Comparison of physicochemical iron removal mechanisms in filters

Water supply companies are continually seeking means to improve the process efficiency of iron removal from groundwater in order to minimize the deposition of iron in distribution networks, backwash water use, and volume of the sludge produced. It is expected that the process efficiency can be optimized with a better understanding of the different mechanisms involved in the iron removal process and their optimal utilization. In order to investigate the options to improve the performance of an iron removal plant, pilot experiments were conducted with filters operating in two different iron removal modes, namely floc filtration and adsorptive filtration. The results were then compared with the performance of the full-scale plant Gilze, which operates mainly in the floc filtration mode. Results demonstrated that the performance of WTP Gilze in terms of ripening time, filter run time and filtrate quality (residual iron and turbidity) could be improved by operating the filters in (a) adsorptive mode with fine sand (0.5–0.8 mm) and low influent oxygen, or (b) in floc filtration mode with dual media (anthracite 0.8–1.2 mm and sand 0.5–0.8 mm). In general, for a single media fine sand filter, the adsorptive mode gave a longer run time than the floc filtration mode with comparable filtrate quality. In the adsorptive mode, oxygen content and pre-oxidation time should be kept as low as possible. However, to ensure sufficient removal of manganese and ammonium present in groundwater, a minimum oxygen concentration is required (about 2 mg l⁻¹ in case of WTP Gilze). Experimental results suggested that filters with coarser single media or dual media operating mainly under adsorptive mode with short pre-oxidation time could further improve filter performance and probably result in significant cost savings.