Experimental and measurement based model approach for site on demand wind energy management.

Achim, Moise ; Risteiu, Mircea ; Cabulea, Lucia 等

1. INTRODUCTION

Today's research developed numerous urban or insulated canopy schemes mesoscale models in order to approximate the drag and turbulent production effects of a city, or small hills on the air flow. Somehow, the little data exists by which to evaluate the efficacy of the schemes since "area-averaged" wind measurements in cities are difficult to obtain owing to the large number of wind sensors required to obtain a reasonable statistical sample. In this paper, we will describe the experimental approach and obtained data we have used, and we show vertical profiles of area-averaged wind speed for several realistic and idealized multi-building array configurations. We will finish by discussing how the area-averaged wind speed may change as a function of plan area and relief configuration, as main parameter that influences the Betz' Theorem (Kastner-Klein et al., 2000).

2. BACKGROUND

Groups of buildings, as well as small hills on average, act to slow down the wind through drag and obstacle deflection of the flow, regions of reverse flow in obstacle-induced circulations, and zones of calm winds between obstacles. A number of research groups have been adding canopy parameterizations to mesoscale models in order to approximate the sub-grid effects of small scale obstacles on the mean flow and turbulent kinetic energy fields (Brown & Gowardhan, 2006). The drag term, usually a function of the frontal area density of obstacles (f(z)), results in a mesoscale-model-produced wind speed profile:

u(z) = [u.sub.H] [e.sup.(a(z/H - 1]))] (1)

Where H is evaluation measurement point height, U is the wind at canopy height, and a is an attenuation coefficient proportional to the porosity of the canopy. We have introduced the evaluation point height at ax point of wind rotor, because the wind turbine rotor blades must take power over entire length of the blade, and the lift force must be minimized.

The wind field produced by the mesoscale model can be thought of as representing the average wind over the computational grid cell (generally on the order of 1 km in horizontal dimension). Since winds in small scale obstacles areas can be extremely complex, with flow on one side of the obstacle opposite to that on the other, a single vertical profile of wind speed measurements from a tower, for example, will not be representative of the mesoscale model grid cell value. The same situation is related to the fact that the wind layer thickness is variating very fast, when the vertical forces (figure 1) are variating in accordance with the position of measuring system (Kjersti Rokenes & Per-Age Krogstad, 2008).

[FIGURE 1 OMITTED]

For evaluating winds produced by mesoscale models, a large number of wind sensors at nearly the same height distributed horizontally over at least a several block area is required to obtain a "measurement" at a particular height. If multiple horizontal planes of instrumentation are available, then a vertical profile of these "areaaveraged" measurements can be used to evaluate the mesoscale model urban canopy parameterizations.

3. MEASUREMENT-BASED MODELING

Fot the selecting of the optimal vertical point of the wind rotor ax, the economical constrains must be also taken into consideration. (Bourgeois et al., 2008) have analysed such situation by using SODAR and LIDAR measurements methodology. Both data sets revealed almost constant turbulence intensities between 30 m and 100 m above ground. Turbulence intensities remained below class A of the IEC 61400. SODAR and LIDAR measurements were normalised to 50 m for all twelve 30[degrees] wind direction sectors. Determination of turbulence intensity is difficult due to different sampling rate. More investigations on calculation of turbulence intensities are needed.

For this reason, the instant vertical speed profil have to be determined. A low cost measuring system is presented next.

3.1 Measurement setup

The measurement setup is based on wireless data acquisition system, with a datalogger facility (WDAQ). The system comprises 3 main components: i) wireless sensor nodes which acquire and transmit data from thermistor-based anemometer (MF51E+ AD620), wind direction (CXTLA02), atmosphere temperature and humidity (MPX4115A serie), WDAQ altitude, ii) base station which receives and passes the data to a host, and iii) software which operates the system. When base station is not available, the WDAQ is swhitched into datalloger mode, where up to 2 millions of data could be stored. For five types of data, up to 400000 sets of data, could be stored. For a sampling rate of 2 samples/sec (Bourgeois et al., 2008), there is enough storage area for more de 55 hours of experiments. The WDAQ and measurement sensors are supplied by a 6 Ah accumulator. The WDAQ and sensors are rised up, and lowering down with a helium balloon. In order to ensure the measuring speed (0.5 m/sec), the diameter of balloon is 3 m. By running the anchorage system of the ballon, with a constant speed, the measuring system acquires data during rising and lowering stage.

