Globally, 2.1 billion people lack access to safe drinking water at home. While improved water sources exist, challenges remain, including high costs of infrastructure, contamination during distribution, and inconsistent supply. Point-of-use (POU) treatment is an alternative, but reliance on individual household behavior change limits its success. Point-of-collection (POC) treatment, like manual chlorine dispensers, offers a solution, but it still requires user involvement. Automatic in-line chlorinators alleviate this burden, allowing users to collect any volume of chlorinated water without altering their routines. However, most electricity-independent options utilize solid chlorine tablets, which can be unavailable. This study aimed to address these limitations by designing and testing a low-cost, electricity-independent liquid chlorine doser, building on previous work with liquid chlorine dosers in Bangladesh. The device was field-tested at water kiosks in Kisumu, Kenya, to evaluate its technical performance, customer demand for chlorinated water, and kiosk owner willingness and ability to pay.

Existing literature highlights the global challenge of providing safe drinking water, particularly in low-income settings. The limitations of centralized water treatment and distribution due to cost, infrastructure challenges, and recontamination are discussed. Point-of-use (POU) water treatment, while effective in improving water quality and reducing diarrheal disease when compliance is high, suffers from low and inconsistent adoption. Point-of-collection (POC) treatment offers a promising alternative, especially for communities where 26% of the population collects water off-premises. Studies on manual and automatic chlorinators at community water points are reviewed, revealing variable uptake rates for manual dispensers and challenges with existing in-line chlorinators due to reliance on solid chlorine tablets or inconvenient wait times.

The Stanford-MSR Venturi, a novel in-line chlorinator employing the Venturi principle, was designed and tested. This device automatically adds liquid chlorine to the water stream without electricity or moving parts. The design process involved optimizing the Venturi geometry using a test stand to ensure accurate dosing across various flow rates. Issues like non-dosing at low flow rates and bubble formation in the chlorine supply tubing were addressed through design modifications, such as changing tubing material and optimizing valve placement. The Venturi was calibrated in the lab to achieve a target free chlorine residual of 0.5–1.5 mg/L. The field study involved 26 water kiosks in Kisumu, Kenya, which were offered four service packages: lease, lease with chlorine delivery, lease-to-own, and lease-to-own with chlorine delivery. Weekly visits to each kiosk involved collecting water samples to measure chlorine residual using a Hach Pocket Colorimeter II. Kiosk operators provided sales data and feedback. Post-intervention surveys were conducted with customers and kiosk operators/managers to collect information on water purchase behavior, perceptions, and preferences. Data analysis was performed in Stata/SE 14 and R.

Seven of 26 kiosks (27%) enrolled in the study. The Venturi consistently chlorinated water (97.6% of samples had detectable free chlorine residual, with most non-detects attributable to one device with iron deposits). Free chlorine residual ranged from 0.5 to 3.0 mg/L in lab tests and 0.05 to 1.59 mg/L in the field, mostly within the target range (0.2–1.2 mg/L). Six of seven kiosks fulfilled their service package payments, with most purchasing the device after the initial 6-month period. Chlorinated water sales accounted for 19.1% of total sales. Customer surveys revealed that 94.4% purchased water at least twice a week, with 66% buying chlorinated water. Customers reported high satisfaction with the water quality. Kiosk operators were largely satisfied with the device's performance, though some suggested improvements (longer spout, lower price, longer payment periods). While some kiosks subsidized the device through grants, others funded payments solely through water sales. Payment compliance was an issue, with only 60% of payments made on time.

The Venturi's consistent chlorination performance and high demand from kiosk owners demonstrate its potential for widespread adoption in improving safe water access. The successful payment and purchase rates suggest a financially viable model, particularly given the additional health benefits from chlorinated water. While the initial adoption rate was moderate (27%), the high rate of purchase after the 6-month trial period suggests that satisfaction with the device contributed significantly to its sustained adoption. Late payments pose a challenge, suggesting that flexible payment plans might be beneficial. The varying strategies used by kiosks to maximize chlorinated water sales highlight the importance of tailoring approaches to local contexts and preferences.

The Stanford-MSR Venturi demonstrates strong potential for increasing access to safe water in low-income communities. Its robust technical performance, coupled with the observed demand from kiosk owners and consumers, indicates the viability of this technology and the associated service model. Future research should focus on optimizing payment plans, exploring pre-filtration for iron-rich water sources, and conducting longer-term studies on operational sustainability and health impacts.

The relatively small sample size of kiosks (26) limits the generalizability of the findings. The study duration (6 months) might not fully capture long-term device performance or seasonal effects on water sales. Microbial water quality measurements and health outcome data were not collected. The involvement of an NGO in providing services and research support means that the true profitability of the service model for a commercial provider remains unknown.