Slow sand filtration pre-treatment with alum is efficient, but is it effective?

Abstract

Enhancing the performance of slow sand filter pre-treatment with alum (i.e. direct or contact filtration) has been proposed as an alternative to overcome limitations of conventional gravel (roughing) pre-filtration. Experimental results revealed high turbidity reduction efficiencies in alum-dosed pre-filtration. However, due to the alum coagulation, the nature of the particulates in the pre-treatment effluent changed and had a potential to shorten the downstream slow sand filter run lengths by approximately 50% under the conditions tested. Hence, depending on the effluent turbidity levels the effectiveness of the alum-dosed pre-treatment could be compromised, despite its high efficiency. Relatively low turbidity levels (<2 NTU) were needed to minimise excessive headloss development in alum-dosed filters. However, the necessary resources to achieve such process control may not be available in developing country contexts. Furthermore, full-scale field experiences indicated the limited effectiveness of alum dosing prior to slow sand filters and the difficulties in maintaining an adequate chemical dosing in under-resourced settings.