Quantification of power losses due to wind turbine wake interactions through SCADA,meteorological and wind LiDAR data

Power production of an onshore wind farm is investigated through supervisory control and data acquisition data, while the wind field is monitored through scanning light detection and ranging measurements and meteorological data acquired from a met‐tower located in proximity to the turbine array. The power production of each turbine is analysed as functions of the operating region of the power curve, wind direction and atmospheric stability. Five different methods are used to estimate the potential wind power as a function of time, enabling an estimation of power losses connected with wake interactions. The most robust method from a statistical standpoint is that based on the evaluation of a reference wind velocity at hub height and experimental mean power curves calculated for each turbine and different atmospheric stability regimes. The synergistic analysis of these various datasets shows that power losses are significant for wind velocities higher than cut‐in wind speed and lower than rated wind speed of the turbines. Furthermore, power losses are larger under stable atmospheric conditions than for convective regimes, which is a consequence of the stability‐driven variability in wake evolution. Light detection and ranging measurements confirm that wind turbine wakes recover faster under convective regimes, thus alleviating detrimental effects due to wake interactions. For the wind farm under examination, power loss due to wake shadowing effects is estimated to be about 4% and 2% of the total power production when operating under stable and convective conditions, respectively. However, cases with power losses about 60‐80% of the potential power are systematically observed for specific wind turbines and wind directions. Copyright © 2017 John Wiley & Sons, Ltd.     

LiDAR‐based characterization of mid‐altitude wind conditions for airborne wind energy systems

Engineers and researchers working on the development of airborne wind energy systems (AWES) still rely on oversimplified wind speed approximations and coarsely sampled reanalysis data because of a lack of high‐resolution wind data at altitudes above 200 m. Ten‐minute average wind speed LiDAR measurements up to an altitude of 1100 m and data from nearby weather stations were investigated with regard to wind energy generation and impact on LiDAR measurements. Data were gathered by a long‐range pulsed Doppler LiDAR device installed on flat terrain. Because of the low overall carrier‐to‐noise ratio, a custom‐filtering technique was applied. Our analyses show that diurnal variation and atmospheric stability significantly affect wind conditions aloft which cause a wide range of wind speeds and a multimodal probability distribution that cannot be represented by a simple Weibull distribution fit. A better representation of the actual wind conditions can be achieved by fitting Weibull distributions separately to stable and unstable conditions. Splitting and clustering the data by simulated surface heat flux reveals substate stratification responsible for the multimodality. We classify different wind conditions based on these substates, which result in different wind energy potential. We assess optimal traction power and optimal operating altitudes statistically as well as for specific days based on a simplified AWES model. Using measured wind speed standard deviation, we estimate average turbulence intensity and show its variation with altitude and time. Selected short‐term data sets illustrate temporal changes in wind conditions and atmospheric stratification with a high temporal and vertical resolution.     

An experimental and numerical study of the atmospheric stability impact on wind turbine wakes

In this paper, the impact of atmospheric stability on a wind turbine wake is studied experimentally and numerically. The experimental approach is based on full‐scale (nacelle based) pulsed lidar measurements of the wake flow field of a stall‐regulated 500 kW turbine at the DTU Wind Energy, Risø campus test site. Wake measurements are averaged within a mean wind speed bin of 1 m s^?1 and classified according to atmospheric stability using three different metrics: the Obukhov length, the Bulk–Richardson number and the Froude number. Three test cases are subsequently defined covering various atmospheric conditions. Simulations are carried out using large eddy simulation and actuator disk rotor modeling. The turbulence properties of the incoming wind are adapted to the thermal stratification using a newly developed spectral tensor model that includes buoyancy effects. Discrepancies are discussed, as basis for future model development and improvement. Finally, the impact of atmospheric stability on large‐scale and small‐scale wake flow characteristics is presently investigated. Copyright © 2015 John Wiley & Sons, Ltd.     

