Probability Prediction Method for Assessing Cascading Faults in Power Systems during Magnetic Storms

When a strong magnetic storm occurs, the reactive power compensation device becomes overloaded, resulting in reduced transformer load capacity and bus voltage drop. These conditions can trigger cascading failures, potentially leading to a complete collapse of the power grid. This paper introduces a probability prediction method for assessing cascading faults in power systems during magnetic storms. To address this issue, our proposed method combines the calculation of Geomagnetically Induced Current and transformer reactive power loss, considering various intensities of geomagnetic storm events. We also incorporate the transformer load capacity model under Geomagnetically Induced Current influence, along with the probabilities of transformer aging failure and accidental shutdown faults. By obtaining the initial fault probability of the transformer, we employ the Matlab and PowerWorld Simulator platforms to analyze the probability of state transitions across all levels of faults within the system. Furthermore, we utilize a Markov process to simulate the system's response in extreme scenarios, enabling the prediction of cascading fault development trends. Finally, we validate our development model and method using the GIC Benchmark model, demonstrating their feasibility and effectiveness. This approach aims to prevent the occurrence of cascading faults, reduce system risks, and enhance the power supply capacity of the overall power system.