Turbulence Intensity Becomes a Key Factor When Planning Wind Farms
Wind power plays a substantial part in worldwide plans to increase renewable energy utilisation. In future existing wind farms will need to be repowered and new wind farm sites are going to need to be explored. However, in countries like Germany new sites are becoming rare and there is a limit to the amount of energy that can be generated from a given area. When placing wind turbines too close together turbulence generated by the wakes of the wind turbines may seriously affect their structural integrity and so becomes a limiting factor in wind farm layouts. This holds especially true for complex terrain, where the terrain-generated turbulence makes a significant contribution to the overall load.
By Thomas Hahm, F2E Fluid & Energy Engineering, Germany
The international IEC standard 61400-1 defines the design requirements for wind turbines and organises different wind turbine classes in terms of reference wind speed and turbulence intensity. The intention of these classes is to cover most applications, but they do not give a precise representation of any specific site, or even of a specific wind farm layout, where the individual wind turbine is subject to the influence of multiple wakes from upwind machines. Therefore the suitability of a wind turbine in a wind farm has to be assessed.
According to the standard the assessment must take into account the deterministic and turbulent flow characteristics associated with single or multiple wakes from upwind machines, including the effects of the spacing between the machines for all ambient wind speeds and wind directions relevant to power production.
Inside a Single Wake
The wake of a wind turbine is characterised by high velocity gradients, an enlarged turbulence and a random meandering. Especially in the near-wake region (below approximately 3–5 rotor diameters) where a sharply defined wake is still present, the loads on downwind turbines can be severe. For more than 10 years we have applied three-dimensional Computational Fluid Dynamics (CFD) to simulate the turbulent flow behind a wind turbine. These simulations can be validated with high resolution measurements and give an insight into the wind conditions a turbine experiences inside a wind farm.
Figure 1 shows a snapshot of a meandering wake behind an ENERCON E-66 wind turbine (blue represents high wind speeds and white represents low speeds). It is taken from a transient calculation which has been run for a couple of weeks on a small computer cluster and it gives an impression of what a turbine experiences inside a wind farm.
Effective Turbulence Intensity
For a day-to-day assessment of the suitability of a wind turbine at a site in a wind farm the effects from multiple wakes are transferred into a single parameter – the effective turbulence intensity. The effective turbulence intensity is a substitute value to be applied for the whole lifespan of the wind turbine. It rates the load caused by the ambient turbulence intensity and the additional load induced by the wake situation.
There are two general approaches to assess the structural integrity of a wind turbine at a site:
Assessing Structural Integrity
Using the first approach the effective turbulence intensities have to be lower than or equal to the design values. If this fails, one can use the effective turbulence intensities as input for a site-specific load calculation. The site-specific loads are then compared to the design loads to demonstrate the structural integrity in accordance with the second approach. In this process high turbulence intensities have to be compensated by other wind conditions (namely annual mean wind speed) that are below the design values assigned to them.
Figure 2 shows the role of the effective turbulence intensity in the process of assessing the structural integrity of a wind turbine within a wind farm.
Ambient Turbulence Comes First
The assessment of turbulence intensity requires two steps. These are: the evaluation of ambient turbulence intensity at the site, and the additional wake-generated turbulence intensity. The following description of the calculation procedure is based on the software code ‘wake2e’ from the author’s company.
The first step is to evaluate ambient turbulence intensity at the site. A good approach would be to measure on site. However, even a measurement period of one year will probably not cover all necessary wind speeds at all wind directions. Furthermore, measurements may not be easily extrapolated to hub height or to all wind turbine sites in a large array with varying terrain.
An alternative approach, in the case where no data or only incomplete data of ambient turbulence intensity is available, would be a calculation based on surface roughness or land cover data.
CORINE Data
For the latter the CORINE satellite data on land cover can be used for most of the European countries. The CORINE data provides terrain information on a 100m by 100m raster and can be used to describe the surroundings of every wind turbine. Distances of at least 10 to 15km around the wind turbines should be included. Subsequently roughness categories have to be assigned to the individual terrain sectors. On the basis of this roughness classification ambient turbulence intensities need to be calculated in the next step. The ambient turbulence intensities are determined for the various directions and wind speeds at hub height. Figure 3 shows the current coverage of raster data for the CORINE land cover 2006 inventory.
More Input Data
Once the ambient turbulence intensities are evaluated, effective turbulence intensities can be calculated. The required input data is as follows:
No Measurements of Effective Turbulence Intensity
One should be aware that a direct validation of the results is not possible as the effective turbulence intensity is a substitute value to be applied for the whole lifespan of the wind turbine and is not accessible by measurements. Therefore results have been compared to load calculations using common approaches for wake deficit and turbulence modelling. These calculations assume that the wind turbine operates for 100% of its lifetime under half-wake conditions (i.e. half of the rotor is exposed to the wake and the other half to the undisturbed wind field), which should cover any loading the wind turbine experiences in a real wind farm. The results show that under these conditions closer distances between the wind turbines will be predicted, compared to the approach described above and used, for example, in the software ‘wake2e’. So this is one proof that the concept of effective turbulence intensity is conservative for close distances.
