January 20, 2010 - Ddegeya

The clinic received 108 patients yesterday. Today has been equally busy, with the screams of babies being tortured. The notable point is that a boy came in with some 4-cm wood splinter in his leg. Apparently, 3 months ago, he was attached by a cow and fell onto some wood. The splinter had stayed in his body for all this time. The doctors applied some local anaesthetic to remove the splinter, and the boy reportedly did not flinch. Probably this is not his worst experience.

This morning, we set up two solar cookers again, one with the aluminium box made yesterday as the inner box (containing the kettle), and one with a traditional cardboard box as the inner box (containing the pot). The outer boxes are the same in both case. However, Kevin clumsily broke one of the framed glass, and we therefore did not use any seal for the aluminium box (but covered the outside with the transluscent sheet bought yesterday). For the other prototype, the inner box was covered with an unframed piece of glass, and the outer box, initially none, but later when the heat was not enough, the transluscent sheet (which appeared to improve the performance). The day was sunny and cloudy on and off, and by 1:30 p.m., the sky turned permenantly cloudy, and the WAPIs had not melted. So we called the day to the experiments. Tomorrow we will paint the kettle and the pot black and repeat, hopefully with stronger sunlight.

Happily, the bucket from yesterday did not seem to have made any leaks, so we washed the larger gravel and the smaller gravel to put into the bucket. The fine sand needed many more washes, as the wastewater continued to be cloudy/yellowish. At the end, we still didn't think that our sand was clean, as it was impossible to reach the depths of the bucket. Overall, we had 11 cm of large/small gravel, and about 20 cm of fine sand on top. We filled the sand slowly into the water in order to ensure that no bubbles were trapped. Seeing that we still had some headspace at the top after the sand had been lain, we decided to use a larger bucket as the diffuser plate (in lieu of the smaller, less deep blue bucket). The thing is that with the larger bucket, the standing water level will actually be above the bottom of this bucket, but we do not think that this will affect the oxygen supply of the biological layer too much, and this makes pouring easier (e.g. the user does not have to aim specifically into the smaller bucket). It was afternoon when everything was completed, and we started pouring water into the bucket. Over the afternoon, Kevin must have hauled over 200 L of water from BH4. Our greatest concern is that because we really do not have coarse sand but rather small gravel, the sand would pass through the pores and come out into the efflux, resulting in a constant loss of sand and a cloudy efflux that would not clear. Initially, the efflux was cloudy, as expected. However, probably because we washed the sand quite thoroughly, the efflux quickly cleared. By the first 20 L, the turbidity dropped to below 10 NTU, and after about 60 L, it was below what is measurable using the turbidity tube (< 5 NTU). The flow rate, we found, is dependent on the water level. When the bucket was completely filled, 1 L of water took about 53-57 seconds; when the water level was about halfway to the standing level, filling 1 L could take 64-77 seconds. But this is within reasonable flow rate of Biosand filter. We just hope that this does not get clogged up too soon, which can dramatically drop the flow rate. We plan to test this prototype for one or two more days, before passing the ownership to Joseph. Joseph initially watched and helped a little with the assembly of the sand filter this morning, but then he was called away to accompany someone to Entebbe to pick up some delayed luggage. We will make sure to quiz Joseph to make sure he understands the details of the workings of the Biosand filter.

We pretty much exhausted the majority of our washed sand in building the first Biosand filter prototype. We do not have a lot of sifted sand to wash with (and sifting more fine sand will take an enormous amount of time). Kevin has attempted to wash the coarse sand pile---while this sand is supposedly mixed with the undesirable clay/mud particles too small to be separated by hand, Kevin has so far observed reasonable results in the number of washes for a certain amount of coarse sand to clear. If this goes well, we might start a coarse sand layer in the second prototype.

While Kevin was doing most of the Biosand filter tests, Mike and Dave made 3 trips to BH5. Equipped with new tools they managed to disconnect the pumping chain from the metal rod and, after much heaving, pull a total of 17 three meter sections of metal rod from the riser pipe until the pump valve appeared. As this was a Mark 3 pump this was possible without also removing the riser pipe - something which was then attempted but found to be an impossible task without the use of a rig. The riser pipe appeared to be wet from only around 10 meters depth. The valve seal was also torn, but without being able to remove and inspect the riser pipe it is impossible to tell whether the valve was the sole problem causing the pump not to function. After much straining Mike and Dave managed to replace the 40+ meters of pipe (minus the broken valve) and headed to bore hole 1 to check the depth of the bore hole with an ingenius tape-measure-with-hammer-attached-device to check whether a longer pipe could be used to replace the rusted 10 ft one removed the previous day. They found the bore hole depth to be around 14 ft, making the 10 ft pipe removed around the optimum length for the bore.