State estimation approach for fault diagnosis in anaerobic digester:application to India's largest dairy unit

Improper treatment in sewage treatment plants and effluent treatment plants of industries account to a major amount of water pollution. This made Governments impose strict rules and regulations on reatment plants. As Anaerobic digestion is one of the cheapest and effective waste water treatment the work focuses on improving its efficiency by early Fault detection and diagnosis. In the work concerned, Model based State estimation approach is applied for fault identification in anaerobic digester. The proposed model is validated by measurements of Anaerobic digester from Amul dairy plant. The model is fed to EKF algorithm to estimate the other unmeasured states.EKF is able to estimate 16 other states out of 24 states in the model. Some faults are simulated and EKF is able to estimate the unmeasured states in faulty condition. This information of states aids in early detection of fault and helps in quick recovery of the process without disturbing other sub processes and enhances performance