Reserve market scheduling considering both capacity and energy bids of reserve

•Proportion of reserves’ energy cost to total operation cost is forecasted by ANN.•Sensitivity analysis based method is proposed for features selection.•Effects of reserves’ applied energy costs are entered to the reserve market model.•Performance of proposed method is evaluated under normal and contingency conditions.•Significant improvements are achieved in comparison with conventional models.

In conventional Ancillary Service Markets (ASM), where Independent System Operator (ISO) purchases requirements for system safe and reliable operation, capacity and energy of reserves have always been considered as individual commodities. Market participants offer their capacity bids and energy bids to the ASM, then the ISO decides on purchasing the required capacity using an optimization model based on capacity bids while energy bids are neglected. During the operation time, the ISO has to call some of the purchased capacity to provide required energy for frequency response, and pay them accordingly at their reserve bids. Therefore, the ISO has to pay for both capacity and applied energy while ISO’s model considers capacity costs alone, thus it cannot reach the overall optimum point. To develop ISO’s model for considering energy bids, an Artificial Neural Network (ANN) based method is proposed in this paper to define a combination of energy and capacity bids to be substituted for solo capacity bids in ISO’s model of market. A modified 24-bus IEEE test system is employed to illustrate the proposed methodology.