Skip to main content

Dynamic loads in wind farms

Objective

To investigate the dynamic loads experienced by wind turbines operating in wind farms through the use of full scale measurements, wind tunnel model measurements and computational methods. The results of the study will allow more accurate life time prediction and better farm lay-out.
Extensive work regarding the prediction of the response of the wind turbine operating in free stream conditions has taken place. The sophistication of the models has been increased until a high degree agreement has been achieved. This has required detailed modelling of the rotor, control system, drive train and tower.

The data base of continuous power measurements and extensive detailed wind speed records are being used to assess the ability to predict the changes to mean and turbulent properties of the wind within a wind farm.

The task of predicting the machine within a wind farm is now underway. Work to date has demonstrated that it is possible, with present analytical techniques, to make quite reasonable estimates of the behaviour of the dynamically active machine operating in a wind farm.
The level of understanding of the loads experienced by isolated wind turbines is now quite high. It is possible to predict their dynamic response to real winds and obtain good predictions of the fatigue and extreme loads. As the understanding of the machine response has improved so has the ability to predict the flow conditions inside a wind farm. These techniques are less mature than the load prediction techniques and certainly much less well validated. Large scale exploitation of wind energy will be through the use of wind farms and hence it is vital that the behaviour of machines in farms is properly understood.

In many parts of Europe the wind farms will be built in complex terrain which gives rise to difficulties simply in the prediction of the flow, even in the absence of a wind farm. In order to isolate the machine interaction from the topographical effects this project is limited to flat terrain.

The majority of the work will concentrate on the Sexberium wind farm in the Netherlands. This farm is equipped wit very comprehensive instrumentation, both to measure the wind and to measure the loads in some of the machines. The investigation will be undertaken using three separate approaches :

. Direct, full scale measurements;
. Physical modelling of the Sexberium lay-out in a large wind tunnel using miniature wind turbine models;
. Computational modelling of the flow and machine response to predict both the model and full scale behaviour.

After completion of the main body of the work using the Sexberium farm the tools developed will be used to predict similar quantities in the Norrekaer Enge farm in Denmark. The results of this phase will be an assessment of the validity of the methods when applied to another, significantly different, machine in flat terrain.

This project will allow an accurate assessment to be made of the maturity of wind farm prediction tools and also produce a better understanding of the physics of the flow and machine response in wind farms. Both results will permit improved machine design and farm lay-out.

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

GARRAD HASSAN AND PARTNERS LTD
Address
Saint Vincent Works, Silverthorne Lane
BS2 0QD Bristol
United Kingdom

Participants (3)

Business Unit of TNO Built Environment and Geosciences
Netherlands
Address
Van Mourik Broekmanweg 6
2628 XE Delft
ENERGY RESEARCH CENTRE OF THE NETHERLANDS
Netherlands
Address
Westerduinweg 3
Petten
KEMA NEDERLAND BV
Netherlands
Address
Utrechtseweg 310
9035 Arnhem