Skip to main content
EnglishEN
European Commission logo print header
Content archived on 2024-04-16

Wind measurements and modelling in complex terrain

Objective

To deeper knowledge of local wind characteristics with focus on regions of complex topography where concentration effects can be very important (valleys, hill tops, islands, etc.).
In close cooperation with some of the participants measuring points in complex terrain for continuous measurements and campaigns have been selected. Equipment has in some cases been provided.

The turbulence reference data work has continued using the Lammefjord data.

The Wind Atlas Analysis and Application Programme (WAsP) model has been further improved, especially with regard to the graphical performance and handling of 3-dimensional topographical data input.

It is attempted to model flow over an escarpment with separation using kappa epsilon model and comparing with measured data.

Wind speed measurements are the foundation for establishing wind resource maps and determining the power curves of wind turbines and their power production. Most measurements are performed by means of cup anemometers hence it is of extreme importance to have an exact understanding of their behaviour in natural wind. 2 studies have been undertaken:
the influence of the mounting setup on the measurements;
causes of overspeeding.
The result of the first study showed that the mounting very often has a rather large influence on the measurements. This effect is often wrongly neglected. Guidelines have been made for anemometer mounting and anemometer design.

The result of the second study points to data processing error being the most prominant cause of overspeeding. The largest contribution to the data processing error comes from neglecting the influence of the cross wind component on the cup anemometer performance.

The European wind Atlas was produced by means of the Wind Atlas Analysis and application Programme (WAsP). However, there are clear limitations to the use of WAsP. On the micro scale very steep slopes and separation are not treated correctly and on the large scale the dynamics are too simplified by the neglect of the large scale stratification and also in relation to local (eg thermally driven) flow systems. Time has now come to start considering the best pres ent dynamic mesoscale numerical models and from there work in parallel on:
simplifying if possible (considering the use in relation to wind energy);
establishing methods for the use of such models.
2 mesoscale models have been selected:
the United Kingdom Meteorological Office Mesoscale Model;
the University of Karlsruhe Mesoscale Model.

The measurement campaign is running as it was planned. Most of the 12 automatic stations have been running satisfactorily.

Measurements are made at 3 different sites:
Arctic fells in northern Finland;
coastline on the southern coast;
in the Aaland archipelago.

The first results of the measurements and tests of wind atlas analysis and application programme (WASP) in the arctic fells were reported. WASP proved it's ability to describe the wind climate upon the fells when the true inversion height (height of the surface layer) was used. The results are even better if monthly data instead of annual wind speed distribution is used.

The main problem in the archipelago and on the coastline is the determination of surface roughness. Islands can be a part of areal roughness or they can be handled as obstacles or rough hills. The areal roughness can be calculated by many different equations giving different results. There are also quite few statistics about the variation of inversion height above sea level. Typically the flow field close to the surface is stable during spring and more or less unstableduring the autumn.

A 30 m mast has been instrumented and erected at the CRES test site in Aghia Marina to test and correct possible instrumentation and data logging problems. Detailed measurements of UVW wind component power spectra have been collected and analyzed. Friction velocities, roughness heights and deviations from the logarithmic law have been computed.

An inland location (Kalivari) on the island of Andros has been chosen as the test site. 3 30 m masts, 4 10 m masts and 1 20 m mast have been instrumented and erected at that site.

On all these masts, 10 minute averages of wind speed, direction and standard deviation have been recorded for 6 to 10 months continuously. A first campaign to measure turbulence parameters, and notably spectra, has been carried out. A large data set has been collected which has been reduced to a large number of measurements with a sonic anemometer at 30 m height of the summit mast.

The analysis of the power spectra of the sonic anemometer brought forth a number of discrepancies with usual formulation of the spectra. Substantial sensitivity to the direction has been found as befits the terrain complexity.

A comparison of the standard deviation accuracy of the cup anemometers used to equivalent measurements by a sonic anemometer has been carried out.

A number of comparisons with various codes, most notably wind atlas analysis and application programme (WAsP) and a mass consistent one have been carried out.

A study to ascertain the effect of a number of the model parameters on the accuracy is now being carried out. Some of the possible approaches to improve accuracy that are being examined include splitting of the input site wind data according to time of day, focusing grid techniques, modelling of separation zones and variable flow field initiation.

The region of Portugal that is being studied is the northernmost part of the country, at latitudes above 40 degrees where the highest mountains are concentrated.
In the current project, 10 points were selected for installation of anemometers. This limited number precluded the possibility of studying the whole area in detail. For practical reasons, perference was given to locations closer to Porto.

