Dynamic Real-Time Optimization of Air Conditioning Systems in Residential Houses under Different Electricity Pricing Structures

This paper investigates potential cost savings and peak shifting in operating residential houses air-conditioning systems through dynamic real-time optimization (D-RTO). Standard design data collected from BEopt (Building Energy Optimization) software were used in Matlab/Simulink to simulate cooling energy consumption of 60 different types of houses, which were scaled up to represent a community scale of 6000 houses for grid-level studies. Four different electricity pricing structures were employed for each house; flat pricing (FP), time-of-use (TOU), critical-peak pricing (CPP), and realtime pricing (RTP). The D-RTO was formulated as a linear programming problem with the air conditioning temperature setpoint and the cooling energy being decision variables of the model. The D-RTO determines day-ahead values for the variables of the system based on the weather forecast and energy price signals. The D- RTO uses a model predictive control (MPC) like approach in which the problem is solved for 24 hours in advance, but the solutions are implemented on a receding horizon, where the prediction interval moves forward by one hour each time step. Results show significant energy cost and energy consumption reductions for the optimized cases versus the non-optimized cases of each pricing structure. Also, the four pricing structures are ranked based on their capabilities for cost and energy savings and peak shifting in the optimized cases.