulbert Spring 2005 Based on Coulbert s MIT M.Eng. thesis of the same title (Feb. 2005)

2 Peru Project Site: La Joya, Arequipa

3 The Peru Team

4 The Water Problem 1.1 billion people lack access to safe drinking water 1.7 million people die each year from diarrheal diseases related to unsafe water, sanitation, and hygiene Fact source: WHO, 2004 Picture source: Susan Murcott, MIT

5 The Problem in Peru 50% of Peru s 28.4 million people live below the poverty line GDP per capita: $5,000 In 2002, Peru had only 66% improved drinking water coverage in rural areas Improved does not mean safe. It may simply indicate a household connection or protected well. (With this definition, this report evaluates almost entirely households with improved access to drinking water, even before treatment.) Source: WHO Joint Monitoring Programme, CIA World Factbook

6 The Solution UN Millennium Development Goal: Halve, by 2015 [as compared to 1990], the proportion of people without sustainable access to safe drinking water and basic sanitation. WHO International Network to Promote Household Water Treatment and Safe Storage This report is part of the effort by MIT and members of the Network to seek ways to increase access to safe water for people worldwide through efforts including household water treatment systems (HWTS) and safe water storage.

7 The Technologies The Table Filter The Safe Water System

8 Safe Water System Outline Tests & results from Peru Table Filter Tests & results from Peru and the MIT lab Household interviews Cost comparison & financing options

9 Study of HWTS Solutions as Currently Implemented in Peru Field Study in southern Peru, mostly in La Joya, Arequipa January 2004 Tested Safe Water Systems (household chlorination) & Table Filters in homes Interviewed users MIT Lab Study February - July 2004 Tested 2 Table Filters with different grades of sand Also tested Table Filters without sand

10 Regions of Study: Arequipa & Tacna (focused in Cerrito Buena Vista, near La Joya, Arequipa) Lima Arequipa Tacna

11 Safe Water System (Household Chlorination)

12 Safe Water System (SWS) 20-L safe storage containers ( bidones ) 0.5% sodium hypochlorite solution generated at a local hospital & distributed to towns in 200-mL bottles Users add half cap of solution to 20 L of water stored in safe storage containers and wait 30 minutes before drinking

13 Chlorine Generator Electricity + Water + Salt = Chlorine solution

14 Family in Peru using their Safe Water System

15 Safe Water System Tests E.coli & Total Coliform (TC) tests These bacteria indicate the presence of fecal contamination, which can cause severe diarrheal sicknesses Zero E.coli or TC colony forming units (CFU) should be present in any drinking water sample Chlorine residual tests Chlorinated water sources should have a residual amount of chlorine to ensure that the chlorine dose was enough to deactivate all harmful bacteria

16 E.coli Concentrations Before & After SWS Chlorination in Peru E.coli CFU/100ml Log Scale 1.E+05 1.E+04 1.E+03 1.E+02 1.E+01 1.E % reduction of E.coli by SWS B-11 E-3 H-12 C - 7 LL - 18 LL - 27 M - 8 I - 1 LL - 22 House ID E.coli Raw E.coli SWS Treated

17 B-11 H-12 LL - 18 M - 8 LL - 22 E -2 L - 16 M - 17 M - 17B LL - 4 L - 10 M - 19B LL - 9 CBV Post B - 2 A' - 7 A' E+06 1.E+05 1.E+04 1.E+03 1.E+02 1.E+01 1.E+00 Total Coliform Concentrations Before & After SWS Chlorination in Peru 95% reduction of TC by SWS TC Raw Avg. TC SWS Treated Avg. House ID TC Raw TC SWS Treated TC CFU/100ml Log Scale

18 Valdivia CBV School A-8 E-3 LL 18 LL - 2 LL - 22 E -2 M - 1 M - 17 H - 10 L - 11 M - 19 LL - 23 N - 3 CBV Kindergarden A' - 2 A' - 5 A' - 5B Free Chlorine Residual Measured in Bidones in Peru (The dark black lines indicate the target concentration range.) The average chlorine residual (0.18 mg/l) was less than the recommended range. Average = 0.18 House ID Free Chlorine [mg/l]

