Modeling and performance assessment of wastewater treatment plant under uncertainty using Fuzzy Bayesian Networks

Wastewater treatment Plant (WWTP) system play an important role in protecting eco-system and public health. The performance of a WWTP system depends on several interconnected components performance, and their performance often changes under uncertain operating conditions. Traditional reliability methods do not properly handle the uncertainty and vagueness related to different component failure/repair data. This study aims to evaluate the performance and reliability of a WWTP components under uncertainty using a Fuzzy Bayesian Network (FBN) approach. The main components considered are bar screen, grit chamber, primary and secondary sedimentation tank, aeration tank, filtration unit, and disinfection unit. Each component is studied in three states namely fully working, partially working and complete failed state. The proposed model helps to estimate fuzzy reliability, posterior probabilities, failure analysis and identify weak components of the system. The reported reliability value of 84.60% denotes the probability that the WWTP operates in acceptable performance states (fully operational or degraded) as estimated by the fuzzy Bayesian network model, and blowers, air diffusers, and backwash system are Identified as the most critical components during system failure. The model also provides fuzzy posterior probabilities, failure analysis and performs sensitivity analysis to rank the components of the WWTP system. The main contribution of this study is the integration of fuzzy logic and Bayesian network method for performance analysis of WWTP under uncertain and multi-state operating conditions. This approach provides a robust framework for enhancing system reliability, informing maintenance prioritization, and supporting sustainable wastewater management.