In each selected site a 10 m mast was erected, composed of a lattice tower, kept in position by 3 steel cables and a cross boom that supports the cup anemometer and direction sensor. The base of the mast and the cables are anchored to concrete foundations produced on site. The measuring and recording equipment was supplied, on a rental basis, and installed by Risoe National Laboratory. Each station comprises a cup anemometer (Risoe 70) a direction sensor (Aanderra Instruments 2750), a sensor scanner board (Aanderra Instruments 3010), a wind gust board, a data storage unit (Aanderra Instruments 2990), a battery and interconnecting cables. The extraction of data from the data storage units is performed at Porto, using a reading unit (Aanderra Instruments 2995).

One of the programmed tasks is the production of digitized terrain data for the area surrounding each station. This is currently being carried out with the help of a digitizing tablet connected to a personal computer (PC) microcomputer. The data is taken from topographic maps in 1:25000 scale and transferred into computer files under operation of wind atlas analysis and afflication programme (WASP). Due to limitations in the size of the digitizing tablet, a separate file is built for a rectangle of 5 by 10 km surface area, containing all the contour lines at 50 m level intervals. WASP is able to subsequently process data from several files, in order to construct maps covering larger areas.

The wind data collected has been processed and preliminary results were compiled. For each station the mean wind velocit y and wind power density (at 10 m above ground level), the maximum recorded 10 minute average and the maximum gust (3 second average), are recorded. The wind roses for all stations (12 sectors) are also recorded.

At Acqua Spruzza 11 wind stations, 15 m above ground level, have been installed. Wind data have been collected, validated and stored. Each wind station includes a combined cup anemometer and wind vane (Thies), and a data acquisition unit (SEB-Schlumberger Model VIT 2000) that records the following data, on the basis of a 10 minute integration time:
mean wind speed and standard deviation;
maximum wind speed;
sector of wind direction with maximum occurrence (out of a total of 16 sectors).
The sampling rate is 0.125 Hz.

Statistical data processing has been carried out according to standard procedures.

A 40 m tower with 3 wind sensors at 3 heights (15 m, 30 m and 40 m) and a full set of meteorological sensors has also been installed at Acqua Spruzza. The profile of the wind speed over the top of the hill (where the Acqua Spruzza test station will be set up) is now being estimated using wind data recorded on the 40 m mast.

A study of the wind flow on the area has been carried out using wind atlas analysis and application programme (WAsP) and AVENU simulation codes. Work has been done to further improve the orography and the roughness descriptions (adjustment and integration of the digitized contour lines, investigations on the spot). Predictions obtained from the models have been compared with the measured values.

According to first results, substantial improvements appear to be necessary if codes have to be utility tools for micrositing in complex terrain. Wind data so far collected at Acqua Spruzza might be used for this purpose within the framework of the extension of this project. Diskettes with these data are ready to be prepared as soon as a suitable format is agreed.

Wind measurements are being recorded on a hill facing prevailing winds, near the Mediterrean sea at Narbonne. The hill is situated in the north face of La Clape mountains, facing Tramontana winds. The upwind plain is very flat over more then 5 km, its altitude is 4 m above sea level; the hill summit is a 150 m above sea level, which gives an altitude difference of 145m and an average upwind slope greater then 0.2. Along hill slopes the ground is covered with bushes, vineyards and small pines. The plain is typically an open country around the reference site with some buildings and hedges.

In the plain, a reference mast of 24 m height is equipped 2 levels (10 m and 24 m) with respectively a vane propeller anemometer and a 3 directoral Gill propeller anemometer and temperature sensors. Mean wind speed, variances and direction are calculated with an average time of 10 minutes and recorded every hour. On the summit, a 40 m high mast has been erected and equipped at 3 levels:
at 10 m a vane propeller and a ventilated temperature sensor;
at 24 m a 3 directional Gill propeller;
at 40 m, a sonic anemometer and another ventilated temperature sensor. At the hill summit, wind speed is logged continuously with a scanning rate of 16 Hz for the sonic anemometer and 4 Hz for the propeller anemometers. Data are stored on optical disks that allows one month of continuous record.

A statistical analysis with the help of wind atlas analysis and afflication programme (WASP) has been applied to the wind speed data measured at 10 and 24m. To compare summit observations and these of the plain, mean wind speed (10 minutes average) at the summit, extracted from the continuous data every hour has been compared with plain observations. The average wind speed ratio between plain and hill for all directions, is 1.36 at 10 m above ground level, but at the upper level (24 m above ground level) this ratio is reduced to 1.27.

After the acquisition of more than one year of wind data time series, from 3 locations in the Iberian peninsula, models for the evaluation of wind climates as well as for validation exercises are now being applied.

