I needed to conduct an experiment to test the design goal: Maintain or improve filtered water quality standards. 

The primary objective of this experiment was to evaluate the performance of three prototype BSFs against two standard plastic control filters over an 8-week period. The study aimed to assess:

• Flow rate 

• Turbidity reduction 

• E. Coli and Total Coliform reduction 

• Usability improvements

Designing the experiment in Cambodia challenged me to work within strong limitations. I had to consider how often we could realistically collect water (100L/day) and how often I could ask staff to drive samples to the lab. Space was also a factor: where could I place filters to keep conditions consistent, especially during the oncoming rainy season?

Budget limitations influenced everything: the number of prototypes I could build, how often I could run tests, and focussing more on turbidity than bacterial testing due to cost. The testing equipment available locally was limited or very expensive, which impacted data collection. With this in mind, reproducibility was a key concern, so I made sure to build at least two of each type of filter.

Contamination was another major issue, especially in a field setting, so I focused on controlling the variables I could, by sanitizing collection containers with alcohol, disinfecting the area, creating a protection for the spout, alternating charge order to avoid systematic bias etc. 

While these constraints made the experiment more complex, they also reflect how the filters would be used in real rural households. I looked at methods from other BioSand Filter studies and adapted them to fit the realities of working in this context.

I ran into several problems during the experiment, mostly due to the limitations I mentioned earlier, and had to adjust as I went. One issue was water splashing onto the filter spout, which would normally be covered by a storage container in my designs. Since we weren’t using containers during testing, I made a simple shield to reduce splash and lower the risk of contamination.

At first, we collected 200L of water by hand every other day, but by day two it was clear that wasn’t going to work. We bought a small pump that fit our budget and cut the collection time in half. I also kept adjusting the sterilization process as I learned what was available and practical: using more alcohol, soaking parts for longer, and wearing gloves when possible. These changes helped keep the setup consistent.

Initial results indicate that reduced height and volume did not measurably compromise turbidity and microbial filtration performance. This may be because most removal happens at the top of the filter in the biolayer.

This indicates a strong proof-of-concept and provides support for further investigation.