A Generalized Decomposition Framework for Large-Scale Transmission Expansion Planning

In this paper, we propose a scalable and configurable decomposition framework for solving large-scale transmission capacity expansion planning with security constraints under uncertainties. This framework is capable of using both progressive hedging and Benders decomposition algorithms to decompose and parallelize a large-scale problem both vertically and horizontally. A scenario bundling method is also developed to create bundles through three steps, i.e., classification, clustering, and grouping with the objective of maximizing similarity between bundles. This bundling method can improve both quality of results (decreasing optimality gap) and performance (reducing computational time) of the proposed framework. To verify capabilities of the proposed method, it is applied to a reduced ERCOT system with 3179 buses, 4458 branches, and 10 scenarios. The numerical result for this case study shows that the proposed framework can make solving large-scale problems tractable, and provides high quality results (with less than 1% optimality gap) in a reasonable time (around 2.8 days).