Filter Media Modification In Rapid Sand Filtration
The main objective of this research was to improve the filtration technologies to make them more sustainable and accessible. This study focused on developing improved operating methods for rapid sand filtration technology that would not require costly infrastructure upgrades to implement. Unlike previous filter media modifications conducted by other researchers, the filter media in this study was modified in situ. Process Controller software was used for automation of the test apparatus facilitating systematic variation of parameters and replication of results. Several different coagulants at varying dosages were applied by either downflow or upflow to modify a sand filter medium. Initially coagulants were added in downflow mode to the top of a 7.5 cm deep sand column prior to challenging the filter with an otherwise untreated kaolin suspension. Three coagulants, alum (Al2(SO4)3(dot)14H2O), ferric chloride (FeCl3), and polyaluminum chloride (PACl), were utilized separately to modify the sand filter medium. After modification of the filter medium, an initial particle removal of 96% was achieved by all coagulants versus 60% removal in the absence of pretreatment. Pretreatment with PACl and alum showed increasing particle removal with increasing dosage up to a maximum of 550 mmol Al/m2 after which filter performance declined. Pretreatment with FeCl3 increased particle removal over the entire range of dosages evaluated (70 to 2200 mmol Fe/m2) but headloss through FeCl3 treated filters became prohibitive at the highest dosage. Although downflow application of coagulant showed promise there was strong evidence that better performance and lower head loss would be possible if a more uniform application of the coagulant throughout the filter could be attained. Thus, a novel fluidized-bed pretreatment process was developed to modify the sand medium at the end of the backwash cycle of the filter. During backwash, a mixture of alum, base, and tap water were pumped into the filter upward from the bottom. The ensuing precipitation of Al(OH)3(am) in the filter pores enhanced the efficiency of turbidity removal from untreated raw water (up to 99.6%) without a substantial increase of head loss (≈14 cm). While pretreatment with Al(OH)3(am) was effective at enhancing particle removal, measurements of dissolved aluminum in the filter effluent showed that this process modification should only be considered for waters with circumneutral pH. At pH of 8 a pretreatment dosage of 16 mol Al/m3 resulted in effluent dissolved Al in excess of the EPA secondary drinking water standard of 0.05~0.2 mg/L. To clearly understand the fundamental aspects of the enhanced particle removal by fluidized-bed pretreatment with Al(OH)3(am), two alternative mechanisms were hypothesized: (1) Al(OH)3(am) coats the sand filter medium, and alters its porosity. (2) Precipitated Al(OH)3(am) embedded within the media pores acts as an additional filter medium that enhances the particle removal efficiency. Mathematical models for these two mechanisms were constructed and compared with experimental data for particle removal and head loss. Model predictions suggest that particle removal by the second mechanism, filtration through Al(OH)3(am) flocs, can account for the observed improvement of the filter performance.
dissertation or thesis