Improving Access to Safe Water in West Bengal, India: From Arsenic and Bacteria Removal to Household Behavior Change

Millions of people in rural West Bengal, India, are exposed to groundwater containing toxic concentrations of arsenic, unpleasant levels of iron and non-negligible fecal contamination. Before publicly-provided piped water becomes widely available, a promising approach to address this unacceptable situation is to treat groundwater in small decentralized plants that sustain themselves by selling water to households at an affordable price. This approach hinges on two hypotheses: first, that groundwater can be treated for all contaminants –arsenic, iron, and microorganisms– at very low-cost and with locally available materials and labor; second, that a large fraction of households –ideally all households– will purchase treated water. This dissertation examines these two hypotheses in detail. First, it analyzes the concurrent removal of arsenic and bacteria from contaminated groundwater using iron electrocoagulation (Fe-EC), a simple and robust process suitable for operation in resource-limited settings. Second, it investigates household water practices and drivers of behavior change in rural West Bengal.Fe-EC relies on the dissolution of an Fe(0) anode to produce strong oxidants and Fe(III) precipitates with a high sorption affinity for arsenic. Building on previous work, a computational model is used to investigate the combined effects of pH and operating conditions (Fe dosage rate and O2 recharge rate) on arsenic removal by Fe-EC. A relationship is established between the impact of operating conditions and the process limiting arsenic removal (As(III) oxidation versus As(V) adsorption), which depends on the pH and the O2 concentration. The robustness of the trends predicted by the model, which operates at constant pH, is evaluated against lab experiments reproducing more realistic conditions where pH increases during treatment –as a result of groundwater equilibrating with atmospheric CO2. The results provide a nuanced understanding of the levers that an operator might use to optimize Fe-EC performance in a range of groundwaters under economic constraints.Bacteria attenuation by Fe-EC is investigated in detail using a series of laboratory experiments. After showing that Fe-EC can attenuate bacteria in synthetic Bengal groundwater without detriment to arsenic remediation, the impact of operating parameters (Fe dosage and dosage rate), groundwater composition (pH, HCO3-, Ca, Mg, Si, P, and natural organic matter (NOM)), and bacteria type are systematically investigated, with a focus on elucidating the mechanisms of bacteria attenuation. The results show that attenuation is primarily due to bacteria encapsulation in Fe(III) flocs and removal by gravitational settling, while inactivation by germicidal reactive oxidants remains limited in the presence of HCO3- and at pH>7. Fe(III) precipitates are found to adhere to the surface of bacterial cells, primarily through interactions with bacterial phosphate groups, resulting in bacteria enmeshment in precipitate flocs. The effect of major groundwater ions is interpreted in light of this mechanism: Ca and Mg reduce attenuation by complexing bacterial phosphate groups; Si and NOM, which do not strongly compete with phosphate groups for sorption to Fe(III) precipitates, do not affect attenuation; by contrast, P decreases attenuation significantly, except in the presence of bivalent cations, which can bridge between P sorbed to precipitates and bacterial phosphate groups. Finally, Fe-EC is shown to be equally effective towards Gram-positive and Gram-negative bacteria, smooth and rough alike, likely due to the universal presence of phosphate moieties on bacterial cell walls. Altogether, results show that Fe-EC can effectively remove all types of bacterial contamination from a range of groundwaters.Results of a 501-household survey about water practices in the arsenic-affected district of Murshidabad, in West Bengal, India, are presented. 53% of the surveyed population was found to use alternatives to shallow groundwater including domestic filters, water purchased from small private entrepreneurs, government tubewells, and municipal piped water. The analysis shows that following socioeconomic status, risk perception of gastric illness and dissatisfaction with iron are the primary predictors of the use of alternatives, prevailing over arsenic risk perception. This finding indicates that households react to readily noticeable water problems more than to an invisible contaminant with long-term effects. The factors affecting the choice amongst available alternatives are investigated. The results show that purchased water does not currently provide universal access and that it only reduces, but does not eliminate, arsenic ingestion. Overall, the findings suggest that the provision of treated water through small independent entrepreneurs can be an interim, but partial, solution to the arsenic crisis until piped water becomes widespread in rural West Bengal