27.02.2013 Views

Wind Energy

Wind Energy

Wind Energy

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

13 Gusts in Intermittent <strong>Wind</strong> Turbulence 75<br />

Engineers [7]. However this scale belongs to the inertial range; it is usually<br />

not a characteristic time scale, so that the results obtained here are expected<br />

to be valid for any other scale τ belonging to the inertial range.<br />

Here we consider the recurrent times of gust events: the times between<br />

successive gust events. The recurrent times of meteorological events have<br />

practical importance, and correspond to the knowledge of some dynamical<br />

properties of wind fluctuations. Here we not only focus on the mean<br />

return times, but also on their pdf. We consider the evolution of this pdf<br />

with increasing velocity thresholds δ. For illustration purposes, we have<br />

taken an atmospheric turbulent velocity data base: velocity measurements<br />

taken 25 m from the ground in south-west France, with a sonic velocimeter<br />

sampling at 10 Hz (see [8, 9]). We analyze here three portions of duration<br />

55 min each, recorded under near-neutral stability conditions. A portion of the<br />

data is shown in Fig. 13.1: this illustrates the intermittent behavior of wind<br />

turbulence. The power spectrum of the 3 portions is shown in Fig. 13.2,<br />

in log–log plot, indicating a very nice −5/3 scaling law over most of the<br />

available dynamics (between 0.2s and about 10 min). We have then transformed<br />

the velocity time series into an amplitude increment time series:<br />

Y3(t) =|V (t + τ0) − V (t)| with τ0 =3s. The new time series is shown in<br />

Fig. 13.3, together with an example of threshold (δ =1 m −1 ), and the associated<br />

recurrent times. Successive return time form a new time series which<br />

is represented in Fig. 13.4.<br />

An interesting question is now to evaluate the tail behavior of the pdf of<br />

return times: since gust events are associated to large return times, their<br />

probability of occurrence is given by the tails. There are in fact many results<br />

providing the form of the tail of the pdf of recurrent times (also called<br />

first-passage times) for scaling processes. One of the most classical scaling<br />

U (m/s)<br />

6<br />

5<br />

4<br />

3<br />

2<br />

1<br />

0<br />

0<br />

50 100 150 200 250 300<br />

t (s)<br />

Fig. 13.1. A portion of the turbulent time series analyzed here for illustration<br />

purposes

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!