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 'Optimising measurement strategies to maximise project value: Is the industry making false economies at the expense of project value?' taking place on Tuesday, 11 March 2014 at 11:15-12:45. The meet-the-authors will take place in the poster area.

Davide Medici DNV GL, Denmark
Co-authors:
Jean-François Corbett (1) F P Ulrik Horn (1) James Bleeg (1) Richard Whiting (1)
(1) DNV GL, Copenhagen, Denmark (2) DNV GL, Bristol, United Kingdom (3) DNV GL, Peterborough, United States of America

Printer friendly version: printer.gif Print

Presenter's biography

Biographies are supplied directly by presenters at EWEA 2014 and are published here unedited

Dr. Davide Medici has been working in the field of wind energy for more than 12 years,
with several published papers on turbine wake meandering and wakes in yaw.
He has a PhD in fluid mechanics at the Linnè Flow Centre, KTH Mechanics, Stockholm and
has joined GL Garrad Hassan in 2006. After several years as Team Leader managing a team
of six engineers, he is currently involved in R&D and CFD in several countries across Europe within
DNV GL. Dr. Medici is also supervising energy assessment analyses in Italy and
Middle East Countries as a senior engineer.

Abstract

A systematic validation of CFD flow modelling for commercial wind farms sites

Introduction

The ultimate link of the model chain is the microscale model, dealing with wind flow variations at the scale of the individual wind farm. Computational fluid dynamics (CFD) models have long promised to increase accuracy compared to traditional linear models, but proof in the form of large-scale validation has been wanting. With few exceptions, published validations have generally rested on very few test sites each. We present here what we believe to be the most comprehensive validation of the commercial application of CFD seen to date, with over 80 very diverse sites and in excess of 600 mast pairs.

Approach

DNV GL simulates the flow of air over terrain using in-house software powered by STAR-CCM+, a flexible, state-of-the-art CFD solver, set to solve the Reynolds-Averaged Navier-Stokes (RANS) equations with a two-equation turbulence closure. Among the sites to which we applied our standard commercial method in the past year, over 80 sites were equipped with two or more meteorological masts, thus allowing for cross-predictions between masts and a comparison between calculated and measured mean wind speeds (MWS). Calculations were also performed using a linear model for comparison.

Main body of abstract

In the spirit of a proof-of-principle, the validation we previously published at EWEA 2012 focused on a narrow band of site types: we had chosen from our database a set of multi-mast sites that had complex terrain but a moderate elevation range, and where neither forestry nor thermal stratification were dominant factors. This allowed us to demonstrate the usefulness of RANS CFD at least for an idealised class of sites.
The validation we present here is of a different character: all multi-mast sites on which we performed CFD modelling in the course of our commercial operations were systematically included. The sites are located across a range of continents, and represent a broad variety of terrain types and atmospheric conditions.
CFD yielded a smaller error than the linear model at a large majority of sites. On average, CFD reduced the modelling error on the MWS by one fifth or 1 percentage point compared to the linear model. CFD was 50% more likely than the linear model to keep errors below 2%; conversely, the linear model occasionally yielded errors greater than 20% whereas CFD did not.
CFD-modelled vertical shear also agreed well with the shear measured at site masts equipped with multiple anemometers. As highlighted by the EWEA Power Curve Working Group, shear is an important factor influencing real world power curves, and CFD promises to offer valuable estimates of this key parameter.

Conclusion

Linear models have remained the microscale workhorse of the wind industry since the 1980s despite their known weaknesses: complex terrain effects, flow separation, atmospheric stability, and forestry effects are not adequately captured, leading to large errors. CFD has oft been cited as the remedy to these issues, but these claims have been met with industry scepticism in the face of the scarce validation published to date. With an increasing commercial application of CFD at DNV GL and elsewhere, large-scale validation has now become possible, and a solid case is emerging that CFD brings significant value to energy production assessments.


Learning objectives
CFD has stepped into a permanent role alongside traditional linear models and can help reduce flow modelling errors, leading to better designed wind farms, better project returns, and ultimately lowering the cost of wind energy. DNV GL aims to help the financial community make informed judgments regarding the bankability of the latest modelling techniques and to this end we have consistently published growing validations as our CFD service has matured.