Two-Stage Stochastic Resilience Optimization of Converter Stations Under Uncertain Mainshock-Aftershock Sequences

Reducing vulnerability and enhancing resilience of power systems, particularly converter stations, against earthquakes are crucial for maintaining their safe operation. Nevertheless, earthquakes are often accompanied by aftershocks and encompass plenty of uncertainties, posing significant challenges for the development of a joint pre-earthquake preparation and post-earthquake restoration strategy. In this article, a novel two-stage stochastic programming model is put forth to bolster the resilience of converter stations under uncertainties associated with mainshock-aftershock sequences. The model is bifurcated into two pillars: the first stage focuses on designing equipment hardening strategy (EHS) and spare parts strategy (SPS) for converter stations before earthquakes. The second stage is dedicated to the optimization of recovery scheduling (RS) in the wake of mainshock-aftershock sequences. To mitigate the endogenous uncertainties of the model, a tolerated random number generation method is employed to generate deterministic scenarios in the second stage. The resulting optimization model is, then, efficiently solved using a combination of the sample average approximation and progressive hedging algorithm. To evaluate the performance of the proposed resilience optimization method, comparative studies are carried out on a real ±800 kV converter station located in the city of Yibin, Sichuan Province, P.R. China. Results from this application demonstrate that the EHS pre-dominantly focuses on hardening the equipment of the same circuit to guarantee the low transmission capacity under diverse earthquake scenarios, while the SPS is proven to be more effective than the EHS in scenarios with high values of peak ground acceleration (PGA) of mainshock and occurrence intensity of aftershocks.