Other Uncertainties
The methods used to calculate ambient turbulence intensity on the basis of land cover or roughness data have been validated in several projects in the north of Germany and show good agreement with measurements. Few comparisons have been made outside this area. These results indicate that the use of turbulence structure correction parameters, in accordance with the international standard IEC 61400-1, should be inevitable in complex terrain. Otherwise the results may under predict ambient turbulence, especially at high wind speeds. In this context it may be necessary to treat the edges of forests as an escarpment of the terrain. It should be noted that in some cases of complex terrain, especially in hill top positions, the results may considerably overestimate the ambient turbulence intensity. In general it is our experience that ambient turbulence intensities based on land cover or roughness data tend to be more conservative at low wind speeds if the course of the wind speed follows the Normal Turbulence Model of the IEC 61400-1 standard, as is done, for example, in ‘wake2e’. One can therefore be quite confident in obtaining conservative results with regard to effective turbulence intensities.
Work Flow
Figure 4 gives a summary of the evaluation process and indicates methods of dealing with exceedances. This may be done by the use of effective turbulence intensities for site-specific load calculations, shut-offs, or curtailments of wind turbines, or as a last measure, the complete change of the wind farm layout. To avoid a nasty surprise at the end of the planning phase, and to prevent delays in the granting of a building permit, one should control the turbulences arising in the wind farm from the very beginning of the planning process.
Biography of the Author
Thomas Hahm has a PhD in engineering sciences from the University of Dortmund. From 1999 to 2007 he worked for TÜV NORD SysTec, where his last position was Head of Competence Center Dynamic Systems. Since 2008 he has been Senior Expert Wind Energy and CFD (Computational Fluid Dynamics) at F2E. He has more than 10 years of experience in the assessment of site-specific wind conditions and structural integrity of wind turbines.{/access}
Wind power plays a substantial part in worldwide plans to increase renewable energy utilisation. In future existing wind farms will need to be repowered and new wind farm sites are going to need to be explored. However, in countries like Germany new sites are becoming rare and there is a limit to the amount of energy that can be generated from a given area. When placing wind turbines too close together turbulence generated by the wakes of the wind turbines may seriously affect their structural integrity and so becomes a limiting factor in wind farm layouts. This holds especially true for complex terrain, where the terrain-generated turbulence makes a significant contribution to the overall load.By Thomas Hahm, F2E Fluid & Energy Engineering, Germany
{access view=!registered}Only logged in users can view the full text of the article.{/access}{access view=registered}This article explains how and why turbulence intensity must be considered during the design stage of creating a wind farm layout and gives an overview of the factors that must be considered in this process, the influencing parameters and the uncertainties.
Design ValuesThe international IEC standard 61400-1 defines the design requirements for wind turbines and organises different wind turbine classes in terms of reference wind speed and turbulence intensity. The intention of these classes is to cover most applications, but they do not give a precise representation of any specific site, or even of a specific wind farm layout, where the individual wind turbine is subject to the influence of multiple wakes from upwind machines. Therefore the suitability of a wind turbine in a wind farm has to be assessed.
According to the standard the assessment must take into account the deterministic and turbulent flow characteristics associated with single or multiple wakes from upwind machines, including the effects of the spacing between the machines for all ambient wind speeds and wind directions relevant to power production.
Inside a Single Wake
The wake of a wind turbine is characterised by high velocity gradients, an enlarged turbulence and a random meandering. Especially in the near-wake region (below approximately 3–5 rotor diameters) where a sharply defined wake is still present, the loads on downwind turbines can be severe. For more than 10 years we have applied three-dimensional Computational Fluid Dynamics (CFD) to simulate the turbulent flow behind a wind turbine. These simulations can be validated with high resolution measurements and give an insight into the wind conditions a turbine experiences inside a wind farm.
Figure 1 shows a snapshot of a meandering wake behind an ENERCON E-66 wind turbine (blue represents high wind speeds and white represents low speeds). It is taken from a transient calculation which has been run for a couple of weeks on a small computer cluster and it gives an impression of what a turbine experiences inside a wind farm.
Effective Turbulence Intensity
For a day-to-day assessment of the suitability of a wind turbine at a site in a wind farm the effects from multiple wakes are transferred into a single parameter – the effective turbulence intensity. The effective turbulence intensity is a substitute value to be applied for the whole lifespan of the wind turbine. It rates the load caused by the ambient turbulence intensity and the additional load induced by the wake situation.
There are two general approaches to assess the structural integrity of a wind turbine at a site:
- Demonstrate that the site conditions are no more severe than those assumed for the design of the wind turbine.
