Forecast and energy management system (F-EMS) framework for optimal operation of sewage treatment plants

Aeration processes at Gujarat Infrastructure and Financial Tec (GIFT) city's sewage treatment plant (STP) account upto 40% of entire STP's energy demand. Solar photovoltaic (PV) rooftop rated 15kW and 10kWh battery energy storage system (BESS) will be installed to address the energy demand of aeration blower locally. In this paper an extreme gradient boost (XGboost) based solar PV forecast algorithm together with linear programming (LP) based energy management system (EMS) algorithm is developed as a framework namely “F-EMS” to optimally charge or discharge BESS to address varying energy demand of aeration blower. The analysis is carried out considering realistic load data of aeration blower and constraint of aeration process, BESS, and utility grid. The efficiency of EMS algorithm depends on accuracy of solar PV forecast where the XGboost based forecast algorithm predicts the solar PV generation with good accuracy i.e. less than 4% of root mean square error (RMSE) during worst fluctuating day. On analysis it was found that by implementing the F-EMS framework, a maximum of savings upto 20% on aeration blower's electricity bills can be achieved.