Harder, better, faster, stronger: understanding and improving the tractability of large energy system models

TL;DR

This paper investigates the computational challenges of large linear programming energy system models, exploring scaling, formulation tweaks, and solver preferences to enhance tractability and runtime efficiency.

Contribution

It introduces an automatic model scaling method and evaluates various solver configurations, providing practical recommendations for energy system modelers.

Findings

Scaling improves numerical stability and reduces runtime

Barrier method without crossover can be faster but alters solutions

Model formulation tweaks significantly impact computational performance

Abstract

Energy system models based on linear programming have been growing in size with the increasing need to model renewables with high spatial and temporal detail. Larger models lead to high computational requirements. Furthermore, seemingly small changes in a model can lead to drastic differences in runtime. Here, we investigate measures to address this issue. We review the mathematical structure of a typical energy system model, and discuss issues of sparsity, degeneracy and large numerical range. We introduce and test a method to automatically scale models to improve numerical range. We test this method as well as tweaks to model formulation and solver preferences, finding that adjustments can have a substantial impact on runtime. In particular, the barrier method without crossover can be very fast, but affects the structure of the resulting optimal solution. We conclude with a range of recommendations for energy system modellers.