19 Pros & Cons of the SWS Pros 99.6% E.coli removal 95% TC removal Very inexpensive: $6 container + $3/year Easy to use Local chlorine generation possible Cons Less effective with turbid source water (particle removal pre-treatment is needed) Average chlorine residual was found to be too low Chlorine solution sometimes difficult to obtain due to poor technical support People don t like the taste

20 Peruvian Table Filter Indigenous filter developed by CEPIS and the Belgian development organization, DGCI Filtering media: geotextile cloth, sand, & ceramic candle filters Made of two 20-L (5-gal.) plastic buckets

21 Table Filter Media Geotextile Sand Bed Ceramic Candles

22 Family in Peru using their Table Filter

23 Coliform tests Table Filter Tests Thermotolerant Coliform (TTC) Indicates fecal contamination E.coli & TC Heterotrophic Plate Counts (HPC) Indicates general bacteria levels Turbidity tests Flow rate tests

24 Lab in La Joya, Peru

25 E.coli & TC Concentrations Before & After Treatment by Table Filters in Peru Coliform [CFU/100ml] LOG Scale 1.E+05 1.E+04 1.E+03 1.E+02 1.E+01 99% reduction of E.coli by TFs 98% reduction of TC by TFs 1.E+00 H-3 F-7 H-3 (2) TC Raw Avg. TC TF Treated Avg. CBV Post H - 3 E - 9 H - 2 O - 3 H - 1 B - 7 House ID (all in CBV) G - 8 H - 13 B - 5 N - 1 E.coli Raw E.coli TF Treated TC Raw TC TF Treated

26 Summary Table of Coliform Tests on Table Filters and SWS in Peru Table Filters SWS Chlorination Coliform CFU/100ml Raw Water Treated Water % Removal Raw Water Treated Water % Removal E.coli 5.3x % 4.5x % TC 3.5x x % 2.1x x %

27 Comparison of Average Coliform Concentrations Before & After Table Filter & SWS (in Peru) 1.E+05 Table Filter SWS Chlorination 1.E+04 before before CFU/100ml LOG Scale 1.E+03 1.E+02 after after 1.E+01 1.E+00 TF - Raw TF - Treated (Filtered) SWS - Raw SWS - Treated (Chlorinated) E.Coli TC

28 Turbidity Concentrations Before & After Treatment by Table Filter in Peru Turbidity (NTU) % reduction of turbidity by TFs (70% by working TFs) 0 H - 3 F - 7 Villa Hermosa E - 3 CBV Post* H - 3 Input Avg. Output Avg. Filter Input E - 9 H - 2* O - 3 H - 1 B - 7 G - 8 H - 13 B - 5* N - 1 *Filter reported as "not working" House ID Filter Output

29 Table Filter Flow Rate in Peru Average: 3.1 L/hr Arequipa average: 2.3 L/hr Tacna average: 3.8 L/hr Faster than other ceramic filters ( L/hr) Slower than BioSand Filter (15-30 L/hr)

30 CBV (Ramirez) La Cano Post Cruz de Mayo (Belisario) Villa Hermosa C-16 CBV H - 3 CBV E - 9 CBV O - 3 CBV H - 1 CBV B - 7 CBV G - 8 CBV H - 13 CBV N - 1 Chucatamani (Chavez) Chucatamani "E" Chucatamani "F" Pistala (Toma) Table Filter Flow Rates in Peru Tacna Avg: 3.8 L/hr Arequipa Avg: 1.7 L/hr House ID Flow Rate [L/hr]

31 MIT Lab Work -- Table Filter February - July 2004 Simulate conditions in Peru All materials, except sand, brought from Peru Charles River water & sewage added to filters to increase water contamination levels Two filters Medium Sand Table Filter (MSTF): Sand size used for BioSand filter, larger grains than in Peru Fine Sand Table Filter (FSTF): Sand size specified in Peru Coliform tests Turbidity Tests

32 Table Filters in MIT Lab

33 TTC Concentrations Before & After Treatment by the Medium and Fine Sand Table Filters 1.E+05 1.E+04 98% average reduction of TTC by MSTF & FSTF 1.E+03 1.E+02 TTC [CFU/100ml] LOG Scale 1.E+01 1.E+00 Feb 20 Feb 23 Feb 27 March1 March5 March 8 March 12 March 15 March 19 June 1 June 15 June 16 June 18 June 21 June 22 June 23 June 24 MSTF Avg. FSTF Avg. Date Source MSTF FSTF

34 Table Filters Tested Without Sand The two Table Filters showed slightly different results on average. Since sand grain size was the only known difference between the two, the sand was removed from the filters before additional testing to see if the grain size was the true reason for the differences in % removal (as opposed to unexpected differences in the ceramic candles).

