Estimating cascading failure risk: Comparing Monte Carlo sampling and Random Chemistry

This paper presents a computationally efficient approach to estimate cascading failure risk in power systems. The method uses the previously published Random Chemistry algorithm [1] to find combinations of branch outages that lead to large blackouts, and then estimates risk by computing the expected blackout size based on the probabilities of various contingencies. We compare this method with Monte Carlo simulation, and show that the method is at least an order of magnitude faster than Monte Carlo simulation. Results from the IEEE RTS-96 and the 2383-bus Polish grid are presented in the paper.