Wind turbine aerodynamic response under atmospheric icing conditions

This article deals with the atmospheric ice accumulation on wind turbine blades and its effect on the aerodynamic performance and structural response. The role of eight atmospheric and system parameters on the ice accretion profiles was estimated using the 2D ice accumulation software lewice Twenty‐four hours of icing, with time varying wind speed and atmospheric icing conditions, was simulated on a rotor. Computational fluid dynamics code, FLUENT, was used to estimate the aerodynamic coefficients of the blade after icing. The results were also validated against wind tunnel measurements performed at LM Wind Power using a NACA64618 airfoil. The effects of changes in geometry and surface roughness are considered in the simulation. A blade element momentum code WT‐Perf is then used to quantify the degradation in performance curves. The dynamic responses of the wind turbine under normal and iced conditions were simulated with the wind turbine aeroelastic code HAWC2. The results show different behaviors below and above rated wind speeds. In below rated wind speed, for a 5 MW virtual NREL wind turbine, power loss up to 35% is observed, and the rated power is shifted from wind speed of 11 to 19 m s^?1. However, the thrust of the iced rotor in below rated wind speed is smaller than the clean rotor up to 14%, but after rated wind speed, it is up to 40% bigger than the clean rotor. Finally, it is briefly indicated how the results of this paper can be used for condition monitoring and ice detection. Copyright © 2012 John Wiley & Sons, Ltd.     

A neural network approach to local downscaling of GCM output for assessing wind power implications of climate change

A methodology is presented for downscaling General Circulation Model (GCM) output to predict surface wind speeds at scales of interest in the wind power industry under expected future climatic conditions. The approach involves a combination of Neural Network tools and traditional weather forecasting techniques. A Neural Network transfer function is developed to relate local wind speed observations to large scale GCM predictions of atmospheric properties under current climatic conditions. By assuming the invariability of this transfer function under conditions of doubled atmospheric carbon dioxide, the resulting transfer function is then applied to GCM output for a transient run of the National Center for Atmospheric Research coupled ocean-atmosphere GCM. This methodology is applied to three test sites in regions relevant to the wind power industry—one in Texas and two in California. Changes in daily mean wind speeds at each location are presented and discussed with respect to potential implications for wind power generation.     

Influence of atmospheric stability on wind turbine loads

Simulations of wind turbine loads for the NREL 5 MW reference wind turbine under diabatic conditions are performed. The diabatic conditions are incorporated in the input wind field in the form of wind profile and turbulence. The simulations are carried out for mean wind speeds between 3 and 16 m s ^{? 1} at the turbine hub height. The loads are quantified as the cumulative sum of the damage equivalent load for different wind speeds that are weighted according to the wind speed and stability distribution. Four sites with a different wind speed and stability distribution are used for comparison. The turbulence and wind profile from only one site is used in the load calculations, which are then weighted according to wind speed and stability distributions at different sites. It is observed that atmospheric stability influences the tower and rotor loads. The difference in the calculated tower loads using diabatic wind conditions and those obtained assuming neutral conditions only is up to 17%, whereas the difference for the rotor loads is up to 13%. The blade loads are hardly influenced by atmospheric stability, where the difference between the calculated loads using diabatic and neutral input wind conditions is up to 3% only. The wind profiles and turbulence under diabatic conditions have contrasting influences on the loads; for example, under stable conditions, loads induced by the wind profile are larger because of increased wind shear, whereas those induced by turbulence are lower because of less turbulent energy. The tower base loads are mainly influenced by diabatic turbulence, whereas the rotor loads are influenced by diabatic wind profiles. The blade loads are influenced by both, diabatic wind profile and turbulence, that leads to nullifying the contrasting influences on the loads. The importance of using a detailed boundary‐layer wind profile model is also demonstrated. The difference in the calculated blade and rotor loads is up to 6% and 8%, respectively, when only the surface‐layer wind profile model is used in comparison with those obtained using a boundary‐layer wind profile model. Finally, a comparison of the calculated loads obtained using site‐specific and International Electrotechnical Commission (IEC) wind conditions is carried out. It is observed that the IEC loads are up to 96% larger than those obtained using site‐specific wind conditions.Copyright © 2012 John Wiley & Sons, Ltd.     

Diurnal cycle RANS simulations applied to wind resource assessment

Microscale computational fluid dynamics (CFD) models can be used for wind resource assessment on complex terrains. These models generally assume neutral atmospheric stratification, an assumption that can lead to inaccurate modeling results and to large uncertainties at certain sites. We propose a methodology for wind resource evaluation based on unsteady Reynolds averaged Navier‐Stokes (URANS) simulations of diurnal cycles including the effect of thermal stratification. Time‐dependent boundary conditions are generated by a 1D precursor to drive 3D diurnal cycle simulations for a given geostrophic wind direction sector. Time instants of the cycle representative of four thermal stability regimes are sampled within diurnal cycle simulations and combined with masts time series to obtain the wind power density (WPD). The methodology has been validated on a complex site instrumented with seven met masts. The WPD spatial distribution is in good agreement with observations with the mean absolute error improving 17.1% with respect to the neutral stratification assumption.     

