Optimized Self-Healing Protocols Utilizing Conservative Voltage Reduction

Abstract

Smart protocols have become prominent in the past decade as they provide significant benefits to the grid. These benefits include allowing distributed generation at the distribution level and self-healing protocols. The ability to self-heal after a contingency is the most attractive feature of the smart grid as it leads to increasing the efficiency and reliability of the system. It is commonly carried out using reconfiguration protocols that reconfigure the system to reroute the power to the loads affected by the contingency. However, the benefit of rerouting the system is constrained by the generation constraints, thermal limitations, and voltage constraints. In the case of grid-connected systems, reconfiguration alone could restore a significant portion of the load. Still, in isolated systems with limited generation, reconfiguration is not very efficient in carrying out the self-healing scheme. This work proposes enhancing the self-healing scheme by adding a concept commonly used in the conventional power grid called conservative voltage reduction (CVR). CVR is deployed on voltage-dependent loads across the whole system as it reduces the voltage on all the loads, resulting in reducing the power consumed by the load. The effect of CVR is more significant in systems with limited generation as it provides more available power to be rerouted to the out-of-service loads. Combining the reconfiguration with the CVR to carry out the self-healing scheme has been shown to significantly affect the number of loads reconnected and improve the self-healing scheme significantly. The proposed method has been implemented on the IEEE-69 bus test system considering its thermal and voltage limitations. The system has been constructed to include both dispatchable and non-dispatchable distributed generation (DG). The non-dispatchable generation is modeled probabilistically to mimic the stochastic characteristics of wind DG. The probabilistic method has been considered with the combined generation and load model to simulate the realistic pattern of the renewable DGs and loads. This has been modeled using GAMS software as an MINLP which minimizes the amount of unserved power.

Date of Award	Jul 2022
Original language	American English