Mechanistic understandings of the convoluted impacts of hydraulic loading rates, temperature, chlorine doses, and media sizes on the performance of biologically active filtration applied for full-scale drinking water treatment

This study filled the knowledge gap remaining in understanding the convoluted contribution of factors governing the contaminant removal and headloss development performance of Biologically Active Filtration (BAF). This new knowledge was interpreted with an existing model framework built for simulating the performance of BAF by taking into consideration the effects of hydraulic shear stress on biofilm detachment, the effects of temperature on viscosity, and "first dose phenomenon" on chlorine inhibition, which have been confirmed to be critical for accurately interpreting the full-scale headloss data. This redefined BAF model was calibrated and validated using 240 days of continuous data from four full-scale BAF systems operated in a local drinking water treatment plant operated at different media sizes and chlorine doses under fluctuating loading rates and temperatures. The model successfully predicted that BAF runtime is less sensitive to loading rate than to temperature after taking hydraulic shear stress into model consideration. Also, increasing media size from 1.0 to 1.4 mm or applying 1 mg/L chlorine dose could lead to equivalent improvements in BAF runtime and reductions in TOC removal. However, when choosing between these two strategies for real world applications, their differences in flexibility and costs should be taken into consideration.