Extrapolating wind speed time series vs. Weibull distribution to assess wind resource to the turbine hub height: A case study on coastal location in Southern Italy

Increasing knowledge on wind shear models to strengthen their reliability appears as a crucial issue, markedly for energy investors to accurately predict the average wind speed at different turbine hub heights, and thus the expected wind energy output. This is particularly helpful during the feasibility study to abate the costs of a wind power project, thus avoiding installation of tall towers, or even more expensive devices such as LIDAR or SODAR.The power law (PL) was found to provide the finest representation of wind speed profiles and is hence the focus of the present study. Besides commonly used for vertical extrapolation of wind speed time series, the PL relationship between “instantaneous” wind profiles was demonstrated by Justus and Mikhail to be consistent with the height variation of Weibull distribution. Therefore, in this work a comparison is performed between these two different PL–based extrapolation approaches to assess wind resource to the turbine hub height: (i) extrapolation of wind speed time series, and (ii) extrapolation of Weibull wind speed distribution. The models developed by Smedman–Högström and Högström (SH), and Panofsky and Dutton (PD) were used to approach (i), while those from Justus and Mikhail (JM) and Spera and Richards (SR) to approach (ii). Models skill in estimating wind shear coefficient was also assessed and compared.PL extrapolation models have been tested over a flat and rough location in Apulia region (Southern Italy), where the role played by atmospheric stability and surface roughness, along with their variability with time and wind characteristics, has been also investigated. A 3-year (1998–2000) 1–h dataset, including wind measurements at 10 and 50 m, has been used. Based on 10–m wind speed observations, the computation of 50–m extrapolated wind resource, Weibull distribution and energy yield has been made. This work is aimed at proceeding the research issue addressed within a previous study, where PL extrapolation models were tested and compared in extrapolating wind resource and energy yield from 10 to 100 m over a complex–topography and smooth coastal site in Tuscany region (Central Italy). As a result, wind speed time series extrapolating models proved to be the most skilful, particularly PD, based on the similarity theory and thus addressing all stability conditions. However, comparable results are returned by the empirical JM Weibull distribution extrapolating model, which indeed proved to be preferable as being: (i) far easier to be used, as z₀–, stability–, and wind speed time series independent; (ii) more conservative, as wind energy is underpredicted rather than overpredicted.     

Wind turbine fatigue loads as a function of atmospheric conditions offshore

In recent years, there has been a growing interest by the wind energy community to assess the impact of atmospheric stability on wind turbine performance; however, up to now, typically, stability is considered in several distinct arbitrary stability classes. As a consequence, each stability class considered still covers a wide range of conditions. In this paper, wind turbine fatigue loads are studied as a function of atmospheric stability without a classification system, and instead, atmospheric conditions are described by a continuous joint probability distribution of wind speed and stability. Simulated fatigue loads based upon this joint probability distribution have been compared with two distinct different cases, one in which seven stability classes are adopted and one neglecting atmospheric stability by following International Electrotechnical Commission (IEC) standards. It is found that for the offshore site considered in this study, fatigue loads of the blade root, rotor and tower loads significantly increase if one follows the IEC standards (by up to 28% for the tower loads) and decrease if one considers several stability classes (by up to 13% for the tower loads). The substantial decrease found for the specific stability classes can be limited by considering one stability class that coincides with the mean stability of a given hub height wind speed. The difference in simulated fatigue loads by adopting distinct stability classes is primarily caused by neglecting strong unstable conditions for which relatively high fatigue loads occur. Combined, it is found that one has to carefully consider all stability conditions in wind turbine fatigue load simulations. Copyright © 2016 John Wiley & Sons, Ltd.     

Difficulties in using power laws for wind energy assessment

A 1/7 power law is often used to estimate atmospheric wind profiles. The use of this relationship, while perhaps appropriate in a climatological sense for daytime wind profiles in the lower atmosphere, frequently results in serious underestimates of wind speeds aloft at night. Several years of wind speed data from a 45 m tower and from use of a Doppler acoustic sounder (sodar) indicate that low-level wind speed maxima seem to form on 50 per cent of summer nights and 15 per cent of winter nights in northeastern Illinois. Even with 24 hr averages, the extrapolation of 6 m wind speeds to those at 45 m calculated with the 1/7 power law expression are found to be approximately 15 per cent too small, corresponding to a 40 per cent underestimate of wind power potential.     

