Gathiru Research III: Sunlight, Water and the Policy Gap

Part III of the foundation series informing Kilimora’s work. A Kenyan team left twelve bottles of Obunga well water in the sun for a day. This brief reads that trial against Kenyan health trials and the record of the African Union, the World Bank, FAO and WFP, and asks what it takes for safe water to reach households and the farms that feed them.

On 21 September 2018, water from four open wells in Kisumu’s Obunga slum sat in clear plastic bottles on the roof of a sub chief’s office. By evening E. coli, the standard marker of faecal contamination, had fallen by 99.99 per cent, a 4 log reduction. Total coliforms fell by only 0.81 log, or 84.64 per cent. The result is in a paper by Marylyn Gathiru, Emilly Obuya, Naumih Noah and Eric Masika, published online on 2 May 2026 in the Journal of Water, Sanitation and Hygiene for Development (doi 10.2166/washdev.2026.257). It is the same Kenyatta University team behind the watermelon rind work in Part II, with partial support from EPA grant SU839297, and Gathiru is Kilimora’s Research Lead. That overlap is a reason to read the paper sceptically, so this brief does.

What the Obunga trial found

The design was simple: one shallow well in each of four zones, three replicate 2 litre PET bottles per well, hourly Petrifilm counts and a UV-A radiometer. Every well was over the WHO limit of zero E. coli per 100 mL before treatment. After six hours the E. coli was gone in all of them. Central, where livestock graze beside the well, started with the most contamination and improved least.

The surprise is the dose. E. coli was fully inactivated only once cumulative UV-A passed 516.13 kJ/m². The Nairobi pre trial run in the same month needed 145.8 kJ/m² on 11 September and 100.76 on 13 September, and the literature range is 90 to 250. Obunga’s figure is 3.5 times the higher Nairobi value and 5.1 times the lower. The authors blame turbidity, suspended solids and dissolved organic matter that dim the UV, and conclude that SODIS rules should be set by measured dose and local water quality, not by hours in the sun.

Reading it honestly

• One day, one roof, twelve bottles, no dark controls. The paper says so itself: heat and UV effects cannot be separated, and weather on that day shaped the result.
• 516 is an upper marker, not a threshold. The dose is ambient UV-A, not in bottle exposure, and with hourly sampling it marks where zero was recorded.
• The log figure rests on small counts. Starting E. coli ranged from 1.67 to 12.33 CFU/100 mL, so a “4 log” reduction is bounded by what Petrifilm can detect. Read it as “none detected”.
• Nairobi versus Obunga is a hypothesis. The authors call the pre trial a methodological calibration, “not a site to site comparison”, yet the abstract and conclusion compare the two. The turbidity explanation was not measured by instrument, only by the newspaper headline test.
• A table that does not reconcile. The four zone reductions in Table 2 average about 94.8 per cent, against 84.64 per cent for total coliforms, so it is unclear which organisms Table 2 covers.

What Kenyan field trials say SODIS delivers

The Obunga result is about water quality. The health case rests on earlier trials. In a 1996 trial among Maasai children, two week prevalence of severe diarrhoea was 48.8 per cent with SODIS against 58.1 per cent in controls, an attributable fraction of 16 per cent (Conroy et al.). A randomised trial in peri urban and rural Nakuru between 2007 and 2009, with 555 children in SODIS households and 534 controls, cut dysentery episodes by 45 per cent and other diarrhoea episodes by 27 per cent, and SODIS children grew 0.8 cm more over a year (du Preez et al., 2011). The same group found in South Africa that dysentery fell 64 per cent among the most motivated households and showed no significant effect at lower motivation (McGuigan, Bristol deck). A cleaner bottle only helps if the bottle goes on the roof.

Why households still need it

WHO and UNICEF’s monitoring programme counted 2.1 billion people without safely managed drinking water in 2024, with sub Saharan Africa furthest behind (JMP data via Our World in Data). Measured at the tap or bottle the picture is worse: a University of Leeds analysis of 37,105 household surveys and 6,395 water samples in rural areas of 14 low and middle income countries found only 7 per cent of households had safely managed water free of contamination at the point of use, and argued the official statistics overstate coverage (Leeds study). Kenya’s own SDG 6 data show 36 per cent of people using safely managed sanitation, 11 per cent of domestic wastewater safely treated and 37 per cent of monitored water bodies in good condition (UN SDG 6 data, Kenya). In Kisumu about 70 per cent of residents are tenants in informal settlements (Practical Action).

The African policy frame, and who pays

The African Union’s water architecture is old, and finance is its weak point. The Africa Water Vision 2025 was prepared in 2000 and costed at US$20 billion a year (AMCOW summary). In 2008 heads of state adopted the Sharm El Sheikh Commitments and in 2015 AMCOW issued the N’gor Declaration on sanitation and hygiene (UNECA review). In 2019 AMCOW adopted the Africa Water Investment Programme, which AU heads of state took up within the continental infrastructure programme (PIDA), and a high level panel put the need at US$64 billion a year (GWP). The first Africa Water Investment Summit met in Cape Town on 13 to 15 August 2025 and called for multilateral finance reform to close the gap (Cape Town Declaration). A post-2025 water security vision was endorsed by ministers in September 2025 and adopted by the AU in February 2026, with a first implementation plan for 2026 to 2033 in preparation and the annual funding gap put at US$35 to 38 billion (GIZ).

