Back to the programme printer.gif Print

Tuesday, 11 March 2014
11:15 - 12:45 Optimising measurement strategies to maximise project value: Is the industry making false economies at the expense of project value?
Resource Assessment  

Room: Llevant
Session description

Striking the right balance between costs and benefits when designing a measurement campaign has always been a challenge. Nowadays the situation is more complex due to:

  • sophisticated instrumentation options (and their limitations);
  • wind farms of larger spatial extent located in more diverse climates;
  • advanced flow modelling.

It is no longer a straightforward process deciding on the optimal measurement strategy to minimise uncertainties in the energy assessment for a specific project. Assessing the resulting financial benefit is just as challenging. The interpretation of the data for site classification and thus choice of turbine has also become more complex.

Learning objectives

  • Evaluate the most efficient use of instrumentation for a specific site
  • Understand and quantify the connection between measurement uncertainty and project economics and loads
  • Make a more accurate choice of turbine type
  • Express uncertainty variations across the site as the basis for cost-efficient measurement campaigns
Lead Session Chair:
Wiebke Langreder, Wind Solutions, Denmark

Jan Coelingh, Vattenfall
Doron Callies Fraunhofer IWES, Germany
Tobias Klaas (1) F P
(1) Fraunhofer IWES, Kassel, Germany

Printer friendly version: printer.gif Print

Presenter's biography

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

Mr. Doron Callies has been working as a scientist at the Fraunhofer IWES research institute for the last six years. His research interests lie in the field of site assessment with focus on remote sensing and the use of lidars in complex terrain. He studied environmental technology at university of applied sciences in Hamburg and did his master degree in renewable energies at university of Kassel.


Lidar errors in complex terrain and strategies for an optimized measurement design in resource assessment


Lidars are becoming important tools for wind resource assessments in all kinds of terrain. Compared to mast measurements mobility and flexibility are their greatest benefits. However, a broad data basis for verification is needed to increase acceptance. There are several effects that need to be considered when setting up a measurement campaign. Site dependent orography effects can cause considerable errors. The magnitude of these errors depends on the level of terrain complexity and measurement location. High errors lead to significantly increased measurement uncertainties. Their estimation with models underlies uncertainties regarding parameterization, e.g. forest height and directional offset settings.


Measurements from Fraunhofer IWES 200m mast and three tall masts (100-140m) in forested and moderately complex terrain with co-located lidar measurements are used to examine the influence of terrain on the lidar performance in detail. Mast data was kindly provided by the German project developers juwi, ABO-Wind and OSTWIND as part of the project “Utilization of inland wind power”.
A RANS CFD model is used to estimate lidar errors at all masts. Model results are being evaluated depending on direction offset settings and model parameterization. Subsequently, the variation in errors with location is analyzed in terms of an error map.

Main body of abstract

Wind measurements of the masts and pulsed doppler lidars (Windcube) placed next to the masts are analyzed in terms of a detailed comparison. This leads to a comprehensive overview over the most important factors influencing measurement uncertainty of lidars. In terms of evaluation, the model results show good performance compared to measurement. The site specific lidar errors are well estimated.
Comparison results show that at flat, forested and even at moderately complex sites without steep slopes in the direct proximity, lidar results are comparable to those from met masts. The influence of complex terrain can be neglected if the lidar is positioned in sufficient distance to hills and ridges.
To demonstrate variations of lidar errors due to choice of location the CFD model is run to calculate the wind flow in an area of 4x4km² 10x10m² resolution, 36 directions).
The model results are used to simulate the lidar measurement error in 50x50m² spatial resolution. This leads to a visualization of the estimated errors to characterize measurement locations.
Results show that choice of location significantly influences measurement errors. Around the 200m mast site these vary between +/-3% in areas of under-estimation respectively over-estimation, both suboptimal choices for lidar measurements. The model can be used to identify areas of low errors to improve measurement design.
Varying the modeled lidar offset setting from 0 to 85° leads to a spread in results of up to 1.5%, depending on site and wind direction. This reveals another uncertainty factor in error estimation for Windcube lidars.


Lidar measurements can be applied for resource assessments at flat and moderately complex, forested (and non-forested) sites with good results. If complex orography is relatively far away the lidar error due to terrain effects was also negligible.
For measurements in complex terrain the error can be reduced by using CFD models. Careful planning of a measurement campaign helps identifying suitable locations with low expected errors before setting up the device. This is a valuable approach for project planners that reduces the uncertainty and leads to better results in wind resource assessments.

Learning objectives
Delegates will learn about applicability of lidars at sites of different complexity. Lidar performance in comparison to tall masts is presented, which enables users to easier evaluate possible lidar measurement sites.
A reasonable choice of location is crucial for a successful lidar campaign. Delegates will learn how the assessment of site specific errors beforehand increases the value of measurement data and reduces uncertainty.