Our findings have implications for future research and interventions that may improve water security in urban slums, especially in India where an estimated 50% of slums are non-notified communities that are frequently excluded from municipal water supplies. With regard to water quality, bacterial contamination occurred at two key points: in point-of-source water from KB pipelines during the monsoon season and in household level stored drinking water during all seasons. Household level water contamination was especially notable in the winter and summer, given the lack of point-of-source water contamination in those seasons. Even in the monsoon, when half of point-of-source water from motors was contaminated, most contaminated drinking water samples were from households that had received clean source water, suggesting superimposed household level contamination.
Our finding on the importance of household level contamination is similar to results of studies from other urban slums [15, 16, 20]. In KB, the vast majority of homes store drinking water in wide-mouthed containers that allow contamination when water is accessed with hands or vessels. Some studies suggest that encouraging safe storage of water in narrow-mouthed containers that minimize hand contact along with household water treatment (e.g., chlorination) may reduce diarrhea rates [21–24]. While a meta-analysis suggests significant benefits from these combined interventions , concerns remain regarding their scalability and possible overestimation of benefits due to publication bias . If further evidence confirms that these interventions are viable and cost-effective, our data suggest they may be beneficial in urban slums.
Our data also highlight seasonal variations in contamination, with summer and monsoon having higher contamination rates than winter. The higher contamination rate in monsoon is partly attributable to point-of-source water contamination, though augmented household level contamination from ambient fecal matter due to flooding of homes is likely another major contributing factor. The higher contamination rate in summer may be due to frequent accessing of water containers due to heat-related body fluid losses.
The finding of seasonal variation in contamination rates may have practical implications for interventions. Given higher rates of vector and water-borne diseases in the monsoon (such as malaria, dengue, and leptospirosis), many municipal health departments in India, including Mumbai’s, dedicate resources in this season for public education campaigns. Public education on household water treatment could be paired with these larger campaigns in a cost-effective manner. Seasonal chlorination may be especially beneficial given the presence of point-of-source contamination during the monsoon, which is not addressable solely by switching to safe storage containers. Some studies question the sustainability of household chlorination of water, since changes in taste may decrease long-term adherence . Encouraging temporary chlorination during seasons when risk is the highest may therefore be a more feasible approach.
While this study highlights household level issues, it also highlights major structural problems with KB’s core water supply that are probably more detrimental to health and social equity outcomes, given their impact on water quality, quantity and cost. Contamination of point-of-source water samples during the monsoon season is a case in point. Such point-of-source contamination likely reflects adulteration that occurs due to backflow of dirty water into the corroded fire brigade pipes that are the community’s primary water source. Since water is only provided to KB’s area for a couple of hours a day, there is no water pressure in these fire brigade pipes for most of the day. The absence of pressure in these pipes leaves them vulnerable to backflow from adjacent leaking sewer lines or from above-ground flooding, which is common during the monsoon season. One study suggests that point-of-source water contamination may be more detrimental to health than household-level contamination, since contaminated point-of-source water introduces new pathogens against which people are unlikely to have immunity .
In addition, the failure of the informal water distribution system to provide an adequate per capita quantity of water is another major structural problem. Use of an inadequate quantity of water may be as detrimental to health as poor water quality . A considerable proportion of homes use less than 20 l/c/d, a consumption level associated with high health risk, since hygiene maintenance becomes difficult . The estimated mean water consumption of 23–32 l/c/d in KB also highlights massive inequality in access, as city-wide data estimate the average water consumption in Mumbai to be 191 l/c/d .
Cost of water was a significant predictor of inadequate water consumption, which is not surprising given the extremely high cost of water in KB compared to the standard municipal rate. In addition, delivery of an inadequate volume of water for the population’s needs, the absence of community taps, limited water delivery timings, and long lines at water collection points may explain the low levels of water use. Periodic failures of the informal distribution system (as captured by our data collected during “system failure”) also introduce huge variability and escalation in the cost of water, as well as being a major cause of chronic stress and emotional distress .
Since point-of-source quality and quantity of water used are functions of these structural issues, these problems can only be remedied by providing equitable access to the municipal water supply, including a new pipeline, public water taps, and improved water provision timings. Furthermore, most remedies may provide only partial benefit unless piped water is provided directly to individual homes, as this is the only intervention that would obviate the need for in-home storage of water. Indeed, data suggest that provision of direct water connections to individual households decreases diarrheal morbidity much more than improving water quantity and access through public water taps alone .
The current informal distribution system exacts a massive economic toll: KB’s residents spend 8.4% of their yearly income on water, and the excess amount paid for water by the entire community could pay for comprehensive water infrastructure in the slum five times over every year. In addition to saving money for residents by lowering water costs, new water infrastructure would generate revenue for the municipality, as most residents are willing to pay for improved water access. Citywide data show that Mumbai’s operating ratio (overall operating expenses divided by revenue generated) is 0.49, suggesting that the municipality generates twice as much revenue supplying water as it spends on operating costs .
In summary, extending formal water infrastructure to slums such as KB would be a “win-win” situation for everyone. It would improve quality of life, reduce water costs, and improve health outcomes for slum dwellers while also decreasing waterborne disease burden in municipal hospitals, generating revenue for the municipality, and decreasing conflicts between the government and slum residents (i.e., by averting government raids on water motors). If widely applied, a policy of extending water access to non-notified slums would also facilitate meeting the Millennium Development Goal for water in urban India .
One limitation of this study is the relatively small sample of households chosen for the seasonal assessment of per capita water consumption and water quality. Future community-based studies with larger sample sizes may help to reinforce our findings, especially with regard to seasonal variations in water quality. In addition, our data on the economic costs borne by community residents only reflect the high cost of obtaining water under informal circumstances. The lack of an adequate quantity of uncontaminated, potable water likely exacts additional costs by greatly increasing the burden of diarrheal, upper respiratory, helminthic, and skin diseases. These diseases may take an economic toll on households by contributing to lost days of work, increased spending on medications, increased health care provider visits, and decreased productivity . Indeed, while we have already highlighted a substantial economic toll on the community secondary to the informal water distribution system, if anything, our calculations are likely to be an underestimate of the overall costs.