Search the EPA Research Database

Project title

Wind Energy Forecasting with an Application to Trading on the European Power Exchange

Primary Funding Agency

Irish Research Council

Co-Funding Organisation(s)

n/a

Lead Organisation

Galway-Mayo Institute of Technology (GMIT)

Lead Applicant

n/a

Project Abstract

High levels of wind energy forecast error result in significant balancing and environmental costs on our electricity networks. State of the art wind energy forecasting relies on the accuracy of numerical weather prediction systems from around the world. These ensemble predication systems are in themselves, multiple runs of mathematical models created from sources of observed meteorological data. Wind speed estimates from the ensemble prediction systems are the key input variable to a wind energy forecast model. As it is difficult to know exactly the state of the atmosphere at any one instance, forecasting what will happen over the next forty-eight hours from a system that is truly chaotic, is challenging. Using the output determined from multiple modelled instances of a chaotic phenomenon such as weather as the input to an algorithm which is relied upon to provide an energy forecast is inherently error driven. Our proposed research will seek to quantify the scale of the forecast error and investigate if the use of machine learning can significantly reduce this error. We will deploy various machine learning techniques which are suited to our datasets. A reduction in forecast error will result in lower levels of fossil based operating reserve on electricity networks with higher levels of renewable electricity penetration for environmental and societal benefit. We have identified three main areas in which to target error reduction methodologies (i) meteorological (ii) the wind farm (ii) the market. The proposed research is original in terms of its objective, and will we trust, contribute significantly to improvements in this field.

Grant Approved

�82,500

Research Hub

Climate related research

Research Theme

Achieving climate neutrality by 2050.

Start Date

01/09/2021

Initial Projected Completion Date

31/08/2024