Back to the programme printer.gif Print



Tuesday, 11 March 2014
14:15 - 15:45 Aspects for offshore and complex terrain
Science & Research  


Room: Llevant
Session description

Siting in complex terrain is still a challenge for wind energy developers. This session will shed light on questions including:

What is the best way to estimate the energy resource over hilly and forested terrains? Are linearized models still useful and are computational fluid dynamics (CFD) models mature enough? What is the best way to use meso-scale models for wind energy resource estimation offshore? How does atmospheric stability affect turbulence?

Learning objectives:

  • Judge the performance of linearized flow models in comparison with CFD model for wind resource estimation in complex terrain
  • Understand how meso-scale models are best used for offshore annual energy production (AEP) estimation
  • Appreciate how atmospheric stability affects turbulence over forests and how to use standard measurements to estimate the stability
  • Get an insight into state-of-the-art meso-scale modelling for wind energy resource estimation
Lead Session Chair:
Jakob Mann, DTU Wind Energy

Co-chair(s):
Joachim Peinke, Uni Oldenburg, Germany
Davide Medici DNV GL - Energy, Italy
Co-authors:
Davide Medici (1) F P Antonio Segalini (2) Ebba Dellwik (3)
(1) DNV GL - Energy, Imola (BO), Italy (2) Linnè Flow Centre, KTH Mechanics, Stockholm, Sweden (3) DTU Wind Energy, Roskilde, Denmark

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 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 within
DNV GL. Dr. Medici is also supervising work and energy assessment analyses in Italy and
Middle East Countries as a senior engineer.

Abstract

On the determination of stability conditions over forested areas from velocity measurements.

Introduction

Stability is an important property of the atmospheric boundary layer, where the local state can be classified as being stable, neutral or unstable [1]. Methods have been proposed to classify local stability conditions from limited information, e.g. if only parameters related to wind speed and turbulence intensity are available. These have been however calibrated with data measured over homogeneous low-roughness lands, and it is expected that they must be modified over complex terrains and forests. The present work presents and validates proxies which can be used to estimate the stratification conditions by means of velocity measurements over forests.

Approach

If a large area is intended to be studied by means of one or several masts and especially in presence of factors such as complex terrain, forestry or strong stability effects, computational fluid dynamics (CFD) simulations can provide lower uncertainties and therefore a reduced error in the wind flow predictions [2]. It is therefore important to identify methods for characterising stability in order to improve the analysis of the sites. Atmospheric stability can be described with quantities such as gradient Richardson number, Ri, or Obukhov length, L, with definitions that can be found in the literature [1]. To calculate the Richardson number the mean temperature profile is needed, while the Obukhov length necessitates of a measure of the vertical kinematic heat flux, , a task usually performed by means of a tri-axial sonic anemometer. Often however, meteorological towers used for wind-resource assessment are equipped with cup anemometry and seldom with temperature sensors at multiple levels. The instrumentation may not be always sufficient to characterize thermal stratification through the Ri or L parameters. It is acknowledged that the definition include conditions of neutral flow and transition between strongly unstable and strongly stable, but the aim of this work is to introduce proxies that can be used also with relatively limited information. Therefore the definition of atmospheric stability is here based on the sign of Ri or L.

In the present study, data from several meteorological masts equipped with sonic anemometers, cup anemometers and temperature sensors at multiple levels are available and have been compared in order to investigate the existence of proxies for stability conditions. The data have been measured both with a high sampling rates of 20 Hz and with sampling rate of 1 Hz subsequently re-averaged to 10-minute statistics. The measurement levels extend up to a height of 140 m, starting within the forestry and extending to several canopy heights.

Main body of abstract

Four independent Swedish forest sites have been characterized either by the Obukhov length or by the Richardson number. Each site has been studied in a given sector where the forest properties are nearly homogeneous and feature trees ranging from few meters up to approximately 30 m. The results are consistent across all sites and show that more than 70% of the data over forest are defined as stable, namely are characterised by a positive Obukhov length or Richardson number. The negative quantities are instead assumed to identify unstable/neutral conditions and account for the remaining 30% of the points.

