5. Conclusions
The aim of a noise impact assessment is the estimation of both
the immission and residual components of the noise levels measured
in proximity of the receivers. Several difficulties in the noise
impact evaluation arise when the receiver is placed in a complex
orography, with many wind turbines, where the wind produces
high residual noise and has high and unpredictable shear.
The method proposed in this paper does not consider the
immission levels of a wind farm as a function of the wind speed
at hubs height, but instead as a function of a new parameter that
considers the rotation of all the turbines: the equivalent blades
rotational speed Neq. The residual noise level is evaluated as a function
of the vgr.
The proposed procedure can reconstruct these trends starting
from an acoustic and weather measurement campaign of at least
two weeks without requiring a plant shut down. The periods when
the wind turbines had rotational speed low enough not to increase
the noise at receiver are used as a first estimate of the residual
noise level. The final residual noise as a function of the vgr and
the immission levels as a function of Neq are computed through
appropriate iterative phases of energetic subtraction between the
measured noise levels grouped by classes of vgr and Neq.
The procedure was applied to measurement campaigns performed
in seven Italian wind farms. For all the farms, the maximum
immission noise level for both daytime and night time,
corresponding to the maximum Neq, was compared to the immission
noise level predicted by the NORD2000 noise model for the
maximum sound production conditions.
For all the wind farms, the immission noise levels resulted comparable
with the predicted ones within the uncertainties. On this
basis, the Federal Council of the National and Regional
Environmental Protection Agencies of Italy suggests this procedure
to verify the WTN compliance with the present Italian legislation,
nonetheless nothing prevents its use in other countries. Nevertheless,
a proper scientific validation is still needed in order to verify
the residual output. The results will be presented in a further article,
even if preliminary ones have been already shown [23]. In
order to enable an appropriate analysis of the uncertainties associated
with the procedure, new measurements in different conditions,
Monte Carlo simulations [24], a sensitivity analysis on the
parameters involved [25] are in progress.