How many birds do wind farms kill?
Wind farms kill far less than other hazards, but there are still ways that we can reduce them.
Bird species are under threat from climate change.
It would be worrying, then, if a move to low-carbon energy increased pressures on bird populations. That’s a common concern as countries move to wind power.
It’s true: wind turbines do kill birds (and bats). But how many, and are they a bigger threat than other hazards?
In this post, I take a look at estimates of bird deaths from turbines and try to put them in context. I also explore ways that we can reduce them.
How many birds do wind turbines kill?
Measuring bird kills from turbines is hard. An obvious way to do so is to have humans go out and count bird carcasses in the area. Many studies have done that.
The problem is that humans often miss small birds, such as songbirds. That’s where dog searches come in.
The estimates that I found in the literature vary quite a bit. Partly due to measurement challenges, but also because risks vary by location: some areas will be prime hotspots for wildlife while others will be more barren.
Estimates ranged from 4 to 18 birds killed per turbine per year. More than four times the difference. I’ve detailed some of these studies in the table below.1
The large spread of these estimates isn’t very satisfying, but at least gives us some sense of magnitude. What would this mean for the total number of birds killed each year?
Let’s apply these numbers to the United States (which is where most of the studies came from).
In 2022, the US produced 434 TWh of wind power.2 Taking the numbers above, that gives us a range of 200,000 to 1.2 million. The upper figure seems more likely since it tries to correct for the under-detection of smaller birds. Let’s call it around one million birds per year.
Assume that these risks are the same across the world, and global deaths are probably over 5 million.3
Cats, buildings, and cars kill far more birds than wind power
Around one million birds are killed in the US. Is that a big number?
Not really, compared to other pressures.
The chart below shows estimates of the number of birds killed by different hazards in the US.
You can see that wind turbines kill a few million at most. Cars, buildings, and pesticides kill tens to hundreds of millions each. Cats kill at least a billion.
Do these figures seem credible? I did a bit of a sense check on a few of the numbers below. If you want to follow along, feel free. If not, skip to the next section.
Cats: Estimates range from 365 million to 2.4 billion bird deaths in the US per year:
- Loss et al. (2013) estimate 2.4 billion.
- Subramanian (2012) estimate 365 million to 1 billion. Based on data from the US Fish and Wildlife Service.
Do these magnitudes seem reasonable? In their 2013 study in Nature, Loss and colleagues provide an overview of predation rate studies from across the US and Europe; these estimate the number of birds killed per cat per year. Owned cats range from around 4 to 30+ birds killed per year. One every few weeks or months.
Non-owned cats kill more: typically in the range of 50 to 150. That would mean one bird every 3-7 days.
From that study, there are around 100 million outdoor cats in the US – that’s non-owned (feral) cats plus owned cats that go outdoors.
If 2.4 billion birds are killed by 100 million cats, that’s around 24 birds per cat. That seems like a reasonable estimate given the predation rates above. It might even be a bit low considering rates for feral cats are much higher.Cars: Estimated that 60 million to 200 million are killed by automobiles every year.
- Loss et al. (2015) estimate 200 million.
- Subramanian (2012) estimate 60 million. Based on data from the US Fish and Wildlife Service.
There are around 283 million registered vehicles in the US. That would mean less than one bird per car per year, ranging from 20% to 70% of cars hitting a bird each year. I expect that this will be skewed, with some vehicles – particularly large trucks – hitting more than one per year, and many cars hitting zero.How do wind power impacts compare to fossil fuels?
I’ve seen a bunch of charts floating around that compare different energy sources. Like the one below. And statements such as “fossil fuels kill 17 times as many birds as wind energy per unit of energy.”
They tend to link back to a 2009 or 2012 study by Benjamin Sovacool. I dug into it a bit, and while I think it’s very likely that climate change will have a much larger impact on bird populations than wind turbines, I’m a bit skeptical about putting these definitive numbers on it.
Bird deaths from fossil fuels are attributed to a few factors: acid rain caused by sulphur dioxide, mining impacts, and climate change. These will all harm wildlife, but the cited numbers are uncertain for a few reasons.
First, in many countries – particularly across Europe and North America – sulphur emissions are usually ‘scrubbed’ out of coal power stacks. Acid rain is much less of a problem than it used to be. So while this might be valid in some countries in the world, it’s probably not a fair assumption for a comparison in the United States.
Second, ‘fossil fuels’ are all lumped together. Coal is much ‘dirtier’ in terms of local air pollution and carbon, so you’d expect it to have a much higher impact than oil or gas.
