A modular diagnosis system based on fuzzy logic for UASB reactors treating sewage.

A modular diagnosis system (MDS), based on the framework of fuzzy logic, is proposed for upflow anaerobic sludge blanket (UASB) reactors treating sewage. In module 1, turbidity and rainfall information are used to estimate the influent organic content. In module 2, a dynamic fuzzy model is used to estimate the current biogas production from on-line measured variables, such as daily average temperature and the previous biogas flow rate, as well as the organic load. Finally, in module 3, all the information above and the residual value between the measured and estimated biogas production are used to provide diagnostic information about the operation status of the plant. The MDS was validated through its application to two pilot UASB reactors and the results showed that the tool can provide useful diagnoses to avoid plant failures.