Operational risk quantification of power grids using graph neural network surrogates of the DC OPF

TL;DR

This work addresses the computational bottleneck of Monte Carlo risk quantification in power grid operation by introducing graph neural network surrogates for DC OPF. The authors train bus-, branch-, and system-level GNN surrogates to rapidly predict key QoIs and integrate them into a reliability and risk framework that accounts for realistic, joint distributions of load and renewable generation via Gaussian copulas. Results on four benchmark grids show that GNN surrogates achieve high accuracy (bus/branch/system QoIs) with speedups exceeding two to three orders of magnitude, while producing risk estimates for reserve adequacy and branch overloading that closely match OPF-based ground truth. The approach supports real-time risk assessment under uncertainty and can be extended to temporal dynamics and reinforcement learning to further enhance operational decision-making.

Abstract

A DC OPF surrogate modeling framework is developed for Monte Carlo (MC) sampling-based risk quantification in power grid operation. MC simulation necessitates solving a large number of DC OPF problems corresponding to the samples of stochastic grid variables (power demand and renewable generation), which is computationally prohibitive. Computationally inexpensive surrogates of OPF provide an attractive alternative for expedited MC simulation. Graph neural network (GNN) surrogates of DC OPF, which are especially suitable to graph-structured data, are employed in this work. Previously developed DC OPF surrogate models have focused on accurate operational decision-making and not on risk quantification. Here, risk quantification-specific aspects of DC OPF surrogate evaluation is the main focus. To this end, the proposed GNN surrogates are evaluated using realistic joint probability distributions, quantification of their risk estimation accuracy, and investigation of their generalizability. Four synthetic grids (Case118, Case300, Case1354pegase, and Case2848rte) are used for surrogate model performance evaluation. It is shown that the GNN surrogates are sufficiently accurate for predicting the (bus-level, branch-level and system-level) grid state and enable fast as well as accurate operational risk quantification for power grids. The article thus develops tools for fast reliability and risk quantification in real-world power grids using GNN-based surrogates.

Figures (6)

• Figure 1: Structure of the three GNN surrogates.
• Figure 2: Schematic of test Dataset II generation and reliability and risk assessment
• Figure 3: Reliability assessment at branch level, twenty critical branches are selected to evaluate the GNN surrogate accuracy. (a)-(d): Case118, Case300, Case1354 and Case2848, respectively.
• Figure 4: Conditional probability (Eq. \ref{['eqn:multiple_branches_reliability']}) of branch overloading (Case118). Top: OPF, bottom: GNN.
• Figure 5: Risk assessment at branch level, twenty critical branches are selected to evaluate the GNN surrogate accuracy. (a)-(d): Case118, Case300, Case1354 and Case2848, respectively.