Third, it’s very difficult to quantify the impacts of climate change on bird deaths or species extinctions. Of course, we know that there are many species threatened by climate change, but putting a definitive number on how many individuals might be lost is hard. To be fair, Sovacool acknowledges this:
“The role of climate change on bird extinctions, although indeed worrying, is not conclusive, and as such should be approached with extreme caution.”
His calculations – although not documented transparently – seem simplistic. It looks like he uses a 2004 study that projects that 15% to 37% of all species of birds could be extinct by 2050 in a middle-of-the-road scenario by 2050. Then extrapolates this to the number of individual birds, and assumes they are lost in a linear rate out to 2050. That might not be the case: the at-risk species might have smaller populations than other species, and these extinctions are likely to scale non-linearly with increasing temperatures so most are lost closer to 2050.
Most of the deaths from fossil fuels are attributed to climate change – 9.36 GWh compared to 0.2 GWh for other factors – so these assumptions matter a lot.
Now, I think it’s reasonable to assume that climate change – driven by fossil fuels – will contribute to significant losses of bird species and populations. And I expect the death rates from fossil fuels are higher than for wind power. But I’d be hesitant to put a definitive number on it without more solid research.
The numbers for nuclear power also seem a bit dodgy to me. Paul Lorenzini did a rebuttal of those numbers – so I won’t rehash that here – but I agree with his overall conclusion that they’re probably too high.
Wind power is a threat to particular types of birds, particularly birds of prey
It’s not just the total amount of birds that are killed that matters, but what types. If a particular species of bird is disproportionately affected it could have real impacts on population dynamics and risk of extinction.
A study by Chris Thaxter and colleagues (2017) looked at the collision rates of different bird species from a large literature review. The documented rates of collisions for different groups are shown in the chart below.
If you’re like me, you will have no idea what most of these groups actually mean. So I’ve translated a few of the most impacted orders:
Accipitriformes: birds of prey including eagles, vultures, hawks, and kites.
Bucerotiformes: birds such as hornbills and hoopoes.
Ciconiiformes: storklike birds such as herons, shoebills, and hammerheads.
Charadriiformes: shorebirds such as waders, gulls, and auks.
In short, birds of prey such as eagles, raptors, and hawks; shorebirds; and storklike orders are at much higher risk of collisions than other families, such as songbirds. This disproportionate risk has been found across many other studies.
These species can be at a higher risk for several reasons. First, they will often use ridgetops to get lift from the wind. Incidentally, this is also a good spot for wind turbines. Second, they are often migratory birds; if wind farms are in their migratory route this puts them at higher risk. More indirect impacts of wind farms – which might not be reflected in death statistics – is their effect on the disruption of migratory patterns.
While the total number of birds killed by turbines is low compared to other hazards, the threat to particular species is more concerning. We need better mapping of key hotspots for these species so that wind farms can be deployed in suitable locations. More on how we can reduce these deaths later.
Wind power is probably a bigger threat to bats
I’ve mostly focused on bird fatalities, but wind power also kills bats. I found it harder to get good numbers here, but estimates suggest it’s in the range of 6 to 20 bats per turbine per year. Some estimates are even higher.
The Thaxter et al. (2017) paper that we just looked at also measured collision rates among bats. If you look at the scale of the ‘collisions/turbines/year’ you’ll see that it’s an order of magnitude higher.
Again, this tallies with other research that suggests that bats have higher mortality rates than birds for wind farm collisions.
We can reduce bird and bat deaths from wind power
We’re not completely helpless in this dilemma. There’s a lot that we can do to limit the biodiversity impact of wind farms, even if fatality rates are not reduced to zero.4
Here’s what we can do:
1. Turn off wind turbines at very low speeds when bats are around
Bats tend to get hit by wind turbines when wind speeds are very low. They struggle to fly in windier conditions. That means we can prevent a lot of bat deaths by curtailing – switching off – our turbines when there isn’t much wind.
You might think that this would hinder energy supply and eat into owner profits. But studies suggest it doesn’t make much difference.
A study from Pennsylvania reduced bat deaths by 44% when wind turbines were turned on at 5, rather than 3.5 metres per second.5 And they fell by 93% when this was increased to 6.5 metres per second.
A study in Australia found that raising the wind speed threshold from 3 to 4.5 metres per second reduced deaths by 54%, and the wind farm only lost 0.1% of revenue.6
Another study in Cadiz, in Spain, found that bird deaths were halved with only a 0.07% loss in energy production. That’s because the biggest risk was migratory birds – which pass through very occasionally. Shutting down production during this time was quick and saved many lives.
