Optimizing Low-Voltage Ride-Through (LVRT) Curves for Wind Farms Using Genetic Algorithms: A Case Study of Taiwan Power System

Recently the capacity of installed wind energy has continued to expand, necessitating that power companies develop specified low-voltage ride-through (LVRT) curves to address unexpected power outages caused by wind farms. To date, the literature lacks reports on the specification of LVRT curves, and the state-operated Taiwan Power Company (Taipower) lacks established guidelines for revising the currently utilized LVRT curves. This study aims to specify LVRT curves based on a projected power load for 2025, as forecasted by Taipower. Simulations were conducted using the Power System Simulator for Engineering (PSSE) equipped with a GEWT 4.0 MW wind turbine module. An objective function was defined to minimize the manufacturing costs of wind turbines while ensuring stability and incorporating the critical clearing time (CCT) and other conditions as constraints. For each of the five scenarios, including 69, 69-161, and 161 kV cases, a three-phase short-circuit fault at a point of common coupling (PCC) was simulated as a worst-case scenario to determine an appropriate LVRT curve. The CCT emerged as a pivotal parameter in the LVRT specification process, which also considers additional factors such as transmission line voltage, voltage sag, and the duration and amplitude of fault recovery oscillations following the sag.