Posters

Abstract

In recent years, there has been an increasing interest in having Wind Resource- (WRA) and Energy Yield Assessments (EYA) discuss and account for the potential impacts of climate change on a project's energy production over the wind farm lifetime. In practice, there are many climate models which can be used for this mitigation effort. However, the sheer number of models induces a very large variance on the overall projections of a site in question, and consequently large uncertainty on how to incorporate them into an EYA. In this work, we present a methodology to select those among the large number of climate models which are best suited to describing the conditions at the target wind farm using a quantitative and site-specific set of criteria. The approach uses surface wind speed time series from Phase 6 of the Coupled Model Intercomparison Project (CMIP6). The relevant models are at first filtered based on geographical availability at- or near the target site, the specific Shared Socioeconomic PathwaySSPX-Y scenarios they modelled, and the availability of a historical scenario (ie. prior to 2014). Then, the remaining climate models are evaluated using a series of KPIs related to their comparison against available- and well-validated datasets in the region, such as reanalysis or measurements. The KPIs are related to quantities such as interannual variability of the mean wind speed, presence of long-term trends, among others. A subset of the available models is selected based on the aforementioned evaluation. After proper selection of the relevant climate models and with views to use in WRA and EYA, this work further emphasizes that the goal is not necessarily to prescribe a future mean wind speed gradient or a change in the overall AEP for a given site. Instead, our work highlights the value of the climate models in identifying- and quantifying risks and sensitivities directly related to EYA uncertainty such as potential increases in interannual variability, change in the shape of wind speed- and direction distributions, occurrence frequency of extreme events- or events outside the operational range of typical WTG power curves, etc. In this work we present our methodology along with three test cases: one in the North Sea, one in the Eastern United States, and one in East Asia; where all are offshore wind farm sites.