35 TTC Concentrations Before & After Filtration without Sand TTC [CFU/100ml] LOG Scale 1.E+04 1.E+03 1.E+02 1.E+01 97% reduction of TTC by MSTF w/o Sand 99% reduction of TTC by FSTF w/o Sand 1.E+00 MSTF Avg. FSTF Avg. June 30 July 1 July 2 Date Source MSTF FSTF

36 E.coli & TC Concentration Before & After Treatment by MSTFs & FSTFs (one day of data: June 24) E.coli & TC [CFU/100ml] LOG Scale 1.E+04 1.E+03 1.E+02 1.E+01 1.E % reduction of E.coli by MSTF 0 E.coli 99.95% reduction of E.coli by FSTF 99.95% reduction of TC by MSTF 3 TC Source Medium Sand Filter Fine Sand Filter 99% reduction of TC by FSTF

37 E.coli Concentration Before & After Treatment by MSTF & FSTF Without Sand E.coli [CFU/100ml] LOG Scale 1.E+04 1.E+03 1.E+02 1.E % reduction of E.coli by MSTF w/o Sand % reduction of E.coli by FSTF w/o Sand 5 1.E June 28 June 30 July 1 July 2 Source MSTF FSTF

38 TC Concentration Before & After Treatment by MSTF & FSTF Without Sand TC [CFU/100ml] LOG Scale 1.E+06 1.E+05 1.E+04 1.E+03 1.E+02 23% reduction of TC by MSTF w/o Sand 99.8% reduction of TC by FSTF w/o Sand 1.E+01 1.E+00 June 28 June 30 July 1 July 2 Source Medium Sand Filter Fine Sand Filter

39 Why Only 23% Reduction of TC by MSTF w/o Sand? The low average % removal of TC by the MSTF was caused by one day of testing which showed that the TC concentration in the treated water was higher than that of the untreated water (June 30). Valid data could not be collected from the water treated by the FSTF, so the two filters could not be compared on that day. Therefore, the average values and % removals of the two TFs without sand are not truly comparable to each other for the TC tests.

40 HPC (Heterotrophic Plate Counts) Concentration Before & After Treatment by Table Filters HPC [CFU/100ml] LOG Scale 1.E+07 1.E+06 1.E+05 1.E+04 1.E+03 1.E+02 1.E+01 1.E+00 99% reduction of HPC by MSTF 66 83% reduction of HPC by FSTF An additional 99.8% reduction from MSTF is needed to reach the average blank water concentration of HPC June 1 June 15 June 23 Source Medium Fine Blank 0 247

41 Summary Table of Coliform Removal Rates of both Table Filters, With and Without Sand, in the MIT Lab Coliform Table Filter Media Source Water MSTF Treated Water FSTF Treated Water MSTF % Removal FSTF % Removal TTC Sand 1.2x x x % 98% No Sand 1.9x x x % 99% E.coli Sand 1.8x10 3 < 1 1.0x % 99.5% No Sand 8.9x10 2 < % 99.8% TC Sand 6.5x x % 99% No Sand 5.0x x x % 99.8% HPC Sand 3.1x x x % 83%

42 Sand Grain Size Theory The coliform tests performed on Table Filters without sand did not support the theory that the sand grain size affected the performance of the filters. Our Peruvian teammates sifting sand for a Table Filter

43 Turbidity Indicates amount of suspended particles in water, or its cloudiness Average % removal of turbidity by Table Filters: Peru: 70% MIT: Medium Sand TF: 91% Fine Sand TF: 92%

