Wind Forecasting with Super-Duper Computer

David Appleyard, Chief Editor, Renewable Energy World
August 01, 2013 | 2 Comments

LONDON -- New software has been developed which has been designed to improve current studies for estimating onshore and offshore wind production, and which may be applied throughout the whole operational life of such facilities, its developers say.

Iberdrola and the Barcelona Supercomputing Center – Centro Nacional de Supercomputación (BSC-CNS), with the collaboration of the National Centre for Renewable Energies (CENER), aim to create an innovative information technology model with the research & development project.

Both the development of the project and its subsequent application will take place in the facilities of BSC, using software run by MareNostrum, one of the top 30 fastest supercomputers in the world.

MareNostrum has 48,896 Intel processors and a calculation capacity of 94.21 Teraflops - a measure of computing speed equal to one trillion floating-point operations per second.

By more accurately forecasting yields, the Sedar Project (High Resolution Wind Simulation) is expected to pave the way fo the construction of wind farms with greater guarantees on investment, its developers say.

Current models have a significant limitation in their calculation times, and the project has overcome this shortcoming through the use of a supercomputer. Furthermore, its backers claim the project will develop improvements for the resolution of physical models which, so far, have also been constrained by computing times.

In the initial stage, already being tested, the project will design new wind farms, providing the most suitable locations for installing machines and reducing the uncertainty involved in investing in this type of study. This phase of the project, which uses Alya Green software, can be applied in future onshore and offshore wind farms.

Both organisations have agreed to carry out the development of the model and its subsequent deployment at Iberdrola's new wind projects at facilities of the BSC.

Lead image: Graphic from the wind forecasting model, courtesy Iberdrola

Read more wind power news here.

Lead image: Wind vane via Shutterstock

2 Comments

ANONYMOUS
2013-08-09 16:47:21.0

Garbage in, garbage out, but faster, much faster.

There is no way to predict wind power 24 hours in advance with sufficient accuracy to make a difference.

Meteorology is and art, not a science!

Gary Tulie
2013-08-08 06:21:13.0

This should help to pin down the expected yields of wind farms over their lifetimes assuming climate change does not have a significant impact on wind speeds.

The other important forecasting job which would help wind farms is to predict wind yield accurately 1, 4, 12, and 24 hours in advance.

This will allow wind farms to maximize their revenue by selling with greater confidence into the forward spot market i.e. 4 to 24 hours ahead (with less costs incurred when they have to buy or sell unscheduled power (regulation services) from other suppliers to deliver on their forward promise).

Add Your Comments

To add your comments you must sign-in or create a free account.

David Appleyard

David Appleyard is a contributing editor. A freelance journalist and photographer, he has some 20 years' experience of writing about the renewable energy sector and is based in Europe.