Siemens Energy and Lawrence Livermore National Laboratory are to conduct atmospheric modelling research that is expected to help operaters and owners manage wind farms more efficiently.
Siemens Energy and LawrenceLivermore National Laboratory (LLNL) are to conduct atmospheric modelling research that is expected to help operaters and owners manage wind farms more efficiently.
Under the two-year cooperative research-and-development agreement (CRADA), Livermore, California-based LLNL will provide high-resolution, numerical weather prediction models to forecast power generated by the wind. Siemens will translate Livermore’s forecasts of wind speed and wind direction at each turbine into power collected.
According to Siemens, many US wind parks yield up to 20 per cent less energy than predicted because of uncertain forecasts. More accurate wind predictions will enable wind-farm operators and owners to know hours or days ahead of time how wind conditions will affect power generation.
‘Accurate and timely forecasts of power availability will enable turbine owners and operators to generate optimal bids on wind turbine production and, in turn, maximise financial benefit and grid support,’ said Henrik Stiesdal, chief technology officer of Siemens Wind Power.
‘More accurate predictions also could reduce the investment risks in wind-powered projects and could improve the design of tall wind turbines to withstand the high-turbulence environment higher in the atmosphere,’ he added.
A recent study of 3,300MW of wind generation in New York state quantified improved forecasting to be worth $125m (£88m) a year to that region. Based on a conservative application of this figure, Stiesdal estimated that wind-farm owners may be able to increase revenue by as much as 10 per cent.
‘Knowing the certainty of the forecast can be useful in a day-ahead or futures market where now there are penalties for underperformance,’ said Julie Lundquist, the Livermore atmospheric scientist who is heading the project. ‘At LLNL, we have developed improved methods for simulating the turbulent properties of the lower atmosphere, which we think will translate into a significant predictive advantage for wind energy applications.’