A group of energy resources (including energy storage units) physically distributed in the network which are placed under a unified and integrated management with a central control system is called a Virtual Power Plant (VPP). The bidding strategy of a VPP, as a participant in energy and reserve markets, has significant role in maximizing its profit. This paper proposes a new mathematical approach based on a comprehensive model for bidding of a VPP in energy and reserve markets. In our proposed model network topology, VPP security constraints, constraints of distributed energy resources (DER) composing the VPP, power loss in the VPP and the balance between supply and demand are considered. The method determines the amount of energy and reserve that should be bought or sold in day-ahead markets, commitment of DER units, charge or discharge status of storage units and the amount of load curtailments. We have used Benders decomposition (BD) method for solving the problem. The results obtained using BD technique is compared with those obtained using genetic algorithm (GA). Simulation results confirm that using BD is more advantageous for solving the problem. It is also shown that BD can be applied to bidding problem of large-scale VPPs considering all the constraints. It is also observed that computational time for solving the problem using GA and the optimality of the solution are major obstacles for applying GA to solve large-scale VPPs.

Keywords