Statistical methodology for on-site wind resource and power potential assessment under current and future climate conditions: a case study of Suriname

Abstract

The feasibility and long term sustainability of proposed projects for wind energy production strongly depend on the availability of wind resources. This availability is assessed by means of local wind speed statistics. Under climate change, the future (long term) wind resource availability is, however, highly uncertain. This research proposes a methodology for on-site wind resource assessment in future (climate change) perspective based on stochastic modeling, observed data and the latest generation of climate model results for future projections. Statistical downscaling and bias correction methods, i.e., the Quantile Perturbation Method and Quantile Matching, are applied to enable local scale assessments. It requires observed data for extended periods to facilitate climate change signal assessments and associated future projections. Two types of data sets are considered for observed data, i.e., on-site (local) measurements and reanalysis data. Two stochastic modeling approaches were adopted for the local observations data extension, i.e., a Markov and Weibull model, allowing for a sensitivity assessment. The methodology has been applied to a potential wind site in Suriname. Results reveal significant changes in wind power potential for the end of the century (2070–2100), ranging from − 27 to 89%. Analysis of the extreme conditions reveals an extended range, from − 65 to 282%.