Feasible operation region of an electricity distribution network

In the course of net zero carbon transition, electricity distribution networks are faced with great challenges brought by increasing total and peak demand due to the electrification of transport and heat, as well as the significant uncertainties from renewable power generation and customer behaviors (such as electric vehicle travelling behaviors). Therefore, distribution network operators (DNOs) need effective tools to assess the capability of distribution networks in integrating generation and demand to conduct active management and efficient expansion of networks. However, conventional scenario-based assessment methods only provide conservative and limited information on the network capability, also with exponentially increased computational burden. By contrast, another stream of methods with different philosophy, named as operation region-based methods, describe the overall picture of the network capability analytically with little computational power needed. However, the existing linearized analytical regions, which are mainly applied to electricity transmission networks, cannot describe the capability of distribution networks accurately enough. To solve these problems, a novel feasible operation region (FOR) method with quadratic analytical expressions was proposed to characterize the range of the operating states of distribution networks, where the thermal and voltage constraints will not be violated. The FOR is a geometry in a high-dimensional space, and thus a high-dimensional error analysis approach was further developed for validating the proposed method. The boundary errors are described by multiple distance functions and operational indices, and the conservativeness of the analytical boundaries are quantified. An 11 kV radial distribution network from the United Kingdom Generic Distribution System (UKGDS) was used for the case study. The simulation results show that the quadratic analytical boundaries well approximated the real boundaries of FOR. The maximum errors for thermal and voltage boundaries would maximally cause an overcurrent up to 116 % and an undervoltage down to 0.96p.u., which are able to satisfy the requirements of engineering practice. Compared to the existing linear approximation (termed as hyperplane expressions) of FOR boundaries, the proposed quadratic expressions were proved to have higher accuracy.