LAUSR

This paper investigates an optimal sizing strategy for an islanded building microgrid. The microgrid composites a rooftop Photovoltaic (PV) system, a Battery Energy Storage System (BESS), an ice-Thermal Energy Storage System (ice-TESS), and loads. The loads are divided into two sets based on their ability to participate in demand response: i) Plugged Loads (PL) such as lights, and ii) Cooling Loads (CL) such as air-conditioners. The microgrid is islanded and loads must be supplied with local generation resources. Therefore, the BESS is deployed to offset the PV output's variability, and the absence of PV power supply at night time. However, relying only on the BESS incurs high stress and shortens the BESS's lifetime. Hence, we propose an optimal sizing strategy of the microgrid constituents, where the BESS coordinates with the ice-TESS to maintain the balance between generation and load in the microgrid. Nevertheless, the dispatch commands cannot swing freely between the two ESSs because of the difference in the type of energy delivery. Specifically, the BESS stores electric energy and can supply both PL and CL. On the other hand, the TESS can only supply the CL. Hence, the BESS-TESS coordination is also aided by a customer-friendly shifting and curtailment mechanism of CL. The design incorporates the effect of weather uncertainty on the PV output. Weather variations are imitated using Recurrent Neural Networks trained on 19-years of contiguous hourly weather data. After optimizing the sizes of the microgrid constituents, the optimal sizes are used in a detailed dynamic model of the system for a real-time simulation on the OPAL-RT platform. The validation results demonstrate the successful coordinated operation of the microgrid constituents which are cost-effective in sizing.