CAREER: From Reactive to Proactive Distribution Grid Risk Management

Distribution utilities manage the integration of sustainable yet variable distributed energy resources, which makes it harder to maintain power quality. They also manage the impacts of frequent severe weather events causing sparks that ignite deadly and devastating fires. The current approach is to react to these problems after they arise because of the lack of methods that can proactively assess and mitigate risks.

This CAREER project will bridge this much-needed gap by developing risk-assessment and optimization methods that will bring transformative change to distribution grid operation by moving from reactive to proactive distribution grid risk management. The intellectual merit of the project lies in developing new risk assessment methods to quantify and mitigate risk due to load variability, wildfire ignitions and electric outages.

The broader impacts of the project include improvements in power quality, a reduction in the risk of wildfire ignitions and fewer power outages across the United States. The project will also improve undergraduate power system education through the use of open-source software, and contribute towards increased retention of undergraduate students by developing a learning community that promotes effective teaching practices among teaching assistants.

The goal of the project is to develop risk assessment methods to quantify short-term operational risk as well as formulations and solution algorithms for the associated risk-based and stochastic optimization methods. These data-driven techniques will consider multiple imminent threats including multiple scenarios for renewable energy generation and wildfire ignitions, and identify control actions (e.g., setpoints for distributed energy resources and remote switches) that improve security and reliability of distribution grid operations across all these scenarios.

Modeling distribution grids requires, e.g., the consideration of three-phase power flow calculations and binary decision variables to ensure radial topologies, thus increasing model complexity relative to transmission grids. Solving this complex, multi-scenario problem will require both new models and solution approaches for data-driven stochastic and risk-based optimization.

A main technical focus of the proposal is, therefore, to develop computationally tractable approaches to leverage large amounts of data within an optimization framework, and effectively interface simulations and optimization.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.