Options
Effect of Clogging on Riverbank Filtration: An Experimental Analysis Using Ganges Riverbed Sediment
Journal
Journal of Hazardous, Toxic, and Radioactive Waste
ISSN
21535493
Date Issued
2022-04-01
Author(s)
Poojitha, S. N.
Hari Prasad, K. S.
Ojha, C. S.P.
Abstract
Riverbank filtration (RBF) is one of the most productive and low-cost technologies for obtaining purified water throughout the year. The inevitable problem associated with RBF is clogging, a complex phenomenon that obstructs the flow. The present study conducted laboratory experiments on two different filter materials (uniformly graded) collected from the Ganges River bed from an RBF site in Haridwar. The primary focus is to study the factors influencing the clogging mechanism that affects the hydraulic conductivity, K of the filter material. The variation in the piezometric head, porosity, specific infiltration resistance, and the progressive clogging of the filter material is studied considering the experimental results. It is observed that the retention and intrusion of suspended particles depend on the grain size of filter material and fine sediments. The clogging of pores is more at the initial depths (2-7.5 cm), demonstrating the phenomenon of physical clogging. For every experimental run performed, as time progressed, with increased resistance and head difference, K and porosity of the filter materials decreased (Filter material-1, 47.17% and 48%; Filter material-2, 93.43% and 81%). With an increase in initial discharge, q0, K is partially recovered, presenting the initial desiltation process. Further, as time elapsed and with an increase in the turbidity, C, from 500 to 1,000 ppm, the clogging advanced resulting in decreased K of the filter materials. Therefore, taken as a whole, the ratio of mean size of filter material to suspended particles, q0, C, and time are considered as the dominant factors influencing the clogging process. Mathematical regression models for both the filter materials are formulated and found to estimate K reliably. The correlation coefficients calculated for both the models at a 95% confidence level are 0.814 and 0.965, respectively, and are statistically significant, presenting the functional dependency of K on the ascertained parameters as acceptable.
Subjects