Demand response for real-time congestion management incorporating dynamic thermal overloading cost

Capacity challenges are emerging in the low-voltage (LV) distribution networks due to the rapid proliferation of distributed energy resources (DERs) and increasing electrification of loads. The traditional approach of network reinforcement does not achieve the optimal solution due to the inherent uncertainties associated with the DERs. In this article, a methodology of real-time congestion management of MV/LV transformers is proposed. A detailed thermal model of the transformer is used in order to obtain the costs incurred by overloading. An agent-based scalable architecture is adopted to combine distributed with computational intelligence for the optimum procurement of flexibility. The efficiency of the proposed mechanism is investigated through network simulations for a representative Dutch LV network. Simulation results indicate that the methods can effectively alleviate network congestions, while maintaining the desired comfort levels of the prosumers.