Sub-seasonal to seasonal climate predictions for wind energy forecasting

A Soret,V Torralba,Ll Palma,Ll Lledó,A Manrique-Suñén,N González-Reviriego,F J Doblas-Reyes,N Cortesi,I Christel

doi:10.1088/1742-6596/1222/1/012009

Abstract

Both renewable energy supply and electricity demand are strongly influenced by meteorological conditions and their evolution over time in terms of climate variability and climate change. However, knowledge of power output and demand forecasting beyond a few days remains poor. Current methodologies assume that long-term resource availability is constant, ignoring the fact that future wind resources could be significantly different from the past wind energy conditions. Such uncertainties create risks that affect investment in wind energy projects at the operational stage where energy yields affect cash flow and the balance of the grid. Here we assess whether sub-seasonal to seasonal climate predictions (S2S) can skilfully predict wind speed in Europe. To illustrate S2S potential applications, two periods with an unusual climate behaviour affecting the energy market will be presented. We find that wind speed forecasted using S2S exhibits predictability some weeks and months in advance in important regions for the energy sector such as the North Sea. If S2S are incorporated into planning activities for energy traders, energy producers, plant operators, plant investors, they could help improve management climate variability related risks.

Highlights

Understanding and quantifying climatic conditions from several weeks to months can improve the decision making of renewable energy generation and electricity demand (Figure 1)
We find that wind speed forecasted using S2S exhibits predictability some weeks and months in advance in important regions for the energy sector such as the North Sea
If S2S are incorporated into planning activities for energy traders, energy producers, plant operators, plant investors, they could help improve management climate variability related risks

Summary

Introduction

Understanding and quantifying climatic conditions from several weeks to months can improve the decision making of renewable energy generation and electricity demand (Figure 1) Climate predictions including both sub-seasonal (up to one month) and seasonal predictions (forthcoming months) have witnessed considerable improvements in the last decade demonstrating that probabilistic forecasting can inform better decision making for some forecast windows and regions [1]. Despite this improvement, climate predictions come with new set of challenges for users: information is often un-tailored and knowledge of power output forecasting beyond a few days remains poor. The work done in two European projects: S2S4E (s2s4e.eu/) and NEWA (neweuropeanwindatlas.eu/) will be presented in this contribution

Objectives

Results

Conclusion