A Markov Decision Process to Enhance Power System Operation Resilience during Wildfires

Abstract

The catastrophic impacts of wildfires on the performance of power grids have increased in the recent years. Various resilience enhancement methods have been applied against severe weather events, however only a few have focused on wildfires. Most previous operational-based enhancement methods focus on corrective or restorative strategy during and after extreme events without proactively preparing the system for forecasted potential failures. Also, the propagation behavior of wildfire among various system components induces further complexities resulting in a mathematically involved problem accompanied with many modeling challenges. This paper proposes a probabilistic proactive generation redispatch strategy to enhance the operational resilience of power grids during wildfires. A Markov decision process is used to model system state transitions and to provide generation redispatch strategies for each possible system state considering component failure probabilities, wildfire spatiotemporal properties, and load variations. For realistic system representation, various system constraints are considered including ramping rates and minimum up/down times of generating units, load demand profile, and transmission line constraints. The IBM ILOG CPLEX Optimization Studio is utilized to solve the optimization problem. The IEEE 30-bus system is used to validate the proposed strategy under various impact scenarios. The results demonstrate the effectiveness of the proposed method in enhancing the resilience level of power grids during wildfires.