Modelling and simulation of the biological and physical processes of slow sand filtration
Slow sand filtration (SSF) is the earliest form of engineered potable water treatment and remains one of the most efficient processes for improving the physical, biological and chemical quality of water. However, whilst widely used throughout the world, knowledge of the filtration mechanisms remains limited. This is important in understanding and managing the processes that are responsible for gradually blocking the filter reducing its operational life and filtration efficiency. The objective of this thesis was to develop a mechanistic simulation model of the fundamental physico-chemical and biological processes responsible for the filtration mechanisms operating in slow sand filters. The model solves a set of equations describing schmutzdecke development above the sand and microbial biomass growth within the sand. The model assumes that the schmutzdecke layer contributes to the water purification process and its growth is described as linear function in relation to time. The dynamic interactions between the principal groups of microorganisms including: algae, bacteria and protozoa, were modelled using Monod-type kinetic equations. The filtration performance of the filter media was defined in the model by the removal of particulate material from water and was represented by a combination of headloss and filtration coefficient functions. The model was calibrated and verified using data from full and pilot plant-scale SSF operated by Thames Water Utilities Ltd. Simulation results showed that interstitial biomass was the smallest part of the bulk specific deposit in both covered and uncovered filters. However, microbial dynamics played an important role in the filtration performance. Schmutzdecke development had a major influence on the operation of uncovered filters and was responsible for the significant increase of headloss observed during operation. The model provides a representation of the fundamental nature of SSF processes and could form the basis of an operational management system to optimise SSF.