Helping New Zealand’s birds find the safest path through wind turbines
Advanced Analytics
Using advanced data analysis techniques to develop an avifauna collision risk model to assist Boffa Miskell with the assessment of ecological impacts of wind farms on birds, and, specifically, to enable robust testing of layout and design scenarios that will minimise collision risk to birds.
Boffa Miskell is a leading New Zealand environmental planning and design consultancy company. Its ecology team has worked on over 29 wind farms of which 10 have been built, one is under construction, and 13 are in various stages of consenting. At each site, Boffa Miskell’s ecologists collect tens of thousands of observations of avifauna at each site which require detailed analysis to determine the collision risk profile for each species.
The Challenge
Boffa Miskell engaged Harmonic Analytics to replace a simple collision risk model that had been in use since 2004, with a new model that responds to the growing knowledge of the behaviour of native bird species in New Zealand, and is updated to account for changes in international guidance, and is more statistically robust.
Harmonic Analytics were challenged to build a forecasting model that could assess multiple species at the same time and estimate the number of birds that may collide with wind turbines within a proposed wind farm layout. The resulting predictions can influence the final design of a wind farm to reduce effects, and help to quantify mitigation requirements, if needed.
The model needed to account for the various scenarios using different wind farm layouts, turbines, numbers, and different turbine options. It also needed to allow for analysis of the distribution and abundance of each species of bird; whether they resident within, seasonal visitor to, or migrants traversing each site. It also needed to account for the flight behaviours and interactions with the wind farm, which vary from species to species.
The Solution
Harmonic Analytics worked collaboratively with a senior ecologist at Boffa Miskell, to build from the ground up, a replacement collision risk model in R that estimates the risk of bird flight collision for any proposed wind farm layout.
This R model allows rapid testing of multiple scenarios. These scenarios can account for changes to any or all variables, such as wind farm size, turbine design, wind speed, species abundance, and species behaviour. It applies Monte Carlo simulations to all key variables to produce confidence intervals for each prediction.
The Result
● We worked collaboratively with Boffa Miskell to create a collision risk model that will assist them in minimising and mitigating any avifauna mortality.
● The model is statistically robust and will assist in the resource consenting and hearings process for new wind farm projects.
● The model draws on the most recent guidance from the United Kingdom, updating and expanding on earlier models used in New Zealand. since 2010.
● We have discussed with Boffa Miskell several additional options to expand the model’s capabilities to run different scenarios and automate the pre-processing of raw data.