Data efficient measure-correlate-predict approaches to wind resource assessment for small-scale wind energy

The feasibility of predicting the long-term wind resource at 22 UK sites using a measure-correlate-predict (MCP) approach based on just three months onsite wind speed measurements has been investigated. Three regression based techniques were compared in terms of their ability to predict the wind resource at a target site based on measurements at a nearby reference site. The accuracy of the predicted parameters of mean wind speed, mean wind power density, standard deviation of wind speeds and the Weibull shape factor was assessed, and their associated error distributions were investigated, using long-term measurements recorded over a period of 10 years. For each site, 120 wind resource predictions covering the entire data period were obtained using a sliding window approach to account for inter-annual and seasonal variations. Both the magnitude and sign of the prediction errors were found to be strongly dependent on the season used for onsite measurements. Averaged across 22 sites and all seasons, the best performing MCP approach resulted in mean absolute and percentage errors in the mean wind speed of 0.21 ms⁻¹ and 4.8% respectively, and in the mean wind power density of 11 Wm⁻² and 14%. The average errors were reduced to 3.6% in the mean wind speed and 10% in the mean wind power density when using the optimum season for onsite wind measurements. These values were shown to be a large improvement on the predictions obtained using an established semi-empirical model based on boundary layer scaling. The results indicate that the MCP approaches applied to very short onsite measurement periods have the potential to be a valuable addition to the wind resource assessment toolkit for small-scale wind developers.     

The round robin site assessment method: A new approach to wind energy site assessment

Portability is one of the many potential advantages of utilizing ground-based measurement devices such as SODARs and LIDARs instead of meteorological towers for wind resource assessment. This paper investigates the use of a monitoring strategy that leverages the portability of ground-based devices, dubbed the “round robin site assessment method.” The premise is to measure the wind resource at multiple sites in a single year using a single portable device, but to discontinuously distribute the measurement time at each site over the whole year, so that the total measurement period comprises smaller segments of measured data. This measured data set is then utilized in the measure-correlate-predict (MCP) process to predict the long-term wind resource at the site. This method aims to increase the number of sites assessed in a single year, without the sacrifice in accuracy and precision that usually accompanies shorter measurement periods. The performance of the round robin site assessment method was compared to the standard method, in which the measured data are continuous. The results demonstrate that the round robin site assessment method is an effective monitoring strategy that improves the accuracy and reduces the uncertainty of MCP predictions for measurement periods less than 1 year. In fact, the round robin site assessment method compares favorably to the accuracy and uncertainty of a full year of resource assessment. While there are some tradeoffs to be made by using the round robin site assessment method, it is potentially a very useful strategy for wind resource assessment.     

Comparison between the bivariate Weibull probability approach and linear regression for assessment of the long-term wind energy resource using MCP

A detailed investigation of a measure–correlate–predict (MCP) approach based on the bivariate Weibull (BW) probability distribution of wind speeds at pairs of correlated sites has been conducted. Since wind speeds are typically assumed to follow Weibull distributions, this approach has a stronger theoretical basis than widely used regression MCP techniques. Building on previous work that applied the technique to artificially generated wind data, we have used long-term (11 year) wind observations at 22 pairs of correlated UK sites. Additionally, 22 artificial wind data sets were generated from ideal BW distributions modelled on the observed data at the 22 site pairs. Comparison of the fitting efficiency revealed that significantly longer data periods were required to accurately extract the BW distribution parameters from the observed data, compared to artificial wind data, due to seasonal variations. The overall performance of the BW approach was compared to standard regression MCP techniques for the prediction of the 10 year wind resource using both observed and artificially generated wind data at the 22 site pairs for multiple short-term measurement periods of 1–12 months. Prediction errors were quantified by comparing the predicted and observed values of mean wind speed, mean wind power density, Weibull shape factor and standard deviation of wind speeds at each site. Using the artificial wind data, the BW approach outperformed the regression approaches for all measurement periods. When applied to the real wind speed observations however, the performance of the BW approach was comparable to the regression approaches when using a full 12 month measurement period and generally worse than the regression approaches for shorter data periods. This suggests that real wind observations at correlated sites may differ from ideal BW distributions and hence regression approaches, which require less fitting parameters, may be more appropriate, particularly when using short measurement periods.     