Evaluating winds and vertical wind shear from Weather Research and Forecasting model forecasts using seven planetary boundary layer schemes

The existence of vertical wind shear in the atmosphere close to the ground requires that wind resource assessment and prediction with numerical weather prediction (NWP) models use wind forecasts at levels within the full rotor span of modern large wind turbines. The performance of NWP models regarding wind energy at these levels partly depends on the formulation and implementation of planetary boundary layer (PBL) parameterizations in these models. This study evaluates wind speeds and vertical wind shears simulated by the Weather Research and Forecasting model using seven sets of simulations with different PBL parameterizations at one coastal site over western Denmark. The evaluation focuses on determining which PBL parameterization performs best for wind energy forecasting, and presenting a validation methodology that takes into account wind speed at different heights. Winds speeds at heights ranging from 10 to 160 m, wind shears, temperatures and surface turbulent fluxes from seven sets of hindcasts are evaluated against observations at Høvsøre, Denmark. The ability of these hindcast sets to simulate mean wind speeds, wind shear, and their time variability strongly depends on atmospheric static stability. Wind speed hindcasts using the Yonsei University PBL scheme compared best with observations during unstable atmospheric conditions, whereas the Asymmetric Convective Model version 2 PBL scheme did so during near‐stable and neutral conditions, and the Mellor–Yamada–Janjic PBL scheme prevailed during stable and very stable conditions. The evaluation of the simulated wind speed errors and how these vary with height clearly indicates that for wind power forecasting and wind resource assessment, validation against 10 m wind speeds alone is not sufficient. Copyright © 2012 John Wiley & Sons, Ltd.     

Analysis and parameterisation of wind profiles in the low atmosphere

Knowledge of temporal and spatial variation of the wind is important to obtain accurate calculations of wind power potential. The best estimation method involves direct measurement of wind speed which, however, is not always possible. In such cases, a good parameterisation of wind profile is necessary. A RASS sodar located in Northern Spain has been used in this paper. This device provided a broad database of 10-min averages from August 2002 to January 2004. The vertical range extended from 40 to 500 m in 20 m levels, although the 220 m level was selected as the upper boundary by analysis of wind speed and temperature vertical profiles. Hourly medians were calculated each month in the 10 lower levels, yielding a sharp contrast between day and night. Flat wind speed profiles were clear during day, mainly in summer, due to convection produced by surface heating. However, stable stratification favoured horizontal movement and wind speed values increasing with height were observed during the night. Power and logarithmic laws have been fitted from vertical profiles of hourly wind speed medians. The exponent of the power law showed hourly medians greater than 0.5 during the night and lower than 0.2 during the day. A simple model has been proposed for the parameters of both expressions, consisting of an addition of three harmonic functions with periods of 1 year, 1 day and half a day. Hourly wind speed medians were successfully fitted at the heights of interest, although the fit proved better for the power law. Finally, a slight decrease in fitting at increasing heights was also observed.     

Classifying rotor‐layer wind to reduce offshore available power uncertainty

Rotor‐layer wind resource and turbine available power uncertainties prior to wind farm construction may contribute to significant increases in project risk and costs. Such uncertainties exist in part due to limited offshore wind measurements between 40 and 250 m and the lack of empirical methods to describe wind profiles that deviate from a priori, expected power law conditions. In this article, we introduce a novel wind profile classification algorithm that accounts for nonstandard, unexpected profiles that deviate from near power law conditions. Using this algorithm, offshore Doppler wind lidar measurements in the Mid‐Atlantic Bight are classified based on goodness‐of‐fit to several mathematical expressions and relative speed criteria. Results elucidate the limitations of using power law extrapolation methods to approximate average wind profile shape/shear conditions, as only approximately 18% of profiles fit well with this expression, while most consist of unexpected wind shear. Further, results demonstrate a relationship between classified profile variability and coastal meteorological features, including stability and offshore fetch. Power law profiles persist during unstable conditions and relatively weaker northeasterly flow from water (large fetch), whereas unexpected classified profiles are prevalent during stable conditions and stronger southwesterly flow from land (small fetch). Finally, the magnitude of the discrepancy between hub‐height wind speed and rotor equivalent wind speed available power estimates varies by classified wind‐profile type. During unexpected classified profiles, both a significant overprediction and underprediction of hub‐height wind available power is possible, illustrating the importance of accounting for site‐specific rotor‐layer wind shear when predicting available power.     

