Environmental Engineering Reference
In-Depth Information
determining whether or not to approve the proposed wind farm based on the
estimated impact to particular species.
Collision modelling uses data for particular bird species to ascertain a level of
collision risk. Data collected from the wind farm site includes measures of flight
frequency and flight heights and the number of individual birds on-site, relative to
the number, layout and dimensions of proposed turbines.
To date, it has not been feasible to obtain requisite utilisation data for species of
microbats due to limitations in capacity to discriminate numbers of individuals in
flight and to adequately detect and/or distinguish taxa of bats at relevant heights.
Thus far, the application of technologies including acoustic bat detectors, radar and
thermal imaging has not fully resolved these limitations and modelling has thus
been applicable only to birds.
The model's results are provided for a range of theoretical avoidance rates
because we have little empirical evidence for the capacities of different birds to
avoid collisions with turbines. The first empirical avoidance rates have just been
reported for two eagle species at two Tasmanian wind farms by Hull et al. (
2013a
).
How Does Risk Modelling Compare with Actual Experience?
As outlined above, there is little empirical data that can be used to compare the
modelled projections with actual collision rates. Our capacity to validate the mod-
el'sprojectionsisthuslimited.However,fortheBluffPointandStudlandBaywind
farms in Tasmania, where substantial, rigorous and controlled programs of monitor-
inghavebeenunderwayfor9and5years,respectively,dataareavailableforwhite-
bellied sea-eagles
Haliaeetus leucogaster
and wedge-tailed eagles
Aquila audax
fleayi
(Hydro Tasmania
2012
). A fuller comparison of the model's results with
actual collision rates for these two species at the two wind farms is provided in
Smales et al. (
2013
). However, Table
2
shows the model's results at three avoidance
rates for these species along with the mean annual number of actual collisions
detected over the entire periods of operation of the two wind farms. The model's
estimatesat95%avoidanceratecloselyapproximatethedocumentednumbersof
actual collisions.
Table 2
Annual numbers of eagle mortalities estimated by modelling compared with numbers of
actual mortalities detected for two species at two Tasmanian wind farms
White-bellied sea-eagle Wedge-tailed eagle
BluffPoint StudlandBay BluffPoint StudlandBay
Modelledavoidancerate
90%
0.9
0.8
2.7
1.9
95%
0.5
0.4
1.5
1.1
99%
0.1
0.1
0.4
0.3
Actual mortalities detected
0.4
0.0
1.6
1.1
BluffPoint2002-2011
StudlandBay2007-2011
