Virtual Power Plant (VPP) scheduling with uncertain multiple Locational Marginal Prices

Abstract

Aggregation of heterogeneous distributed energy resources (DERs) and prosumers for profit maximisation by virtue of the market interface has evolved into virtual power plants (VPPs). The VPP operator optimises its aggregated resources to maximise profit by participating in the wholesale electricity market. VPP resources may be geographically dispersed and connected to a transmission grid at various nodes having different locational marginal prices (LMPs). These LMPs vary throughout the day and exhibit mutual correlation. Considering this, the current study presents a profit maximisation framework for VPP operators working under multiple LMPs, with their correlated uncertainties. This work also provides VPP resource scheduling for profit maximisation, considering optimal trading at various LMP nodes for a constrained internal network. This decision is devised using Markowitz's mean-variance (M-V) criterion under expedient correlation between volatile LMPs. A numerical case study on the proposed model with the PJM market demonstrates that LMP correlation alters the role of the VPP operator. This affects its trading and scheduling decisions as a buyer/seller at various time intervals to improve its profit–risk trade-off and consequently the market interfacing of the VPP operator. The proposed model is relevant in practical market conditions for day-ahead/medium-term VPP planning while managing market uncertainty.