What the programmes delivered

Programmes have real outcomes and real shortfalls. On the coast, the World Bank’s Water and Sanitation Development Project had reached 24,000 of a 74,300 person water target and 6,600 of a 130,000 person sewer target by January 2025, with the water target to be met as contracts finish by August 2025 (status report). The US$250 million IDA financed Climate Smart Agriculture Project, in 24 arid and semi arid counties, had reached 230,000 of 278,900 targeted households by its third year (Kenya News Agency). The pattern is that networks deliver slowly and farm support faster, and both leave households waiting.

Kisumu’s own record shows both sides. In Nyalenda and Obunga, household connections rose from 33 to 55 per cent, with 73,257 people benefiting, but 45 per cent of households still had no tap at home. The utility’s KIWASCO pilot in Manyatta connected 102 households after a Sh1.8 million grant subsidised connection fees and first bills, because the fee was the barrier (Kenya News Agency).

Water is a food question

FAO and its four partner agencies report in SOFI 2026 that 645 million people, 7.8 per cent of the world, were hungry in 2025, the third annual fall (UN video library, FAO). Africa now has 309 million, or 20.0 per cent of its people, more than Asia’s 292 million, and 510 to 520 million people could still be hungry worldwide in 2030 (The Reporter). Nearly 67 per cent of Africans cannot afford a healthy diet (Guardian Nigeria). A year earlier SOFI 2025 had Africa at 307 million, up 113 million since 2015, with Eastern Africa steady near 26 per cent (IFPRI food security portal).

In Kenya, the 2026 short rains assessment put 3.3 million people in need of humanitarian assistance (WFP strategic plan revision). WFP assisted 1.2 million in April 2026, reached 362,861 drought affected people in Marsabit, Mandera, Wajir and Garissa with mobile cash, and needed US$194.4 million for May to October, while backing rainwater harvesting irrigation, including 100,000 m³ structures in Dadaab (country brief, May 2026). It also supports the government’s school meals plan for 10 million children (country brief, October 2024). Water and nutrition meet in the Nakuru trial: SODIS children grew 0.8 cm more in a year.

Faecal contamination also reaches food from the other side. A 2023 study of 66 irrigation water samples on the Nairobi, Machakos interface found high loads of bacterial pathogens and judged the water unfit for fruit and vegetables under Kenyan standards and WHO guidelines (Ahogle et al.), and the FAO/WHO expert meeting it cites names irrigation water as the main entry point for bacterial pathogens on fresh produce. SODIS treats two litres at a time and is no answer for irrigation. What carries over is the logic of testing, measuring dose and filtering turbid water first.

What it means for Kilimora

• Treat SODIS as a bridge. The Nakuru and Kibera results show it protects children where taps are absent, and the Kisumu record shows connections will take years.
• Use dose, not hours. The authors recommend colorimetric dosimeters or phone based UV sensors, longer exposure on cloudy days and pre filtration for turbid water. A simple turbidity check belongs in any training Kilimora supports.
• Test the farm’s own water. The irrigation evidence says the same contamination that sickens households reaches produce. Water testing at partner farm sites and packhouses is the cheapest first step.
• Use the policy window. The 2025 to 2026 AU investment documents and the African Water Facility Strategy 2026 to 2030 give water proposals a policy hook, and programme shortfalls show where delivery partners are needed.
• Replicate it independently. The next step is a multi season trial with dark controls, instrument turbidity and in bottle UV measurement, ideally with a team that did not write the first one.

SOURCES: M. Gathiru, E. Obuya, N. Noah, E. Masika (2026), Field assessment of solar disinfection of water in Obunga slum, Kisumu, Kenya; M. M. Gathiru et al. (2024), Biosynthesized and chemically synthesized Ag/TiO2 nanocomposites, Heliyon; R. Conroy et al. (1999), Solar disinfection of water reduces diarrhoeal disease: an update; M. du Preez et al. (2011), Randomized intervention study of SODIS in Kenyan children, Environ. Sci. Technol.; K. McGuigan, University of Bristol, SODIS trials summary, including motivation results; A. M. A. Ahogle et al. (2023), Bacterial hazards in urban stream irrigation, Nairobi Machakos; Pulitzer Center, Clean water in Kenya (SODIS in Kibera); U.S. EPA, Final Report, grant SU839297. WHO/UNICEF JMP via Our World in Data, Share using safely managed drinking water; University of Leeds, Safely managed drinking water at the point of use in 14 countries; UN Water, SDG 6 snapshot, Kenya. African Union (2008), Sharm El Sheikh Commitments on Water and Sanitation; AU Executive Council (2014), Decision on implementing the Sharm El Sheikh Commitments; UNECA, Water security for Africa’s development; AMCOW, capacity4dev, Africa Water Week summary; Global Water Partnership, High Level Panel partnership, US$64 billion estimate; AU and AIP (2025), Cape Town Declaration on Africa Water Investment; GIZ, Vision to Action: Africa’s first Water Security Implementation Plan. World Bank, Water and Sanitation Development Project (P156634); World Bank, Implementation Status and Results Report, February 2025; World Bank, Kenya Climate Smart Agriculture Project, completion report; Kenya News Agency, KCSAP households and funding; Practical Action, KisumuSAN impact brief; Kenya News Agency, Manyatta water project, Kisumu; IRC, Improving water services to informal settlements, Kisumu. FAO and partners, SOFI 2026 launch; The Reporter, SOFI 2026: Africa 20.0% in 2025; Guardian Nigeria, SOFI 2026: healthy diets out of reach for 67% of Africans; FAO, SOFI 2025 launch; IFPRI, Hunger on the rise across Africa, SOFI 2025; WFP, Kenya Country Brief, May 2026; WFP, Kenya country strategic plan, revision 3; WFP, Kenya Country Brief, October 2024.