FUZZY DISSOLVED OXYGEN CONTROLLER APPLIED IN ACTIVATED SLUDGE PROCESS

Abstract

Roughly 80% of waste water treatment plants are based on active sludge technique due to its simplicity and reliability. Under aerobic conditions microorganisms can grow on some organic pollutants present in the wastewater. As such, dissolved oxygen control is paramount in order to save energy, improve effluent quality and achieve high nitrogen removal. However, due to the nonlinearities of bioprocesses, dissolved oxygen (DO) is a challenge in terms of variable control. In Suzano municipal Waste Water Treatment Plant (WWTP) - located in São Paulo Metropolitan area - dissolved oxygen control is conducted based on a set of heuristic rules stablished by skilful operators. Although yielding satisfactory results, staff training is time consuming. Under the light of these facts, this study aimed to develop a fuzzy DO controller applied to activated sludge process. Data collection and implementation of the fuzzy control system were conducted in bioreactors at Suzano WWTP. Three fuzzy inference systems were developed: “aerator number”, “aerator speed” and “submergence height”. The fuzzy controller was tested and calibrated with a dynamic simulator and implemented as a supervisory controller, yielding positive results. Statistically, dissolved oxygen mean was 1.75 mg/L with a standard deviation of 0.25 mg/L. Based on the results the application of fuzzy logic for DO control in activated sludge process has proven not only viable, but an interesting control option. Fuzzy logic overcame DO non-linearities by translating an empirical knowledge into a linguistic, rule-based controller. It is a promising technique, bringing significant advances, both in treatment performance and energy efficiency.