3.2 Data interpretation

As far as we want to understand the behavior of the thickness of the wind layer, and taking into account the IEC 61400.1, IEC 61400.12 standards, we have collected 25 sets of data, in about 2 hours of measuring process. Each set of data has own vertical wind speed profile. For 25 sets of wind speed data we have calculated the mean wind speed. Nest figure (figure 2) shows the average of the wind speed according to the altitude of measuring point. On the same time, after data processing we have pointed some interesting information correlated with vertical wind speed (the two circles might suggest the wind generators diameters).

[FIGURE 2 OMITTED]

The figure 2 is pointing the major tendency of the thickness layer. Based on collected data, at 57.5 m altitude, the wind speed is getting over 5.6 m/sec. This state is up to 100 m. That means that, we can draw up a circle with the radius of 21.25 m. Theoretically, this radius might be the radius of a wind generator blade. According to (Alexiadis, & Dokopoulos, 1999), the power of the generator might be up to 250 kW.

Figuring the situation of taking into account the wind speed over 7 m/sec, the thickness of the wind layer is 24.5 m, which means that theoretically we can design a generator blade with 12.25 m radius. In this figure we must notice that, the middle of the wind layer thickness is situated at 80.5 m altitude.

4. CONCLUSION

Taking into account the filed research, when an urban, or small hiils become the places where we want to install wind generators, before long time measurements, as they are required IEC 61400.1, IEC 61400.12 standards, some preliminary estimation of the vertical wind speed profile has be done.

Because the focuses of the implementation are the small scale wind generators, a low cost instant speed measurement system is required. Vertical wind speed will estimate the thickness of the wind layer on the specific site. So, the long time measurement sensors must be placed inside of the relevant wind speed layer.

On the same time, based on our example, we already have a milestone of the feasibility study. This is related to the fact that, 5.6 m/sec we get into a wind layer of 42.5 m thickness, starting from 57.5 m. The main question here is: can we get enough energy to cover to implementing cost?

Our measuring approach offer one more preliminary information for wind generator costs implementation. On the other hand, we haven't analised the horizontal position of the measuring point (tower), as shown in figure 1. In order to control the lifting force on the blades, the exact position of the rotor position has to be determined.

5. REFERENCES

Alexiadis, M.C. & Dokopoulos, P.S (1999). Wind speed and power forecasting based on spatial correlation models, IEEE Transaction on Energy Conversion, Volume 14, Issue 3, Sept. 1999 Page(s):836-842

Allison M. Berg (2006). The feasibility of sodar wind profile measurements from an oceanographic buoy, B.S. United States Naval Academy, Available from: http://dspace.mit.edu/handle/1721.1/38510?show=full

Kastner-Klein, P. M. W. Rotach, M. J. Brown, E. Fedorovich and R. E. Lawson ( 2000). Spatial Variability of Mean Flow and Turbulence Fields in Street Canyons, 3rd AMS Urban Env. Symp. Davis, CA. LA-UR-00-3025

Kjersti Rokenes & Per-Age Krogstad (2008). Wind tunnel simulation of terrain effects on wind farm siting, Wind Energy, Volume 12, Issue 4, pages 391-410

Michael J. Brown & A. Gowardhan (2006). Experimental And Model-Computed Area-Averaged Vertical Profiles Of Wind Speed for Evaluation Of Mesoscale Urban Canopy Schemes, Sixth Symposium on the Urban Environment AMS Forum: Managing our Physical and Natural Resources: Successes and Challenges, AMS, Available from: http://ams.confex.com/ams/Annual2006/techprogram/paper _105229.htm

Saskia Bourgeois, Rene Cattin, Hans Winkelmeier, Ian Locker (2008). Cfd Modeling Of The Vertical Wind Profile And The Turbulence Structure Above Complex Terrain And Validation With Sodar And Lidar Measurements, Verein Energiewerkstatt, Friedburg, Austria (Public Report)

Torben Skov Nielsen&, Alfred Joensen (1999). A new reference for wind power forecasting, Wind Energy, Volume 1 Issue 1, Pages 29-34

Ulas Eminoglu & Bahtiyar Dursun (2008). Incorporation of a new wind turbine generating system model into distribution systems load flow analysis, Wind Energy, Volume 12, Issue 4, pages 375-390