Solving a large energy system optimization model using an open-source solver. (November 2021)

• Record Type: Journal Article
• Authors: Macmillan, Madeline; Eurek, Kelly; Cole, Wesley; Bazilian, Morgan D.
• Is Part Of: Energy strategy reviews. Volume 38(2021)
• Publication Date: 2021-11
• Subjects: Open-source solvers -- Energy models -- Solve time -- ReEDS; CLP COIN-OR linear programming -- ReEDS Regional Energy Deployment System -- NREL National Renewable Energy Laboratory -- NEOS Network-Enabled Optimization Server -- GAMS General Algebraic Modeling System -- LP linear program -- ERCOT Electric Reliability Council of Texas -- MPS mathematical programming system -- BDMLP Brook, Drud, and Meeraus Linear Program solver -- GLPK GNU Linear Programming Kit -- IPOPT Interior Point Optimizer -- MINOS Modular In-core Nonlinear Optimization System -- PIPS Parallel Interior Points Solver -- SOPLEX Sequential Object-oriented simPlex; Energy policy -- Periodicals; 333.7905
• DOI: 10.1016/j.esr.2021.100755
• Languages: English
• ISSNs: 2211-467X

Abstract

Abstract: Open-source energy models are becoming more widely used for electric power systems planning. The solutions for these models are often computed using commercial optimization solvers, which require licensing fees that can be a potential barrier for certain organizations and researchers. This study explores the ability of the open-source COIN-OR linear programming (CLP) solver to compute solutions for the Regional Energy Deployment System (ReEDS) model—a large-scale, open-access electricity system planning model for the United States developed by the National Renewable Energy Laboratory (NREL). We find that open-source solvers, such as CLP, require some reduction of model size and detail. Although the solutions for reduced-form models may differ from full-featured models, we demonstrate that reduced-form solutions for ReEDS can still provide useful insight about drivers of power sector evolution. This research can help the modeling community better understand how open-source solvers can be applied to large-scale planning tools, and the potential steps that may be required to implement them. Highlights: Evaluate the ability of open-source solvers to solve large electricity models. Summarize techniques to improve the solvability of open-source energy models. Analyze the trade-offs of model detail, solve time, and modeling outcomes.