A novel method for modeling effluent quality index using Bayesian belief network

Reliable estimation of the effluent quality from a municipal wastewater treatment plant is important for safe discharge and reuse of the treated stream as well as control and monitoring of treatment processes. The quality index is a summative index that can be used for a rapid assessment of water and treated wastewater to rank the quality level. Since there is no quality index for different reuse options of reclaimed wastewater, this study aims to propose a quality index for the treated wastewater focusing on reusing purpose. The significant quality parameters associated with EQI were found using the Delphi method and weighted by analytic hierarchy process decision-making tool. Finally, the Bayesian network analysis was employed to estimate the probability of meeting legal reuse and disposal requirements for EQI based on data collected from south wastewater treatment plant in Tehran city, Iran. The results of Bayesian network analysis were compared with the aggregation method as a widely used method for estimating quality indices. Results revealed Bayesian model had great potential for effluent quality index modeling and significantly increased the precision and the accuracy of estimating the EQI formula. The suggested methodology can provide valuable support also to such practice.