Mr Will Usher, UCL Energy Institute
The rapidly maturing field of energy system modelling integrates disciplines that operate under fundamentally different assumptions about uncertainty. With some notable exceptions, these uncertainties have not been given the attention they require, due to the undesirable burden of complexity they introduce. This is problematic as energy system models are constructed to support complex decision making under uncertainty. Increasingly, the uncertainties across these disciplines appear intractable, especially given the undue attention by the media and partisan ‘think- tanks’ that can undermine public confidence in both science and policy. But there are plenty of techniques available to manage uncertainty, including sensitivity analysis and uncertainty analysis. These techniques can increase transparency, highlight areas of models and research that are lacking, and provide insights that are resilient to changes in assumptions.
Prior to a full scale uncertainty analysis, it is essential to deconstruct the way in which uncertainty is understood within each discipline. For example, an economist’s understanding of uncertainty of economic growth (a driver of energy demand) within the complex multi-layered economic system is very different to that of an engineer who manages the uncertain failure of components (a driver of technology cost).
We establish an uncertainty framework in which we analyse a subset of uncertainties relevant to energy system modelling. For example, we differentiate between those uncertainties related to measurement error, inherent randomness and those due to lack of knowledge and discuss the differences. We highlight those uncertainties of interest to policy makers and modellers, and show how insights from the resulting studies can be used as the foundation for a modelling exercise that incorporates uncertainty.