ANFIS based Modelling of dewatering performance and polymer dose optimization in a wastewater treatment plant

Abstract

Anaerobic digestion coupled with dewatering is a common and widely used sludge treatment methods in a wastewater treatment plant (WWTP). Many researchers have employed adaptive neuro-fuzzy logic inference system (ANFIS) successfully to optimize the output of anaerobic digesters. However, there is no published work on the application of ANFIS to optimize the performance of the dewatering section of a WWTP. In this study, ANFIS was applied to model the relationship between biosolid quality (total solid content in biosolid) to total solid, volatile solid content, volatile fatty acid of digested sludge and polymer dose. Historical data on various physico-chemical process parameters such as total solid, volatile solid, volatile fatty acid content of digested sludge, total solid content of biosolid and polymer dose from the dewatering section of a WWTP over a period of four years was used for the analysis. The modelling study showed that the optimum total solid content (DS TS) was in the lower DS TS range of 1.3–1.5% w/w and the optimum volatile solid content (DS VS) was also in the lower DS VS range of 70–75% of TS of the digested sludge. The Optimum polymer dose range of 20 to 60 kg/ton of dry solid (kg/TDS) showing better biosolid capture compared to higher polymer dose ranges for the same total solid and volatile solid content. It was found that lower total solid and volatile solid concentration coupled with relatively low polymer dose yield higher biosolid capture rate. ANFIS based predictive model are useful for determination of optimum polymer dose in efficient dewatering performance.