Errors and uncertainties associated with missing wind data and short records

A near‐complete 4 year data set of 10 min average 80 m wind speeds is used to examine the impact of missing data on monthly and yearly estimates of mean wind speed and energy production from a generic wind turbine. Missing data is a source of uncertainty in wind energy resource assessment studies. Quantifying that uncertainty can improve the reliability of P90 and related wind farm energy production estimates. An empirical relationship between missing data percentage and relative uncertainty in monthly mean wind speed is derived. Relationships between uncertainties in monthly average wind speed and uncertainties in monthly energy production are also explored. In many cases with monthly data losses of 10% or less the contribution to the overall uncertainty in annual energy production will be small (<1%), but with substantial losses in cold winters, typically caused by icing; the uncertainties can become more significant. The data set is also used to indicate uncertainties associated with short data periods. Annual average wind speed estimates based on less than a complete year's data also add significant uncertainty to wind resource assessments. Copyright © 2013 John Wiley & Sons, Ltd.     

Assessment of wind energy potentiality at Kudat and Labuan, Malaysia using Weibull distribution function

The wind resource is a crucial step in planning a wind energy project and detailed knowledge of the wind characteristic at a site is needed to estimate the performance of a wind energy project. In this paper, with the help of 2-parameter Weibull distribution, the assessment of wind energy potentiality at Kudat and Labuan in 2006-2008 was carried out. “WRPLOT” software has been used to show the wind direction and resultant of the wind speed direction. The monthly and yearly highest mean wind speeds were 4.76 m/s at Kudat and 3.39 m/s at Labuan respectively. The annual highest values of the Weibull shape parameter (k) and scale parameter (c) were 1.86 and 3.81 m/s respectively. The maximum wind power density was found to be 67.40 W/m² at Kudat for the year 2008. The maximum wind energy density was found to be 590.40 kWh/m²/year at Kudat in 2008. The highest most probable wind speed and wind speed carrying maximum energy were estimated 2.44 m/s at Labuan in 2007 and 6.02 m/s at Kudat in 2007. The maximum deviation, at wind speed more than 2 m/s, between observed and Weibull frequency distribution was about 5%. The most probable wind directions (blowing from) were 190° and 269° at Kudat and Labuan through the study years. From this study, it is concluded that these sites are unsuitable for the large-scale wind energy generation. However, small-scale wind energy can be generated at the turbine height of 100 m.     

Potential for wind generation on the Guyana coastlands

Guyanas dependence upon imported petroleum fuels can only be offset by the sustained exploitation of its indigenous resources. With its populated coastlands exposed to the northeast trade winds and a history of small-scale wind energy utilisation wind is one such potential energy source. In this study, the coastal wind regime is analysed and historical data from a coastal weather station are used to estimate the potential for wind generation. It is found that a hybrid Weibull probability density function best describes the annual wind speed frequency distribution at the reference height of 10.67 m. With an annual mean wind speed of 5.8 ms, an energy pattern factor of 1.41, and an annual average power density of 159 Wm², this distribution represents a class-3 wind resource, suitable for most wind turbine applications. Site analysis and observed trends in coastal wind availability suggest the strong likelihood of a greater wind resource in more open locations. In view of its apparent potential for wind farm operation, a comprehensive, wind resource assessment programme is recommended for the Guyana coastlands.     