2. Don’t put wind farms in high-risk areas for birds and bats
Areas like ridgetops are prime spots for migratory birds and raptors that use the winds for uplift.
As we saw earlier, these species tend to be disproportionately affected by wind farms, so we should strive to avoid these areas.
3. Fewer larger turbines are better than many small ones
Birds and bats might be more likely to collide with a large turbine than a small one. But the question is whether a wind farm should have a few large turbines or lots of small ones.
The study by Thaxter et al. (2017) suggests the former. Fatality rates for both birds and bats tend to be higher in wind farms with turbines of very low capacity.
Having a small number of large turbines would therefore reduce fatality rates.
4. Paint the turbines black
When birds get close to turbines, the blades spin so quickly that it blurs their vision. But, if you paint the turbines black, it makes them much more visible.
Some tests of this approach in Norway reduced bird deaths by more than 70%.7
It might not be as effective for offshore farms, so that still needs to be tested.
5. Play alert noises to bats and birds to deter them
For some bat species, playing high-pitch sounds (which humans can’t hear) can deter them from the area. A study in Texas reduced the deaths of the deaths of two species of bats by 54% and 78%.8 It was then rolled out to many wind farms in the area.
Other systems can be used to identify eagles in the nearby area, and either emit distracting noises or switch the turbines off automatically.
6. Use GPS to track and find the optimal height for turbines
Surveillance technologies, such as GPS, can help scientists understand the flying patterns of migratory species. That means we can pick more optimal heights for turbines when they’re being constructed.
It can also alert wind farm generators that migratory flocks are in the area, so they can switch turbines off during high-risk times.
While some wildlife deaths from wind power might be unavoidable, there’s a lot that we can do to reduce them. It might come at very little cost to energy output and profit, so at a time when the world’s birds are under threat, it’s worth doing.
Here are the sources:
Subramanian (2012). Nature. The authors give an upper and lower estimate: in the table, I give the mid-point.
Electricity production figures come from Ember Climate.
The world produced around 2000 TWh of wind power in 2022.
1.2 million / 434 * 2000 = 5.5 million.
I found this article really useful when studying this.
https://www.canarymedia.com/articles/wind/wind-turbines-kill-too-many-birds-and-bats-how-can-we-make-them-safer
Arnett, E. B., Huso, M. M., Schirmacher, M. R., & Hayes, J. P. (2011). Altering turbine speed reduces bat mortality at wind‐energy facilities. Frontiers in Ecology and the Environment, 9(4), 209-214.
Bennett, E. M., Florent, S. N., Venosta, M., Gibson, M., Jackson, A., & Stark, E. (2022). Curtailment as a successful method for reducing bat mortality at a southern Australian wind farm. Austral Ecology, 47(6), 1329-1339.
May, R., Nygård, T., Falkdalen, U., Åström, J., Hamre, Ø., & Stokke, B. G. (2020). Paint it black: Efficacy of increased wind turbine rotor blade visibility to reduce avian fatalities. Ecology and evolution, 10(16), 8927-8935.
Weaver, S. P., Hein, C. D., Simpson, T. R., Evans, J. W., & Castro-Arellano, I. (2020). Ultrasonic acoustic deterrents significantly reduce bat fatalities at wind turbines. Global Ecology and Conservation, 24, e01099.
One thing not factored into the cat numbers is how many birds would be killed in a non-suburban, non-house cat habitat anyway. That is, birds are preyed upon by bobcats, other birds, snakes and such. House cats are replacing these apex predators in suburban environments.
Interestingly, I worked for (landscaping) almost 40 years in a neighborhood in Irvine,CA. Many people had outdoor cats and there did not seem to be a noticable deficit of birds (though that's a guess). In the early 90s coyotes moved in and starting on preying cats so everybody brought their pets indoors. The rabbit and rat populations subsequently exploded! I would see hundreds of rabbits mowing down people's lawns in the morning drive to the site. Residents could not grow veggies or fruit because rats would eat them ( A pair of bobcats did take up residence for awhile and controlled the problem but they were killed by rat poison put out by folks). Interestingly the coyotes could never control the varmint problem--in direct opposition to local environmentalist theory.
Anyway, you would think with this dearth of outdoor cats the bird populations would flourish...not so! It is just a relatively barren avian ecosystem with the usual suburban denizen-- lots of crows,a few pigeons, mockingbirds, and a few finches. Why is this? It's probably because the crows and rats are eating nestlings and eggs of more diverse species!
Great article 👏 Thanks.