Predicting offshore wind energy resources

Project Information

Grant agreement ID: JOR3980286

Start date 1 August 1998

End date 31 July 2001

Funded under

• Specific programme for research and technological development, including demonstration in the field of non-nuclear energy, 1994-1998

Total cost € 0,00

EU contribution No data

Coordinated by

COUNCIL FOR THE CENTRAL LABORATORY OF THE RESEARCH COUNCILS
United Kingdom

Deliverables

Predicting offshore wind energy resources

In the coming years, exploitation of offshore wind energy is set to play a central role in Europe's overall energy strategy by assisting EU member state governments to achieve their national greenhouse gas emission reduction targets (both now and in the future), whilst continuing to meet the demand for energy. However, development and integration of offshore wind energy is currently handicapped by significant knowledge gaps, including a scarcity of good quality information on the extent, characteristics and distribution of the offshore wind energy resource. The objective of the Predicting Offshore Wind Energy Resources (POWER) project was to improve the understanding of the nature and distribution of Europe's offshore wind resource. In particular the project team set out to improve upon previous estimates of the European offshore wind energy resource, to consider a number of additional factors that could affect its exploitation on a commercial basis and to present the information in a straightforward, yet useful format. Within POWER, a novel wind resource assessment methodology was developed which can produce long-term and spatially detailed estimates of the wind conditions at offshore sites covering a wide area. Furthermore, the team applied this methodology to the region 30ºN to 70ºN and 15ºW to 30ºE on a grid of 0.5º resolution, an area which covers the major sea areas bordering EU countries - the North Sea, the Baltic, the Mediterranean and the eastern North Atlantic. The POWER project has produced state-of-the-art estimates of the extent and distribution of Europe's offshore wind energy resources not only in the coastal zone - the current focus of the offshore wind industry's attention - but also throughout the region's far offshore areas, where there is potential for wind energy to be exploited in the longer-term. On a local scale, POWER provides detailed first estimates of the long-term environmental conditions at specific offshore locations. This information is useful to the offshore wind energy industry since this is the exact type of data required for initial scoping and feasibility studies for new offshore wind energy developments. It may be possible to base preliminary assessments of the turbine power output as well as other key parameters such as initial values of the design parameters for turbine support structures from the POWER results. The data on wind and wave parameters for European waters produced by the POWER project has been compiled as a set of Microsoft Excel workbooks. The data can be accessed using the ''POWER tool'' - a simple graphical user interface (GUI) that allows the user to display, in both numerical and graphical form, data from the database of wind and wave parameters. The POWER project’s techniques should enable the wind energy industry to exploit the offshore wind energy resource with greater confidence, and hence facilitate a future expansion of the wind turbine manufacturing and installation industry - with the consequent employment opportunities.

Map of sea surface roughness for Europe

One of the project’s objectives was to assess the effect of variable sea surface roughness (z0 ) on offshore wind speed predictions. Values of sea surface roughness were calculated throughout the entire power project area (30ON to 70ON and 15OW to 30OE) using a new method suggested by Taylor and Yelland (2000) for the period January 1987 to December 1996, and detailed maps of the means values created. The work indicated that at some locations (for example, the Atlantic margin and North sea area) there were relatively high values of mean sea surface roughness. However, subsequent work in the project indicated that the variability of 20 found would result in only small differences (<0.5%) in predicted wind speeds (compared to those obtained using existing a constant value of z0).

Methodology to compute confidence limits for POWER wind speeds

A bootstrapping estimation method was used to gauge the reliability of the POWER wind speed estimates by calculating 95% confidence limits in the mean monthly wind speed estimates. Bootstrapping uses the available data to create a many additional data sets using re-sampling with replacement. Using the Weibull parameters for each directional bin generated using WAsP, twelve sets of simulated wind speeds were created, one for each directional bin, with 100 simulated wind speeds per bin. The analyses were carried out under MatLab. The simulated data were used to calculate the monthly mean wind speed for that grid square for each direction bin. To arrive at an overall mean, the bin means were weighted according to the frequency of wind speeds associated with each direction, and then summed. The 95% confidence limits for the overall mean were derived by applying the frequency weighting procedure to the confidence limits for each directional bin. Example results have been produced for six grid squares, for the 95% confidence limits and for a single height (90m a.s.l.) selected to be a typical hub height for offshore wind turbines. Note, however, that the methodology can be readily applied to data for any height at any location.

Dataset of annual, seasonal and monthly mean geostrophic wind speeds over EU waters at resolution of 0.5o latitude by 0.5o longitude

Geostrophic winds are theoretical winds that flow parallel to isobars (contours of equal pressure). They are a good approximation to the actual wind in the free atmosphere. They can be calculated from atmospheric pressure data. This is significant because while measured wind data at offshore locations are sparse and of variable quality, there are good quality data sets of atmospheric pressure available that cover offshore areas. These datasets can be used to construct geostrophic wind conditions, which in turn can be used to calculate near-surface winds. Atmospheric pressure data at mean sea level were obtained from the US National Centers for Environmental Prediction (NCEP) for the period 1985-97, at 6-hourly intervals and on a 2.5º latitude by 2.5º longitude grid. These data were interpolated onto a 0.5º latitude by 0.5º longitude grid using bi-cubic spline interpolation. The interpolated atmospheric pressure data were then used to calculate the sea level pressure gradients in the westerly and southerly directions at each point in the 0.5º by 0.5º latitude/longitude grid. Finally, the pressure gradient at each grid point was used to calculate the geostrophic wind speed and direction for each grid point and time step. The results were validated using radiosonde data.

Seasonal and annual predicted wind speeds for offshore Europe at heights between 10 and 150 m

Mean wind conditions for the period 1985-97 were estimated using WAsP at eight hub heights at each POWER grid point over the sea. The hub height levels (10m, 30m, 50m, 70m, 90m, 110m, 130m and 150m above mean sea level respectively) were chosen to cover the range of expected hub heights of wind turbines that are likely to be sited offshore in the coming years. In addition, the monthly and inter-annual variability of the wind conditions in European waters were also investigated by performing WAsP model runs estimating the mean monthly and mean yearly wind conditions at all offshore grid points. Note: Whilst maps showing the data have been published, the consortium do NOT plan to distribute the raw data for each grid point (i.e. at each 0.5O x 0.5O latitude/longitude grid point in the area 30ON by 15OW to 30OE) at the present time.

Analysis of wind speeds at various temporal scales: Long-term (1871 to present) geostrophic wind speed at 10° longitude by 5° latitude resolution Medium-term: Seasonal and diurnal cycle Short-term: gu

Long-term variability: Historical records of sea level pressure in Europe extend back to the 1870s. These were used to calculate the geostrophic wind speed. This was analysed for the offshore waters of Europe with respect to long-term trend. Medium-term variability: Using a data set of observations from coastal stations for a core 10-year period. The data were partitioned according to whether they approached from over the land or over the sea, and according to season. Then, the diurnal cycle was examined. It was found that there was only a small diurnal cycle in the winds from over the sea. Those that approached from over the land had a strong diurnal cycle that produced variations in wind speed greater than the differences between seasons. These analyses were augmented with information from the SODAR installed on the North Norfolk coast for two summer months, which allowed us to examine changes in the diurnal cycle with height. Short-term variability: Gust speeds were analysed with respect to the return periods of extremes using the Generalized Extreme Value distribution.