Electricity generation and wind potential assessment at Hurghada, Egypt

A technical and economic assessment has been made of the generation of electricity using wind turbines at one of the most promising wind sites in Egypt: Hurghada. In this paper, we used wind data recorded over 23 years for this site. The WASP program was used to calculate the values of wind speed frequency for the station, their seasonally values have been estimated and compared with measured data.Weibull parameters and the power law coefficient (n) for all seasons at different heights (10–70 m) has been estimated and used to describe the distribution and behavior of seasonal wind speed and their frequencies at Hurghada. The monthly and annual values of wind potential at a height of 70 m were obtained by extrapolation of the 10 m data from the results of our previous article [Ahmed Shata AS, Hanitsch R. The potential of electricity generation on the east coast of Red Sea in Egypt. Renew Energy 2006;31:1597–615] using the power law.Also, the monthly plant load factor (PLF) has been estimated, which is used to determine the expected annual energy output of a wind energy conversion system (WECS).Variation of annual capacity factor with rated wind speed for 10 different wind turbines has been studied. The lower the rated speed for the WECS of the same height, the higher will be the capacity factor values. The expected electrical energy cost of kWh produced by the wind turbine (Repower MM82) with a capacity of 2 MW considered for Hurghada station was found to be less than 1.5 € cent/kWh.     

Implications of seasonal and diurnal variations of wind velocity for power output estimation of a turbine: a case study of Grenada

This case study highlights the importance of taking into consideration diurnal variations of wind velocity for wind energy resources assessment. Previous studies of wind energy distribution that are based on the two-parameter Weibull density function have so far neglected to consider time of day fluctuations in wind speed, instead concentrating primarily on seasonal deviations. However, this has serious implications where such a wind energy model is the underpinning of calculations for the potential power production from a wind turbine and in particular where the timing of the energy output is essential to meet electricity loads. In the case of Grenada the energy output from a wind turbine during the day is approximately two times the output at night thereby fluctuating enormously around the seasonal mean distribution. When this is not taken into account the economic and technological viability of a wind turbine project may be overestimated or not even be identified. This work shows how a wind energy resources assessment based on the Weibull distribution model can be done and how the power output of a horizontal axis turbine is calculated. An analysis of the recorded wind data confirms the application of the Weibull density function as a suitable tool for modelling wind regimes. Copyright © 2003 John Wiley & Sons, Ltd.     

测风塔选址对复杂地形风电场风资源评估的影响

为研究复杂地形条件下风电场测风塔的代表性及其对风资源评估的影响,以陕西省靖边县境内某风电场为例,选取3座测风塔资料,利用WindSim软件模拟分析了2011年风电场风能资源分布,并估算了风电场年发电量。结果表明,复杂地形风电场处测风塔数量较少时风资源评估结果的不确定性显著增加,而在考虑地形因素情况下测风塔数量增多,估算发电量更为准确。在地形较为复杂的风电场应根据地形条件布设适当数量测风塔,以得到风电场内较为精准的风资源分布,减少因测风塔位置选择而造成的风资源评估的不确定性。     

Wind power statistics and an evaluation of wind energy density

In this paper the statistical data of fifty days' wind speed measurements at the MERC-solar site are used to find out the wind energy density and other wind characteristics with the help of the Weibull probability distribution function. It is emphasized that the Weibull and Rayleigh probability functions are useful tools for wind energy density estimation but are not quite appropriate for properly fitting the actual wind data of low mean speed, short-time records. One has to use either the actual wind data (histogram) or look for a better fit by other models of the probability function.     

Wind energy resource assessment for five locations in Saudi Arabia

The analysis of recently collected wind data at five sites in Saudi Arabia namely, Dhulum, Arar, Yanbu, Gassim and Dhahran is presented. The five sites represent different geographically and climatologically conditions. The data collected over a period spanned between 1995 and 2002 with different collection periods for each site. Daily, monthly and frequency profiles of the wind speed at the sites showed that Dhulum and Arar sites have higher wind energy potential with annual wind speed average of 5.7 and 5.4 m/s and speeds higher than 5 m/s for 60 and 47% of the time, respectively. The two sites are candidates for remote area wind energy applications. The costal site's, i.e. Yanbu and Dhahran wind speed data indicated that the two sites have lower annual wind speed averages and wind blows at speed higher than 5 m/s during afternoon hours. That makes the two sites candidates for grid connected wind systems for electrical load peak shaving. The data of Gassim site showed that the site has the lowest wind energy potential compared to the others. The annual energy produced by a Nordex N43 wind machine is estimated to be 1080, 990, 730, 454 and 833 MWh for Dhulum, Arar, Yanbu, Gassim and Dhahran, respectively. The analysis showed that the estimated annual energy produced by the machine based on 10 min averaged data is 2.5% higher than the estimated energy based on 30 min averaged data.     

