Finding Dunkelflaute - estimating the probability of successive days of low Variable Renewable Energy (VRE)
Conference
Format: IPS Abstract
Keywords: #energy, renewable resource, time-series
Session: Invited Session 2A - Tools for the Energy Transition
Monday 2 December 1:30 p.m. - 3 p.m. (Australia/Adelaide)
Abstract
Variable Renewable Energy in the form of wind and solar farms plus rooftop solar accounts for more than 70% of energy to the electricity grid in South Australia. The goal is to have it reach 100% by 2027. That would not be all the electricity at every time of the year, but the total supplied to the grid over the year would equal the total demand. In times of excess, it will be supplied via interconnector to other states, and the flow would go in the opposite direction in times of insufficient supply. In order to progress to firm power from VRE, one needs to understand the characteristics of the lack of supply. This study estimates the probabilities of a number of days in succession with supply from VRE under a specified threshold. This type of calculation will inform decision makers on the types of methods to ensure supply in these periods. For instance, is only short duration storage needed, or is there a need for longer duration storage, as in pumped hydro or more esoteric solutions. To perform this risk analysis, given that the time series of output from sources of VRE are short, techniques such as bootstrapping are employed to construct synthetic series that are statistically indistinguishable from the original series. These synthetic sets will undoubtedly contain sequences not present in the historical sets, aiding the determination of risk.