In Jumla, clean water is not guaranteed. Glacial melt carries sediment, surface water is contaminated by animal waste and human activity, and boiling every liter of drinking water is energy-intensive and unsustainable.
We spent three weeks testing four low-cost filtration designs in households across the district. This is what we learned.
The Challenge
Jumla’s water sources vary dramatically by season. During the monsoon, surface water is abundant but heavily contaminated. In winter, springs slow to a trickle and freeze overnight. Most households boil water for 20-30 minutes per batch — consuming precious firewood or kerosene.
What We Tested
We brought four filter prototypes designed for local manufacture:
- Gravity-fed ceramic candle filter — The simplest design. Two stacked buckets with a ceramic element. No moving parts.
- UV purification unit — Solar-powered UV-C lamp for community centers. Higher throughput but more complex.
- Bio-sand filter — Layers of sand and gravel with a biological layer. Low maintenance but requires space.
- Ceramic pot filter with silver coating — Locally fired clay pots treated with colloidal silver. Promising but fragile.
Results
The ceramic candle filter was the clear winner. After two weeks of daily use by a family of five, it delivered 10 liters of clean water per day at near-zero maintenance. The ceramic element cost less than $3 to produce locally.
The UV unit worked well but required reliable solar charging — a challenge during Jumla’s monsoon season. The bio-sand filter was too bulky for household use but shows promise for schools and health posts.
Next Steps
We’re now iterating on the ceramic design: improving the clay配方 for higher altitude firing, adding a pre-filter for monsoon sediment, and designing a 50-liter community-scale version. The full build instructions and BOM will be published on GitHub this quarter.
The water filtration pillar is moving from field testing to prototype refinement. If you want to help — join the Jumla residency or contribute on GitHub.