Strategic planning for the low carbon energy transition is characterised by a high degree of uncertainty across many knowledge domains and by the high stakes involved in making decisions. Energy models can be used to assist decision makers in making robust choices that reflect the concerns of many interested stakeholders. Quantitative model insights alone, however, are insufficient as some dimensions of uncertainty can only be assessed via qualitative approaches. This includes the strength of the knowledge base underlying the models, and the biases and value-ladenness brought into the process based on the modelling choices made by users. To address this deficit in current modelling approaches in the UK context, we use the NUSAP (Numeral Unit Spread Assessment Pedigree) approach to qualify uncertainty in the energy system model, ESME. We find that a range of critical model assumptions that are highly influential on quantitative model results have weaknesses, or low pedigree scores, in aspects of the knowledge base that underpins them, and are subject to potential value-ladenness. In the case of the UK, this includes assumptions around CCS deployment and bioenergy resources, both of which are highly influential in driving model outcomes. These insights are not only crucial for improving the use of models in policy-making and providing a more comprehensive understanding of uncertainty in models, but also help to contextualise quantitative results, and identify priority future research areas for improving the knowledge base used in modelling. The NUSAP approach also promotes engagement across a broader set of stakeholders in the analytical process, and opens model assumptions up to closer scrutiny, thereby contributing to transparency.
Assessing qualitative and quantitative uncertainties in UK energy systems modelling
Pye, S., Li, F.G.N., Petersen, A., Broad, O., McDowall, W., Price, J., and Usher, W. (2018). Assessing qualitative and quantitative uncertainties in UK energy systems modelling. Energy Research & Social Science , 46: 332–344. doi: 10.1016/j.erss.2018.07.028