Back to the programme printer.gif Print




Delegates are invited to meet and discuss with the poster presenters in this topic directly after the session 'The model chain: First steps towards tomorrow's technology' taking place on Thursday, 13 March 2014 at 09:00-10:30. The meet-the-authors will take place in the poster area.

Erik Berge Kjeller Vindteknikk AS, Norway
Co-authors:
Erik Berge (1) F P Øyvind Byrkjedal (1) Rolf Erik Keck (2) Niklas Sondell (3) Rolv Erlend Bredesen (1)
(1) Kjeller Vindteknikk AS, , Norway (2) University of Oslo and Statkraft Development, Oslo, Norway (3) Statkraft Development, Stockholm, Sweden

Printer friendly version: printer.gif Print

Abstract

Properties of a wind farm wake as simulated by a numerical weather prediction model for the Sm�la wind farm

Introduction

Recently the numerical weather prediction model WRF has been extended to account for wind farms in the mesoscale calculations (Fitch et al 2012) by including the turbine drag coefficient. The wind farm will thus represent a sink of atmospheric momentum in the model. The energy extracted from the available kinetic energy of the atmosphere is divided into two parts. Firstly kinetic energy is converted to electric energy which is expressed by the power coefficient. Secondly, kinetic energy is converted to turbulence expressed as an increase in the TKE (turbulent kinetic energy).

Approach

In the work presented here the kinetic energy converted into electrical energy from the Fitch scheme is compared to power production data from the wind farm Smøla which consists of 68 turbines and is located in western Norway. The WRF model is run in the time domain for a period of 45 days from which we have concurrent data on power production from each turbine in the wind farm and from a meteorological mast located near the wind farm.

Main body of abstract

The Fitch scheme produces a deep wake that results in a wind speed deficit of 0.3-0.6 m/s reaching as far as 50 km downstream of the wind farm. The horizontal distribution of the wake is clearly dependent on the atmospheric stability; for the more unstable conditions the wind farm wake is found to reach shorter distances before it is dissipated, but it propagates more vertically. The flow is found to be distorted around the wind farm, typically an increase in the wind speed is evident on the left side of the wind farm facing the downstream direction. The analysis of the power production from the wind farm and data from the meteorological mast indicates that this effect can be observed for certain wind directions.

The wake loss for Smøla is underestimated by 36 % by the Fitch scheme during the analyzed time period, while the total power production from is underestimated by 20 %.

Conclusion

The biases found are partly caused by the underestimation of the local wind speed in the model compared to the measured wind speed from the meteorological mast. However, also the power coefficient given in the Fitch scheme represents a turbine with a less optimal power curve than the turbines at Smøla. This also contributes to the underestimation of the power production from the Fitch scheme.


Learning objectives
The Fitch module can be a useful tool to study the meso scale effects of a wind farm. The scheme is implemented in a meso scale model and as such the two way interaction between the wind farm and the atmosphere will be described. The model will also account for meso scale variation within a wind farm.