Long term pump storage capacity planning for microgrid in model predictive control platform

Abstract

This work considers the optimal long term pumped storage capacity problem for a microgrid equipped with a Wind Farm (WF), Interruptible Loads (IL), transmission line to a main grid and Pump Storage (PS). Typically, the WF is operated according to wind speed to charge PS and supply electrical power to satisfy the load demands. A linear model is established to describe the system dynamics of the central plant. A model predictive control (MPC) problem is formulated to obtain optimal set points to satisfy the consumer demands and minimize daily operation cost. At each 24 hour time scale, the MPC problem is represented as a large-scale mixed-integer linear programming (MILP) problem. After solving the MILP for 26 years, we propose a financial algorithm to obtain best capacity for the PS in longevity. Simulation results clearly show how much IL cost should be expected for different amounts of pumped storage capacity. A comparison between the operating and investment costs is then used to determine the optimal pumped storage capacity. Finally, various sensitivity analyses are performed to assess the effect of key parameters on this optimal capacity.