Share this page on:

Conference programme 

Back to the programme printer.gif Print

Poster session

Lead Session Chair:
Stephan Barth, Managing Director, ForWind - Center for Wind Energy Research, Germany
Ioannis Stylianou University of Manchester, United Kingdom
Ioannis Stylianou (1) F P Gabriel Cuevas Figueroa (1) Tim Stallard (1)
(1) The University of Manchester, Manchester, United Kingdom

Printer friendly version: printer.gif Print

Download poster(0.76 MB)

Presenter's biography

Biographies are supplied directly by presenters at OFFSHORE 2015 and are published here unedited

Mr. Stylianou has recently completed his MSc degree in Renewable Energy and Clean Technology at the University of Manchester and he is currently searching for a job within the wind energy industry. He has also obtained an MEng degree in Electrical and Computer Engineering from the National Technical University of Athens.


Prediction of wind farm energy yield using NWP considering within-cell wake losses


Numerical Weather Prediction (NWP) Models such as Weather Research and Forecasting (WRF) are widely used for predicting the wind resource at potential wind farm deployment sites and, increasingly, for energy yield prediction. Subgrid models have previously been developed to represent wind farms by modification of momentum sink and turbulence kinetic energy source terms within cells occupied by wind turbines. In this study, a group of turbines are parameterised by thrust and power determined using semi-empirical wake models to assess influence of within-cell wake losses on net yield.


Variation of thrust and power with wind speed and direction was obtained for groups of turbines using the modified PARK and Eddy Viscosity methods in OpenWind. Sensitivity to turbine number and spacing relative to the cell were determined. The influence of such wake-losses on yield was evaluated by comparison of energy yield from power curve and predicted wind speed, from use of a standard turbine representation within WRF and from a modified parameterisation to represent wake losses. The case study is based on the Horns Rev farm for time intervals selected to represent the annual wind speed distribution.

Main body of abstract

Atmospheric models such as WRF solve a reduced form of the Navier-Stokes equations with typical resolution of 20-1000 m in the vertical axis and 1-2 km in the horizontal axes. Fitch et al. (Monthly Weather Review 140, 2012) represent wind turbines by a streamwise momentum sink and a Turbulent Kinetic Energy (TKE) source defined as a function of turbine thrust and extracted power, both of which vary with incident flow velocity. Net thrust from multiple turbines within a cell can be estimated by linear superposition although this neglects wake losses between turbines located within a single cell. These losses are estimated using alternative semi-empirical wake models and were found to reduce thrust by up to 8.5% and power output by up to 12.5% for groups of four turbines. The influence of these losses on energy yield has been assessed by considering the Horns Rev wind farm.
An aggregate power curve for the Horns Rev farm combined with wind data from the ERA-Interim dataset provided yield for 2007 to within 16% of published yield. Time varying power output from this standard approach was compared to power predictions for the same boundary conditions and farm dimensions using WRF, including wind turbines using the Fitch scheme and with modified values of momentum sink and TKE to account for within-cell wake losses. A 1.3 km horizontal grid was used with vertical spacing of 15 m over the turbine diameter. The reduction of yield due to these wake losses varies with wind speed and direction.


The influence of wake losses between turbines located within a single cell of an atmospheric model has been studied. Thrust and power variation with wind speed and direction are reported for a range of flow speeds and directions using semi-empirical wake models. Maximum thrust and power for groups of four turbines were up to 10% and 8% lower, depending on wake model, than when neglecting wake losses and this leads to a reduced energy yield prediction when such groups are represented in the WRF model.

Learning objectives
Intended study outcomes: Understanding of the extent to which thrust and power output of small groups of turbines is reduced due to wake losses compared to superposition and the extent to which these losses, when used to represent multiple turbines within a NWP code, influence power production. Data presented concerning the accuracy of wind resource predictions is also expected to be of value for understanding the uncertainty associated with energy yield predictions using existing methodologies.