A Framework for Multi-Regional Real-Time Pricing in Distribution Grids

Abstract

A novel real-time distributed market framework at the distribution grid level is proposed in this paper. The framework coordinates information in a hierarchical manner, while aiming to achieve consensus by sharing physically coupled information among multiple regional operators. In principle, each region maximizes its individual social-welfare problem, which is the local (regional) version of the overall social-welfare maximization of the distribution grid. The individual problem incorporates the coupled physical information from its neighboring regions. The consensus among all regions is then enforced through the proposed consensus-alternating direction method of multipliers structured trust-region (CAST) algorithm. Upon achieving convergence, the distribution locational marginal price (DLMP) is recovered for each region, which is novel in a sense that on the one hand, it is computed in a distributed manner, i.e., with preserving local information, and on the other hand, it accurately represents loss allocation from its neighboring regions. The proposed methodologies are tested on an IEEE 33-bus distribution grid and on a 144-bus network with five regions.