Wind energy potential in Aden-Yemen

The wind energy resource is very large and widely distributed throughout the world as well as in Yemen. Aden possesses a very good potential of wind energy. In this article a number of years data on wind speed in Aden has been studied and presented. A statistical analysis was carried out from which the annual wind speed was found to be 4.5 m/s and most of the time the wind speed is in the range of 3.5–7.5 m/s. The wind speed distributions were represented by Weibull and Rayleigh distributions. It was found that the Rayleigh distribution is suitable to represent the actual probability of wind speed data for Aden. The wind speed data showed that the maximum monthly wind speed occurs in the month of February with the maximum in the month of June. It is concluded that Aden can be explored for wind energy applications.     

Analysis of wind potential and characteristics in North Aegean, Greece

Wind data for the years 2000 and 2001 were analyzed to evaluate the wind potential of the Mikra–Thessaloniki, region in northern Greece. The objective of the analysis was the establishment of the required criteria to answer the question: “are the renewable energy sources capable to maintain the operation of a desalination pilot unit?” The polar diagrams of the wind (wind speed, frequency, direction), the mean monthly and annual wind speed profile and the Weibull distributions for the years 2000 and 2001 are presented.     

Application of Bayesian model averaging in modeling long-term wind speed distributions

Accurate estimation of wind speed distribution is critical to the assessment of wind energy potential, the site selection of wind farms, and the operations management of wind power conversion systems. This paper proposes a new approach for deriving more reliable and robust wind speed distributions than conventional statistical modeling approach. This approach combines Bayesian model averaging (BMA) and Markov Chain Monte Carlo (MCMC) sampling methods. The derived BMA probability density function (PDF) of the wind speed is an average of the model PDFs included in the model space weighted by their posterior probabilities over the sample data. MCMC method provides an effective way for numerically computing marginal likelihoods, which are essential for obtaining the posterior model probabilities. The approach is applied to multiple sites with high wind power potential in North Dakota. The wind speed data at these sites are the mean hourly wind speeds collected over two years. It is demonstrated that indeed none of the conventional statistical models such as Weibull distribution are universally plausible for all the sites. However, the BMA approach can provide comparative reliability and robustness in describing the long-term wind speed distributions for all sites, while making the traditional model comparison based on goodness-of-fit statistics unnecessary.     

Wind characteristics and wind energy potential in Morocco

The wind characteristics of 11 sites in the windy regions in Morocco have been analysed. The annual average wind speed for the considered sites ranged from 5 m/s to 10 m/s and the average power density from 100 W/m² to 1000 W/m², which might be suitable for electrical power production by installing wind farms. On an annual scale the observations of the distribution of hourly wind speed are better fitted by the Weibull hybrid distribution in contrast to the Weibull distribution.The wind power is estimated to be 1817 MW, that is to say, the exploitable wind energy is 15198 GWh, which represents theoretically 11% of the total consumed energy in Morocco in 1994.     

Site‐specific Design Optimization of Wind Turbines

This article reports results from a European project, where site characteristics were incorporated into the design process of wind turbines, to enable site‐specific design. Two wind turbines of different concept were investigated at six different sites comprising normal flat terrain, offshore and complex terrain wind farms. Design tools based on numerical optimization and aeroelastic calculations were combined with a cost model to allow optimization for minimum cost of energy. Different scenarios were optimized ranging from modifications of selected individual components to the complete design of a new wind turbine. Both annual energy yield and design‐determining loads depended on site characteristics, and this represented a potential for site‐specific design. The maximum variation in annual energy yield was 37% and the maximum variation in blade root fatigue loads was 62%. Optimized site‐specific designs showed reductions in cost of energy by up to 15% achieved from an increase in annual energy yield and a reduction in manufacturing costs. The greatest benefits were found at sites with low mean wind speed and low turbulence. Site‐specific design was not able to offset the intrinsic economic advantage of high‐wind‐speed sites. It was not possible to design a single wind turbine for all wind climates investigated, since the differences in the design loads were too large. Multiple‐site wind turbines should be designed for generic wind conditions, which cover wind parameters encountered at flat terrain sites with a high mean wind speed. Site‐specific wind turbines should be designed for low‐mean‐wind‐speed sites and complex terrain. Copyright © 2002 John Wiley & Sons, Ltd.     