Due to the different climatic and orographic characteristics of the zones, corresponding to the complex terrain type, the overall scope of work offers a wide range of application scenarios.

Ebro Valley is a moderate complex terrain which has a great advantage for energy exploitation. The general wind climate gives over 5 m/s of yearly average and some orographic effects may induce flow acceleration along the walley. The integration of the wind potential with the terrain availability implies that the zone offers an energy resource to consider for the medium and long term.

Galicia Coast is, on the contrary, a region with a high wind resource (around 7 m/s yearly average) but relatively low terrain availability for wind energy exploitation, the complexity of the terrain makes the evaluation of resources and the site selection tasks more difficult.

Southern Portugal is a complex terrain zone with a considerable wind energy potential at the south part, relatively low at the north part.

The one year wind climate for each zone and location are now complete and the inputs are being prepared for wind atlas analysis and application programme (WASP) model application by introducing the topography at several levels of definiton.

Data from nearby climatological stations have been collected and analyzed, for application of correlation methods.

The number of sites included in the measurement campaign was greatly increased on that originally planned by cooperation on an informal basis with a wind site assessment study being carried out by the Electricity Supply Board (ESB) and also by a private firm Arigna Fuels Ltd who are investigating the upland, inland area of Arigna. The informal cooperation and consultation took the form of agreement on:
the choice of sites (to obtain the best spread of the wind rich areas);
the choice of data loggers and instruments (NRG system 9600);
the choice of logging interval (10 minutes);
choice of height above ground level of the measuring instruments (10 m, 30 m and 50 m);
choice of logging period;
the sharing of data and site description information.
The installation program for the 10 minutes measurments is well advanced with a total of 14 sites operating at this time.

The data is quality checked by comparison with Met Service data. The data is assembled into weekly and monthly blocks and plotted out to check for any problems. The mean weekly windspeeds at 2 sites in the north of the country and KAN in the southwest were compared. The pattern of variation of the 2 sites is quite similar. However the winds are much higher at KAN than at CDY, the overall means at 10 m and 30 m for the period are 5.6 m/s and 8.2 m/s for CDY and 9.6 m/s and 10.6 m/s for KAN. The other difference is the large windshear effect at CDY (which is a site with a lot of vegetation, bushes and small trees in the area around the base of the mast) compared to KAN which is a well exposed hilltop site (altitude about 300 m) with little vegetation.

The KAN site shows the highest mean windspeed of all those sites being measured.
The cost of wind generated electricity and the cost, performance and lifetime of wind generators are strongly dependent on the wind conditions at the site of the generators. The data and the siting methods given in the European Wind Atlas provide a good basis for detailed resource assessment in most parts of the EC. However, the data coverage and quality can be significantly improved by new measurement programmes and further the siting methods can be refined to include estimates of turbulence, gusts and extreme values. The present multi-partner project addresses these issues, and is divided into the following tasks:

I. Risoe, DK. Coordination of project, collaboration with the other participants regarding instruments set up and data analysis in each of the measuring campaigns. Production of a users manual for climatological turbulence data. Further development of siting computer codes (WASP).

II. CRES, GR. Intensive measurements of turbulence and mean flow around a steep hill. Data analysis, model validation and improvement.

III. ENEL, I. Continuation and extension of measurements at Frosolone. Evaluation of existing models for predicting wind conditions at Frosolone.

IV. Universidade do Porto, P. Installation and operation of measuring equipment in areas where concentration effects are expected. Evaluation of existing models for predicting wind conditions at these conditions. Data analysis.

V. CNR-IFA, I. Study of Mediterranean wind climates. Analysis of data from an area across Italy. Operation of measuring stations. Data analysis.

VI. University College Dublin, EL. Installation and operation of measuring equipment at potential windfarm sites. Analysis of the effects of spatial diversity of the wind resource.

VII. CIEMAT, SP. Study of Iberian wind climate. Installation and operation of measuring equipment at potential windfarm sites. Estimation of site climatology and evaluation and evaluation of model predictions.

VIII. CSTB, F. Intensive measurements of turbulence and mean flow around a steep hill. Establishment of turbulence reference data for a hill top location. Data analysis.

IX. FMI, SF. Operation and analysis of data from a number of high masts. Study of icing conditions and the effects on wind turbine structures.

Fields of science

Programme(s)

Topic(s)

Data not available

Call for proposal

Data not available

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

RISOE NATIONAL LABORATORY
EU contribution
No data
Address
Frederiksborgvej 399
ROSKILDE
Denmark

See on map
Total cost
No data

Participants (10)

Share this page

Download

Last update: 7 May 1994

My Booklet

Permalink: https://cordis.europa.eu/project/id/JOUR0067

European Union, 2024