The influence of the wind speed profile on wind turbine performance measurements

To identify the influence of wind shear and turbulence on wind turbine performance, flat terrain wind profiles are analysed up to a height of 160 m. The profiles' shapes are found to extend from no shear to high wind shear, and on many occasions, local maxima within the profiles are also observed. Assuming a certain turbine hub height, the profiles with hub‐height wind speeds between 6 m s^?1 and 8 m s^?1 are normalized at 7 m s^?1 and grouped to a number of mean shear profiles. The energy in the profiles varies considerably for the same hub‐height wind speed. These profiles are then used as input to a Blade Element Momentum model that simulates the Siemens 3.6 MW wind turbine. The analysis is carried out as time series simulations where the electrical power is the primary characterization parameter. The results of the simulations indicate that wind speed measurements at different heights over the swept rotor area would allow the determination of the electrical power as a function of an ‘equivalent wind speed’ where wind shear and turbulence intensity are taken into account. Electrical power is found to correlate significantly better to the equivalent wind speed than to the single point hub‐height wind speed. Copyright © 2008 John Wiley & Sons, Ltd.     

Evaluation of three mainstream numerical weather prediction models with observations from meteorological mast IJmuiden at the North Sea

Numerical weather prediction models play an important role in the field of wind energy, for example, in power forecasting, resource assessment, wind farm (wake) simulations, and load assessment. Continuous evaluation of their performance is crucial for successful operations and further understanding of meteorology for wind energy purposes. However, extensive offshore observations are rarely available. In this paper, we use unique met mast and Lidar observations up to 315 m from met mast “IJmuiden,” located in the North Sea 85 km off the Dutch coast, to evaluate the representation of wind and other relevant variables in three mainstream meteorological models: ECMWF‐IFS, HARMONIE‐AROME, and WRF‐ARW, for a wide range of weather conditions. Overall performance for hub‐height wind speed is found to be comparable between the models, with a systematic wind speed bias <0.5 m/s and random wind speed errors (centered RMSE) <2 m/s. However, the model performance differs considerably between cases, with better performance for strong wind regimes and well‐mixed wind and potential temperature profiles. Conditions characterized by moderate wind speeds combined with stable stratification, which typically produce substantial wind shear and power fluctuations, lead to the largest misrepresentations in all models.     

Predictions of ice formations on wind turbine blades and power production losses due to icing

Prediction of ice shapes on a wind turbine blade makes it possible to estimate the power production losses due to icing. Ice accretion on wind turbine blades is responsible for a significant increase in aerodynamic drag and decrease in aerodynamic lift and may even cause premature flow separation. All these events create power losses and the amount of power loss depends on the severity of icing and the turbine blade profile. The role of critical parameters such as wind speed, temperature, liquid water content on the ice shape, and size is analyzed using an ice accretion prediction methodology coupled with a blade element momentum tool. The predicted ice shapes on various airfoil profiles are validated against the available experimental and numerical data in the literature. The error in predicted rime and glime ice volumes and the maximum ice thicknesses varies between 3% and 25% in comparison with the experimental data depending on the ice type. The current study presents an efficient and accurate numerical methodology to perform an investigation for ice‐induced power losses under various icing conditions on horizontal axis wind turbines. The novelty of the present work resides in a unified and coupled approach that deals with the ice accretion prediction and performance analysis of iced wind turbines. Sectional ice profiles are first predicted along the blade span, where the concurrence of both rime and glaze ice formations may be observed. The power loss is then evaluated under the varying ice profiles along the blade. It is shown that the tool developed may effectively be used in the prediction of power production losses of wind turbines at representative atmospheric icing conditions.     

Validation of boundary layer parameterization schemes in the weather research and forecasting model under the aspect of offshore wind energy applications— Part I: Average wind speed and wind shear

Five different planetary boundary layer (PBL) schemes in the weather research and forecasting model have been tested with respect to their capability to model boundary layer parameters relevant for offshore wind deployments. For the year 2005 model simulations based on the Yonsei University, asymmetric convection model version 2, quasi‐normal scale elimination, Mellor–Yamada–Janjic and Mellor–Yamada–Nakanishi–Niino PBL schemes with weather research and forecasting have been performed for the North Sea and validated against measurements of the Forschungsplattformen in Nord‐ und Ostsee Nr.1 platform. The investigations have been focused on the key parameters 100 m mean wind speed and wind shear expressed by the power law exponent α. All PBL‐schemes are doing well in reproducing averages and average annual statistics of the 100 m wind speed. However, two of the schemes (Yonsei University and Mellor–Yamada–Nakanishi–Niino) overestimate the wind speed above 15 m s^?1 systematically. The results for the power law wind profile show a large variability between the models and the observations for different atmospheric stability conditions and also differ a lot from the industry standards. Overall, the Mellor–Yamada–Janjic scheme performs slightly better than the others and is suggested as first choice for marine atmospheric boundary layer simulations without apriori information of atmospheric stability in the region of interest. Copyright © 2014 John Wiley & Sons, Ltd.     