Review of the use of Numerical Weather Prediction (NWP) Models for wind energy assessment

Wind energy resource assessment applications require accurate wind measurements. Most of the published studies used data from existing weather station network operated by meteorological departments. Due to relatively high cost of weather stations the resolution of the weather station network is coarse for wind energy applications. Typically, meteorological departments install weather stations at specific locations such as airports, ports and areas with high density population. Typically, these locations are avoided during wind farms siting. According to WMO regulations, weather stations provide measurements for different weather elements at specific altitudes such as 2 m for air temperature and 10 m for wind measurements. For wind energy resource assessment applications, minimum of one year of wind measurements is required to build wind climatology for a certain site. Therefore data collected from a certain site cannot be used before one year of operation. Due to these limitations, wind energy resource assessment application needs to use data from different sources. Recently, wind assessment studies were conducted using data generated by Numerical Weather Prediction models. This paper reviews the use of the Numerical Weather Prediction data for wind energy resource assessment. It gives a general overview of NWP models and how they overcome the limitations in the classical wind measurements.     

Wind energy estimation at Quetta

Hourly wind data at Quetta airport (Samungli) for the years 1984–1985, were obtained using a standard anemometer height of 10 m, with a view to work out the feasibility of wind energy utilization. Quetta (lat. 30°11 'N long. 66°57'E) is elevated at 1799 m above sea level. In this paper, we analyse the wind energy data by using the Weibull distribution. Scaling and shaping parameters are determined by using least-squares approximation to a straight line. Actual wind data, weighted Weibull density function and weighted Rayleigh probability density function for regular and continuous periods of 4 weeks up to a year (wind characteristics are being recorded and summarized as diurnal and monthly wind velocity distributions and wind power density roses) are plotted which shows that the Weibull distribution is generally of the right shape to fit low-averaged wind speed frequency curves. However, density function of a normal distribution is also determined. Deviations in wind speed distributions at very low-averaged wind speeds and at comparatively large-averaged wind speeds are found.     

Offshore wind resource assessment with WAsP and MM5: comparative study for the German Bight

In this study, two different approaches to estimate the wind resource over the German Bight in the North Sea are compared: the mesoscale meteorological model MM5 and the wind resource assessment program WAsP. The dynamics of the atmosphere of the year 2004 was simulated with the MM5 model, with input from the NCEP global model, without directly utilizing measurement data. WAsP estimations were calculated on the basis of six measurement stations: three on islands, two offshore and one onshore. The annual mean wind speed at onshore, offshore and island sites is estimated by both models. The predictions are compared both with each other and with measured data. A spatial comparison of the wind resource calculated by the two models is made by means of a geographical information system. The results show that the accuracy of the WAsP predictions depends mainly on the measurement station used as input. Small differences are shown in the estimations performed by the three island stations, despite the large geographical distance between them. Compared with the measurements of the offshore sites, they seem to be suitable for estimating the offshore wind resource from measurements on land. The two offshore stations show differences when predicting each other's mean wind speed with the WAsP method, while the MM5 calculations show a similar deviation for both sites. The largest differences between the two models are found at distances of 5–50km from the coast. While in WAsP the increase occurs in the first 10km from the coast, MM5 models an increase due to coastal effects for at least 50km. Copyright © 2006 John Wiley &Sons, Ltd.     

新疆达坂城风电场风能资源特性分析

对新疆达坂城风电场的风能资源特性进行了详细的研究。基于在达坂城风电场实测的10m和24m高程的10min平均风速数据,分析了原始风速的分布特性。根据地表风速沿高度呈风剪指数分布的特性,计算了在各个轮毂高度上的风速分布。采用最小误差逼近算法原理,计算了风速韦布尔分布的参数以及平均风速和分布方差。通过对韦布尔分布的分析,计算了各个高度上风电场的平均风功率密度、有效平均风功率密度和可利用小时数等风能资源特性参数,为当地的风能开发提供分析基础。