Characterization of a novel fiber filtration system with a unique mechanism for adjusting the depth of the filter bed

We developed a novel fiber filtration system, which allows the filter bed packing density to be adjusted, to treat raw water of differing turbidities without coagulants. A bench-scale fiber filter is used to remove 5–20 μm poly(methyl methacrylate) particles with high efficiency, without the use of coagulants. The obtained data are analyzed using Yao’s model and regression models. The first filtration cycle achieved a benchmark removal efficiency of 90 % of 5 μm particles for 71 min with a Darcy velocity of 38.2 m/h. Each cycle included a backwash process, but the filtration time gradually decreased to about 46 min by the 5th cycle. However, pressure loss did not increase significantly during the filtration cycles. The shortening duration of filtration over operating successive cycles was postulated as residual particles remaining in the bed after backwash and rinsing. We measured pressure loss at various Darcy velocities for fiber filters with different packing densities and found that the square term of Darcy velocity in Ergun’s model was significant by the regression analysis. Based on the experimental results of the novel fiber filtration system, we compare the performance and running cost of a conventional fiber filtration system and a rapid sand filtration system. The novel fiber filtration system comprising 940 mm fiber filter elements at a packing density of 110 g/L removed 96 % of 5 μm particles at a Darcy velocity about four times greater than that of the rapid sand filter. The running cost of the novel fiber filter was estimated to be approximately one-sixth that of the conventional fiber filter using poly aluminum chloride and less than approximately 30 % of the sand filter.