Incorporating Settling Time Constraints of Energy Storage Systems in Frequency Containment Control Optimization Problems

Response of energy storage units and power plants to a set-point change is not necessarily linear with a constant ramp rate, especially for frequency containment services Which are actuated on a $\leq$ 30s time scale. This is because real systems have time lag in their dynamic behavior Which will cause the output to behave nonlinearly before settling down. This paper shows how to implement this behavior into the linear programming optimization framework using settling time constants of a common energy storage system and power plants. Simulation results show that if linear programming is used to model the units participating in the frequency containment services, neglecting the settling time constraints can lead to decisions Which may not be grid-compliant in practice.