Policy makers are tasked with selecting, designing, and implementing policies to support the transition to a sustainable power system. As part of the task they often turn to models to quantify and compare the options available to them. In this work we investigate the importance of representing a wide range of economic and physical sources of uncertainty in the modelling used to evaluate different decarbonization policies. We investigate six different energy policies across three different methods for incorporating uncertainty into decision making models in a Portugal based case study. We find that the method for incorporating uncertainty into the model used to evaluate policies leads to differences in the resulting capacity expansion plan, in the ability to meet carbon intensity targets, and in the abatement costs of policies. Policies designed in a deterministic way can result in significant violations of the emission target and the expected costs be more than double those estimated in a deterministic way. We also find that the six policies appear roughly equivalent when analysed by the deterministic model but perform very differently when uncertainty is considered potentially biasing decision makers that ignore uncertainty. Finally, we demonstrate that the simplified inclusion of uncertainty, such as scenario analysis, often underestimates the carbon reduction effect of policies and can over or under estimate costs.
Keywords generation expansion planning, electricity market modelling, renewable energy policy, decarbonization