Mihir Gupta's Individual Contribution Page

Fall 2012 Semester

Member of the Low Flow Stacked Rapid Sand Filter team. We need to start by collecting data from the existing model. This data would include velocities, headloss, water levels etc for a given flow rate. We then need to use this data to build a mathematical model which will help us find out any potential problems with the setup. It should also help us realise the points at which maximum headloss occurs, so that we may replace the part, as currently, without any sand, the headloss is too great for actual usage.
After making such changes as necessary, a complete fluid and performance testing will be carried out, so that the design may be approved.

UPDATE:

On implementing the mathematical model, it was determined that the current head losses would make the system unstable and would cause unreliable. It was then decided that the spring valves were a major contributor to this high head loss, and thus, they were swapped out with swing check valves. It was also decided that a larger diameter backwash pipe would help reduce this head loss. 

On making these changes, we noticed that during the backwash run, the LFOM continued to pass water. Ideally, there should not be any water flowing out of the LFOM during the backwash, and thus, we had to add a ball valve.

We then noticed that during the backwash run, there were large variations in the head loss. We determined that these were caused by air bubbles, which cause a hydraulic disconnect in small pipes and thus, increase the head loss. We then worked to make the entire system air tight.

The next order of business was performance testing. We added different concentrations of clay suspensions along with a constant coagulant dosage. We obtained a removal percentage of about 99%.