Stability regimes have been identified for Site 1 following the Richardson number classification.



Figure 1 above shows the stability parameter as function of the time of the day (ToD) for one of the masts, with stable points defined as having a positive Ri. It can be immediately observed that at night most of the stratification is stable, while in day-time both stable and unstable conditions are equally frequent. This result identifies a first simple criterion to separate stability regimes based just on the ToD, since after 17:00 and before 6:00 the flow can be in fact considered as only stable. The validity of the ToD proxy to identify stable flow is also confirmed by the Obukhov length plot in Figure 2 below for Site 2 as an example.



The criterion includes only a minor subset of unstable/neutral points, which occurs at night. It is however clear that during day-time another proxy must be used to complete an accurate characterisation of all points.

The second successful proxy is based on a combination of the local shear evaluated by means of the exponent of a fitting power law of the form U=Aza, and of the Turbulence Intensity (TI) calculated as the ratio between the standard deviation of the wind speed and the local mean velocity, evaluated at a certain height. The proxy requires to some extent the threshold to be arbitrarily selected and slightly changes between sites; however, as shown below in Figure 3 for Site 1, it is effective at identifying those points that during daytime are stable. As expected following the literature, these points are associated to TI values below a certain reference level together with high shear conditions. Similarly to the ToD proxy, a limited number of unstable/neutral points is included together with the stable points.




Figure 4 above shows the scatter plot between these two variables including both night and day points for Site 2. There is a rather large overlapping region which prevents this proxy to be the only approach required for the classification of the stability at the sites. It can be noted by observing the figure that proxies based on turbulence intensity or shear only, i.e. applied with a threshold which is either a vertical or horizontal line crossing the dataset, cannot be as effective in identifying the stability conditions at the sites as their combination.

The results from Site 1 are detailed here as an example. At this site 74% of all points are stable. Approximately 53% occur between 17:00 and 6:00 and are therefore correctly identified by the ToD proxy. Only 2% of unstable/neutral points are included within this proxy. An additional 15% of stable points are identified during daytime by being characterised by low turbulence intensity and high shear, with only 2% (again) of unstable/neutral points meeting the same criterion. Therefore of the 74% of stable occurrences at this site, the two proxies here utilised are capable of correctly identifying 68% of the points without requiring the calculation of either the Obukhov length or Richardson number.

Conclusion

When buoyancy effects are not properly accounted for, significant mismatches may arise between the actual wind conditions and the numerical simulations. Some of the methodologies proposed to reduce these errors entail the study of the stability conditions within the atmospheric boundary layer to better classify the flow. Clearly the preferred approach is to classify the stability conditions by means of a rigorous calculation of the Obukhov length or of the Richardson number; however this is not always feasible on sites commercially developed for wind energy.

In the present study, a simple criterion based on time-of-day together with a combination of turbulence intensity and shear thresholds has been proved to effectively identify the near totality of the stable conditions on forest sites. The authors had access to detailed and high-quality measurements to compare the methodology proposed in this paper based on these proxies with the results from a stability classification based on the Obukhov length or Richardson number. An important finding is therefore that proxies for classification of the stability of the atmospheric boundary layer can be found also in the absence of temperature measurements at multiple levels or tri-axial sonic anemometers. It is also shown that the atmospheric boundary layer over forests is predominantly stable.

This approach to analyse stability has provided consistent results across all four sites and it can be used for improving wind speed simulations by means of numerical models, such as stable CFD, which require the identification of the stability conditions. The approach could easily be refined by taking the actual day-length into account.


Learning objectives
The application of stable CFD models has stepped into a key role alongside traditional linear models and can help reduce wind flow modelling errors, leading to better project refinances and ultimately lowering the cost of wind energy. This however requires information on atmospheric stability. Knowledge of stability conditions over forests is not a trivial task and the proxies presented in this work identify a simple and effective methodology to better understand atmospheric boundary layers.


References
[1] – Garrat, J. 1992 The atmospheric boundary layer, Cambridge University Press.
[2] – Bleeg, J. – Modeling Stable Thermal Stratification and Its Impact on Wind Flow Over Topography. AWEA WindPower 2012.