Optimization of direct filtration: Experiments and mathematical models

A laboratory-scale set-up consisting of helicoidal flocculator and rapid sand filter was used to study the implication of floes in direct filtration for optimizing the design parameters. Two different methods, one using a flocculation and filtration mathematical model and other using Ives' filterability number were used to optimize the direct filtration. The results indicated that (i) there is an optimum range of floe size which resulted in prolonged filter run, (ii) alum dose played an important role in floe size and density, (iii) a good compromise between velocity gradient and flocculation time is essential in optimizing the direct filtration performance, and (iv) the optimum filter depth increased with higher filtration rate. The simulation results indicated that a floe size of 62 μm was found to be optimum for direct filtration which corresponded to 2.5 minutes of flocculation time at the velocity gradient values of 26.5 s^-1. This optimum value shifted with filter medium size, depth and filtration velocity. On the other hand, Ives' filterability number indicated that a floe size of 57-76 im was found to be optimum which corresponded to 4.8 - 7.2 minutes of flocculation time and at a velocity gradient values of 33.6 - 79.6 s^-1. These methods will help in optimizing the design parameters of direct filtration in a rational manner using minimum number of experiments with the specific raw water and chemicals.