44 Pros & Cons of the Table Filter Pros Average % removal of TC in Peru = 98% Average % removal of turbidity at MIT = 92% Provides relatively consistent and significantly improved drinking water Inexpensive: $11.40 each Easy to use Cons Broken spigots Cleaning is bothersome Fragile ceramic candles Parts not easily available Sand sifting (during filter assembly) is timeconsuming More expensive than other treatment methods

45 Interviews 89 households surveyed 35 had Table Filters 49 had Safe Water Systems 5 had neither 66 were from Arequipa (57 from CBV) 23 were from Tacna

46 Interview Statistics Households averaged 5.2 people, including 1.3 children under the age of 5 Average total household* spending was S/ 93 (Peruvian nuevo soles) per month (or $0.58 per person per day) * Of the 84 households with government-sponsored water treatment systems

47 Where do you get water? Where do you get your drinking water? 90 # of Respondents Piped System Water Truck Bring from Arequipa Canal Water Vending Trucks Table Filter Bidon (Chlorine) No System Piped System to Household Taps & Storage Devices Irrigation Canals

48 # of Respondents How often do you clean your How often do you clean your Table Table Filter? Weekly Every 2-6 weeks Every 3-6 months

49 How often do you do add you disinfectant add disinfectant solution to your SWS bidon? solution to your SWS bidon? 20 # of Respondents Twice daily Daily Every 2 days Every 3 days (or every 2-3 days) Every 15 days When did did you last you add disinfectant last add solution disinfectant to your SWS bidon? solution to your SWS bidon? 20 # of Respondents Today Yesterday 2 days ago 3 days ago 1 year ago

50 Does Does the the water taste taste better or better worse now or that worse you now use a treatment system? that you use a treatment system? 80 # of Households Better: Same: Worse: Table Filter Bidon (Chlorine) Do you feel better now that you Do you feel better now that you use a use a treatment system? system? # of Households Yes Sam e No Table Filter Bidon (Chlorine)

51 Do you have any problems/complaints about the treatment system? # of Households Do you have any problems or complaints about the treatment system? Yes No # of Households Is the treatment system easy to use? Is the treatment system easy to use? Yes No Table Filter Bidon (Chlorine) Table Filter Bidon (Chlorine) 100 Do it? Do you like it? # of Households Yes No Table Filter Bidon (Chlorine)

52 Cost Comparison of Each Treatment Option Water Treatment Option Capital Cost O&M Costs / year Total 10- Year Cost Table Filter $6.40 $5 $56 Safe Water System $9.80 $3 $40 Table Filter + SWS $16.20 $8 $96 Water Treatment Plant (w/ piped system, per family of 5) $475 $36 $835 The capital cost of a Table Filter is low because the initial purchase of ceramic candles is included in the O&M costs.

53 # of Respondents Willingness to Pay for a Table Filter by Current Filter Owners (Average Offer: S/25) Average offer: S/ 25 ($7.1) Price Willing to Pay [Soles ] Actual capital cost of Table Filter: S/ 40 ($11.4)

54 Willingness-to-Pay Household surveys indicate that the target population may be willing to pay about half the capital cost of each system How does the program then cover costs?

55 Recommendation for Payment and Financing Options Government (Ministry of Health) and/or outside aid organizations contribute some % of cost or pay for large initial costs (like a $1,400 chlorine generator) Cheaper treatment options are explored or implemented (e.g., chlorine or SODIS) HWTS recipients pay in monthly installments

56 Proposed Monthly Payment Plan for Each Treatment Option (assuming each user must cover 100% of capital and O&M costs) Monthly payment over 12 months Table Filter S/ 3.5 ($1) Safe Water S/ 4 System Table Filter + SWS ($1.1) S/ 7 ($2) Monthly payment over 24 months S/ 2.5 ($0.7) S/ 2.5 ($0.7) S/ 4.5 ($1.3) Approx. monthly O&M after capital is paid off S/ 1.5 ($0.4) S/ 0.9 ($0.25) S/ 2.5 ($0.7)

57 Recommendation: Filtration plus Chlorination is best! Filtration by a Table Filter helps remove the turbidity in water that would make chlorination less effective Chlorination and safe water storage by the Safe Water System kills any bacteria that remain after filtration and protects against recontamination