Hybridizing the fifth generation mesoscale model with artificial neural networks for short-term wind speed prediction

This paper presents the hybridization of the fifth generation mesoscale model (MM5) with neural networks in order to tackle a problem of short-term wind speed prediction. The mean hourly wind speed forecast at wind turbines in a wind park is an important parameter used to predict the total power production of the park. Our model for short-term wind speed forecast integrates a global numerical weather prediction model and observations at different heights (using atmospheric soundings) as initial and boundary conditions for the MM5 model. Then, the outputs of this model are processed using a neural network to obtain the wind speed forecast in specific points of the wind park. In the experiments carried out, we present some results of wind speed forecasting in a wind park located at the south-east of Spain. The results are encouraging, and show that our hybrid MM5-neural network approach is able to obtain good short-term predictions of wind speed at specific points.     

Interaction between heterogeneous thermal stratification and wakes of wind turbine arrays

The thermal heterogeneity between the land and sea might affect the wind patterns within wind farms (WF) located near seashores. This condition was modeled with a large-eddy simulation of a numerical weather prediction model (Weather Research and Forecasting) that included the wind turbine actuator disk model (ADM). The assumed condition was that the downstream surface temperature was relatively higher (unstably stratified condition) than the neutrally stratified upstream wind. Under this condition, a thermal internal boundary layer (TIBL) was developed from an area where a step-changed surface temperature was implemented. The combined effect of the wake deficit due to the WF and velocity recovery as a result of enhanced mixing under unstable stratification showed significant modulation of the wind speed at the hub height when local atmospheric stability affected the wind turbine (WT). We show that TIBL height depends on the variables to be evaluated as the threshold. A precise prediction of the TIBL height is beneficial for better estimation of power generation. A prediction model was proposed as an extension of the internal boundary layer (IBL) model for neutral stratification, and the results tracked TIBL development reasonably well. The effects of WFs on surface properties (e.g., friction velocity, heat flux, and Obukhov length) and the tendency of IBL growth were minor. A single WT wake was also assessed under several TIBL developmental stages (i.e., location) and thermal stratification conditions. The standard deviation of the wake deficit increased vertically during the development stage of the TIBL. In contrast, the coefficients in the horizontal and vertical directions were comparable when the WT was deep inside the TIBL.     

Atmospheric stability‐dependent infinite wind‐farm models and the wake‐decay coefficient

We extend the infinite wind‐farm boundary‐layer (IWFBL) model of Frandsen to take into account atmospheric static stability effects. This extended model is compared with the IWFBL model of Emeis and to the Park wake model used in Wind Atlas Analysis and Application Program (WAsP), which is computed for an infinite wind farm. The models show similar behavior for the wind‐speed reduction when accounting for a number of surface roughness lengths, turbine to turbine separations and wind speeds under neutral conditions. For a wide range of atmospheric stability and surface roughness length values, the extended IWFBL model of Frandsen shows a much higher wind‐speed reduction dependency on atmospheric stability than on roughness length (roughness has been generally thought to have a major effect on the wind‐speed reduction). We further adjust the wake‐decay coefficient of the Park wake model for an infinite wind farm to match the wind‐speed reduction estimated by the extended IWFBL model of Frandsen for different roughness lengths, turbine to turbine separations and atmospheric stability conditions. It is found that the WAsP‐recommended values for the wake‐decay coefficient of the Park wake model are (i) larger than the adjusted values for a wide range of neutral to stable atmospheric stability conditions, a number of roughness lengths and turbine separations lower than ～ 10 rotor diameters and (ii) too large compared with those obtained by a semiempirical formulation (relating the ratio of the friction to the hub‐height free velocity) for all types of roughness and atmospheric stability conditions. © 2013 The Authors. Wind Energy published by John Wiley & Sons, Ltd.