Search

Background: Energy system optimisation models (ESOMs) are commonly used to support long-term planning at national, regional, or continental scales. The importance of recognising uncertainty in energy system modelling is regularly commented on but there is little practical guidance on how to best incorporate existing techniques, such as global sensitivity analysis, despite some good applications in the literature. Methods: In this paper, we provide comprehensive guidelines for conducting a global sensitivity analysis of an ESOM, aiming to remove barriers to adopting this approach. With a pedagogical intent, we begin by exploring why you should conduct a global sensitivity analysis. We then describe how to implement a global sensitivity analysis using the Morris method in an ESOM using a sequence of simple illustrative models built using the Open Source energy Modelling System (OSeMOSYS) framework, followed by a realistic example. Results: Results show that the global sensitivity analysis identifies influential parameters that drive results in the simple and realistic models, and identifies uninfluential parameters which can be ignored or fixed. We show that global sensitivity analysis can be applied to ESOMs with relative ease using freely available open-source tools. The results replicate the findings of best-practice studies from the field demonstrating the importance of including all parameters in the analysis and avoiding a narrow focus on particular parameters such as technology costs. Conclusions: The results highlight the benefits of performing a global sensitivity analysis for the design of energy system optimisation scenarios. We discuss how the results can be interpreted and used to enhance the transparency and rigour of energy system modelling studies., QC 20231115

Data handling for energy system optimisation models is a necessary but tedious task. Depending on the workflow, user skill level, and model implementation, the data interfacing requirements can be significantly different. OSeMOSYS Tools for Energy Work, or otoole, is a Python package providing OSeMOSYS energy modellers with options to use different input dataformats, visualize and validate input data, and process result data. otoole exposes three different input data formats to the user, can process result data from four popular solvers, and is designed to be modular and extensible to enable interoperability between OSeMOSYS models developed in different programming languages., QC 20240909