RESEARCH ARTICLE A bi‐level stochastic framework for VPP decision making in a joint market considering a novel demand response scheme Ehsan Ghorbankhani | Ali Badri Faculty of Electrical Engineering, Shahid Rajaee Teacher Training University, Tehran, Iran Correspondence Ali Badri, Faculty of Electrical Engineering, Shahid Rajaee Teacher Training University, Lavizan, Tehran, Iran. Email: ali.badri@srttu.edu Summary This paper presents a stochastic decision‐making framework for an optimal bid- ding strategy of a virtual power plant (VPP) in a joint day‐ahead and regulation (balancing) market. The VPP seeks to maximize its expected profit in day‐ahead market and compensates its deviation in balancing market. Due to the inherent uncertainties of renewable energy resources and also other market participant strategies, a stochastic programming is used for underlying optimization prob- lem. In this regard, a 2‐stage bi‐level problem is presented in which upper level represents VPP profit maximization and lower level deals with market clearing problem in which power transfer distribution factors are used to incorporate transmission constraints. Uncertainties related to stochastic generation and market participant offer/bid curves are modeled via scenarios. A mathematical programming with equilibrium constraints is obtained by reformulating the lower‐level problem using Karush‐Kuhn‐Tucker optimality conditions. The resulting mathematical programming with equilibrium constraints is converted into a tractable mixed‐integer linear programming problem with strong duality theorem. The considered VPP is assumed to be commercial and consists of sto- chastic generation units (wind and solar), conventional power plant, energy system storage, and adjustable internal loads. A novel demand response scheme is introduced into the VPP portfolio in which VPP is penalized through shifting load amount and operating time interval as well. Finally, a trade‐off between profit and risks associated with uncertainties is explicitly taken into account using the conditional value at risk. To assess the validity and the effectiveness of the proposed model, a 6‐bus test system is chosen to apply the model. 1 | INTRODUCTION The mounting concerns of energy crisis by running out fossil fuels in the near future have attracted worldwide attention to substitute these sources of energy by an alternative energy sources. All these sources have the common and outstand- ing features of very low operation costs, sustainability, and being well replenished. Among them, wind and solar by the high production capacities are of great importance. The most critical drawback of these sources is their unpredictability, which poses a great challenge to the system operator. So the approach towards them is fundamentally different, and the way they are dispatched should be distinct from conventional generation sources. Generally, electricity markets consist of 2 main sections which the vast majority of transactions take place there: the day‐ahead market and the regulation market. In day‐ahead market, the market operator clears the market via market Received: 28 April 2017 Revised: 25 July 2017 Accepted: 17 September 2017 DOI: 10.1002/etep.2473 Int Trans Electr Energ Syst. 2017;e2473. https://doi.org/10.1002/etep.2473 Copyright © 2017 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/etep 1 of 18