Experimental Investigation of Inverter-Battery System Efficiency for Energy Scheduling

Battery energy storage systems (BESSs) are widely used in energy management applications to maximize arbitrage profits. The consideration of the power losses from both battery and inverter components in optimization schemes is vital to ensure the economic BESSs scheduling. Towards this direction, this work develops two optimization models that consider the combined inverter-battery efficiency. The first model represents the non-linear efficiency curve of the inverter using a piecewise linear approximation, resulting in mixed-integer linear programming formulations. The second model approximates the nonlinear efficiency curve using a constant term, resulting in linear programming formulations which can be fast and reliably solved. Moreover, an experimental investigation is performed to derive the efficiency curve of a real inverter-battery system, where the parameters of the two models are identified. Simulation results highlight the necessity to consider the efficiencies from both battery and inverter components to reduce modelling inaccuracies and hence to maximize profits. The results also show the capability of the linear programming model to yield high-quality solutions with an approximation error of less than 1%.