Hybrid Biological Systems for Wastewater Treatment

Abstract

The current trend in sustainable development deals majorly with the environmental management. There is a need for economically affordable, advanced treatment methods for the proper treatment and management of domestic wastewater containing excess nutrients (such as nitrogen and phosphorous) which otherwise may lead to eutrophication. In the present study, the effect of carbon to nitrogen (C/N) ratio, suspended biomass concentration (X), hydraulic retention time (HRT), and dissolved oxygen (DO) on nutrients removal in a lab-scale activated sludge biofilm (AS-biofilm) reactor was monitored. Based on various trials, it was seen that ASbiofilm reactor achieved good removal efficiencies with respect to COD-92%, NH4+- N- 93%, TN- 86% and TP-52%. Further, in order to improve the quality of the treated wastewater, photocatalysis by TiO2 was investigated as a post-treatment technology, using solar and UV irradiations. The UV photocatalysis was found to be better than solar photocatalysis during the comparative analysis. The maximum removal efficiencies of COD, MPN and phosphorous at optimum conditions in the case of UV and solar irradiations were 72%, 95%, 52% and 71%, 99%, 50% respectively. Similarly, to enhance the performance of the system in terms of nitrogen and phosphorous in addition to carbon removal integrated anaerobic/anoxic/oxic activated sludge biofilm (A2O-AS-biofilm) reactor was designed and operated by varying operating conditions such as C/N ratio, suspended biomass (X), HRT and DO. Based on various trials, it was seen that the A2O-AS-biofilm reactor achieved good removal efficiencies of COD-95.5%, TP-93.1%, NH4+-N-98% and TN-80% when the reactor maintained C/N ratio - 4, suspended biomass (X) - 3 to 3.5 g/L, HRT-10hr, and DO - 1.5 to 2.5mg/L. Applicability of soft computing techniques viz, Adaptive Neuro Fuzzy Inference System (ANFIS), Genetic Algorithm Adaptive Neuro Fuzzy Inference System (GA-ANFIS) and Particle Swarm Optimization Adaptive Neuro Fuzzy Inference System (PSO-ANFIS) to performance prediction of hybrid system was studied. ANFIS was applied on real time WWTP of 43.5 MLD capacity. ANFIS models showed better efficiency while modeling wastewater using multivariate analysis. So in the current study, in order to improve the prediction ability of ANFIS,ii hybrid models such as GA-ANFIS and PSO-ANFIS have been applied for the prediction of effluent TN, COD and TP concentration yielded from a hybrid ASbiofilm reactor. From the results, both GA-ANFIS and PSO-ANFIS proved capable to predict the effluent parameters of the reactor with varying operation conditions and can be adopted for modeling the nonlinear data.