Stochastic optimization for retailers with distributed wind generation considering demand response

Abstract

Abstract In this paper, a multi-stage stochastic model is presented for a renewable distributed generation (RDG)-owning retailer to determine the trading strategies existing in a competitive electricity market. Uncertainties associated with wholesale electricity market price, clients’ consumption and power output of wind resources are considered through auto regressive integrated moving average (ARIMA) approach. In the proposed method, three trading floors are addressed for the retailer to hedge against the uncertainties. In the first stage, the retailer participates in day-ahead market to supply the clients and in the second stage, intraday market is addressed to allow the retailer to modify the schedule of its clients’ consumption/RDG production. Due to unfavorable uncertainties, especially in renewable power production, real-time market is considered in the third stage to diminish the uncertainty at power delivery time. Cost function of wind resources considering capital, operation and maintenance (O&M) cost is incorporated in the objective function to increase the applicability of the mechanism. The proposed approach is formulated for risk-averse and risk-taker retailer through conditional value at risk (CVaR) approach. In order to study the impact of retail strategies on consumption pattern and consumers’ electricity bills, time-of-use (TOU) demand response programs are discussed in this paper. Formulating the problem, the mixed integer non-linear programming (MILNP) problem is transformed into mixed integer linear programming (MILP) by jointly using decomposition and disjunctive constraints. Finally, a case study containing wind power resources, energy storage system and retailer is considered to analyze the proficiency of the proposed approach.

Keywords

• Renewable
• Retailer
• Stochastic
• Uncertainty
• Demand response