CAREER: Spatial Complex Network Analysis of Bulk Electric Grids for Long-Term System Planning

This NSF CAREER project aims to address computational and modeling challenges inherent in the design of large-scale electric transmission networks with large numbers of inverter-based resources. As increased electrification and the interconnection of renewables will require a major expansion of the electric transmission grid in the near future, this project will bring transformative change to power system planning with solutions to more intelligently select new transmission line projects for grid expansion. This will be achieved by an innovative methodology leveraging how patterns in the network structure of the grid predict its technical performance. The intellectual merits of this project include demonstrating the performance capability of a transmission expansion planning (TEP) solution framework for ultra-scale problems with algorithms to address both scenario variability and modeling complexity. The broader impacts include supporting electric grid design to reduce renewable curtailments and be more robust to blackouts, along with jumpstarting an open-access digital educational experience that puts students in the power grid operator’s seat to learn more about sustainability and resilience. The TEP problem has an inherent challenge due to the large combinatorial space of possible expansion options, intensified by the multi-faceted engineering assessment required to realistically evaluate the viability of even one potential solution. Hence very little application of integer programming methods can be seen in practical deployment. This project aims to develop a new framework, applying a paradigm of spatially-embedded complex networks, to approach ultra-scale TEP and overcome fundamental limitations in solver efficiency, breadth of scenario coverage, and depth of modeling. For solver efficiency, the project will develop a multi-layered solver architecture starting with spatially-aware candidate production, feeding to an iterative down-selection with an annealing-inspired metaheuristic and a final integer programming step with bounding that exploits spatial embedding. For scenario breadth, the project will compute grid community structure across thousands of uncertainty scenarios, leading to sensitivities that can inform a TEP solver to select lines that will reduce renewable curtailment throughout the year. Finally, depth of modeling will be addressed by connecting network structural properties with low-order models of voltage- and stability-based transmission limitations. 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.