Two-stage interval scheduling of virtual power plant in day-ahead and real-time markets considering compressed air energy storage wind turbine

Abstract

Air pollution, global warming, depletion of fossil fuels, and difficulty of transferring power to remote areas are some factors that have led to the growth of distributed generation. In addition, low capacity, invisibility to the independent system operator, and uncertainty in their output power has led to the scheduling of these units as a virtual power plant (VPP). In this study, a VPP including wind turbines (WT), photo-voltaic (PV) panels, conventional generator (CG), energy storage systems, and controllable loads is studied in a day-ahead and real-time market framework. Uncertainties in WT output power, PV output power, load, and market prices are modeled using the interval uncertainty method. In addition, demand response program is modeled using utility function. A new dispatchable WT, i.e. compressed air energy storage WT (CA-WT) technology, is evaluated in terms of its impact on VPP profit and uncertainty management. In order to evaluate the the benefit of investing in a CA-WT as compared to a traditional WT, two scenarios are considered. The scheduling problem is modeled as a mixed-integer linear program using GAMS and solved by CPLEX. Simulation is performed on a 33-bus distribution network. The results show that CA-WT effectively handles uncertainty, reduces the use of CG, and increases power sales to the markets. CA-WT increased the profit of the VPP by about 19.3 and 20.2 % respectivelu in both day-ahead and real-time market. Sensitivity analysis of VPP profit on different intervals of uncertain parameters is also investigated.