- Demonstrate the structural integrity for conditions equal to, or more severe than, those at the site.
Assessing Structural Integrity
Using the first approach the effective turbulence intensities have to be lower than or equal to the design values. If this fails, one can use the effective turbulence intensities as input for a site-specific load calculation. The site-specific loads are then compared to the design loads to demonstrate the structural integrity in accordance with the second approach. In this process high turbulence intensities have to be compensated by other wind conditions (namely annual mean wind speed) that are below the design values assigned to them.
Figure 2 shows the role of the effective turbulence intensity in the process of assessing the structural integrity of a wind turbine within a wind farm.
Ambient Turbulence Comes First
The assessment of turbulence intensity requires two steps. These are: the evaluation of ambient turbulence intensity at the site, and the additional wake-generated turbulence intensity. The following description of the calculation procedure is based on the software code ‘wake2e’ from the author’s company.
The first step is to evaluate ambient turbulence intensity at the site. A good approach would be to measure on site. However, even a measurement period of one year will probably not cover all necessary wind speeds at all wind directions. Furthermore, measurements may not be easily extrapolated to hub height or to all wind turbine sites in a large array with varying terrain.
An alternative approach, in the case where no data or only incomplete data of ambient turbulence intensity is available, would be a calculation based on surface roughness or land cover data.
CORINE Data
For the latter the CORINE satellite data on land cover can be used for most of the European countries. The CORINE data provides terrain information on a 100m by 100m raster and can be used to describe the surroundings of every wind turbine. Distances of at least 10 to 15km around the wind turbines should be included. Subsequently roughness categories have to be assigned to the individual terrain sectors. On the basis of this roughness classification ambient turbulence intensities need to be calculated in the next step. The ambient turbulence intensities are determined for the various directions and wind speeds at hub height. Figure 3 shows the current coverage of raster data for the CORINE land cover 2006 inventory.
More Input Data
Once the ambient turbulence intensities are evaluated, effective turbulence intensities can be calculated. The required input data is as follows:
- Ambient turbulence intensity (if not based on CORINE data):
- Measurements or data of surface roughness
- Frequency distribution of wind speeds and wind directions:
- Calculated or measured values
- Wind farm layout and information about curtailments and shut-offs
- Elevation data for orographically complex terrain
No Measurements of Effective Turbulence Intensity
One should be aware that a direct validation of the results is not possible as the effective turbulence intensity is a substitute value to be applied for the whole lifespan of the wind turbine and is not accessible by measurements. Therefore results have been compared to load calculations using common approaches for wake deficit and turbulence modelling. These calculations assume that the wind turbine operates for 100% of its lifetime under half-wake conditions (i.e. half of the rotor is exposed to the wake and the other half to the undisturbed wind field), which should cover any loading the wind turbine experiences in a real wind farm. The results show that under these conditions closer distances between the wind turbines will be predicted, compared to the approach described above and used, for example, in the software ‘wake2e’. So this is one proof that the concept of effective turbulence intensity is conservative for close distances.
Other Uncertainties
The methods used to calculate ambient turbulence intensity on the basis of land cover or roughness data have been validated in several projects in the north of Germany and show good agreement with measurements. Few comparisons have been made outside this area. These results indicate that the use of turbulence structure correction parameters, in accordance with the international standard IEC 61400-1, should be inevitable in complex terrain. Otherwise the results may under predict ambient turbulence, especially at high wind speeds. In this context it may be necessary to treat the edges of forests as an escarpment of the terrain. It should be noted that in some cases of complex terrain, especially in hill top positions, the results may considerably overestimate the ambient turbulence intensity. In general it is our experience that ambient turbulence intensities based on land cover or roughness data tend to be more conservative at low wind speeds if the course of the wind speed follows the Normal Turbulence Model of the IEC 61400-1 standard, as is done, for example, in ‘wake2e’. One can therefore be quite confident in obtaining conservative results with regard to effective turbulence intensities.
Work Flow
Figure 4 gives a summary of the evaluation process and indicates methods of dealing with exceedances. This may be done by the use of effective turbulence intensities for site-specific load calculations, shut-offs, or curtailments of wind turbines, or as a last measure, the complete change of the wind farm layout. To avoid a nasty surprise at the end of the planning phase, and to prevent delays in the granting of a building permit, one should control the turbulences arising in the wind farm from the very beginning of the planning process.
Biography of the Author
Thomas Hahm has a PhD in engineering sciences from the University of Dortmund. From 1999 to 2007 he worked for TÜV NORD SysTec, where his last position was Head of Competence Center Dynamic Systems. Since 2008 he has been Senior Expert Wind Energy and CFD (Computational Fluid Dynamics) at F2E. He has more than 10 years of experience in the assessment of site-specific wind conditions and structural integrity of wind turbines.{/access}
