Risk-constrained strategic resilience enhancement of power distribution and natural gas systems through G2P and P2G planning and distributed load Restoration against flood

Abstract

Floods, categorized as high-impact low-probability (HILP) events, pose significant risks to power distribution systems (PDS) and natural gas systems (NGS). Ensuring mutual support during emergencies through energy sharing, facilitated by coupling devices, highlights the necessity for optimal integration. However, managing these systems separately by different operators calls for coordinated yet decentralized operation. Hence, this paper introduces a risk-constrained bi-level model to integrate PDS and NGS. The upper level deals with siting and sizing power to gas (P2G) and gas to power (G2P) units. Meanwhile, the lower level employs decentralized load restoration using the alternating direction method of multipliers (ADMM) algorithms, with each system individually minimizing its own risk. To account for the unpredictable nature of HILP events surpassing a predefined risk threshold (α), this study defines Value-at-Risk (VaR) and Conditional Value-at-Risk (CVaR) for each system as resilience criteria. Validation of the model is demonstrated using an IEEE 33-bus PDS and a 20-node NGS, and applied to a real-world large-scale PDS in Khuzestan province, Iran. Across various scenarios, higher system risks prompt an increase in the number and size of planned coupling devices, particularly to address the most severe 5% of scenarios. This optimization leads to enhanced energy sharing, resulting in a resilience improvement of nearly 35% in specific cases, along with a notable 25% reduction in both active and reactive power loss. Additionally, the sensitivity analysis explores the impact of varying risk levels and thresholds on outcomes, demonstrating how investors can utilize the proposed model across diverse risk perspectives.