Impact of Mechanical System Modeling on Compressed Air Energy Storage Models for Frequency Regulation

In this paper, the modeling of compressor and turbine maps, their associated controls, and the cavern pressure of Compressed Air Energy Storage (CAES) system for frequency regulation studies are discussed. Artificial Neural Networks (ANNs) and a polynomial function are used to model the compressor map, while a model of a choked nozzle is used to model the turbine map. Special control strategies are proposed to ensure that the compressor and turbine operate within their allowed pressure ratios, as limited by the cavern pressure. A CAES facility is then simulated, connected to a power system with high penetration of wind generation, to study frequency regulation, demonstrating that for cavern pressures below nominal values, the frequency regulation performance of CAES is limited, while at higher pressures it improves considerably.