Do wind farms produce less than expected?

Posted on Updated on

Almost two years ago I wrote a post called “Are there limits to large scale wind power?”  The post looked briefly at a paper by Adams and Keith which suggested that large scale wind farms will produce less electricity than expected.

From the paper’s abstract: ” New results from a mesoscale model suggest that wind power production is limited to about 1 W m−2 at wind farm scales larger than about 100 km2. ”

At the time I suggested a possible test of this would be the output of the London Array wind farm, which is 100 square kilometres in size. At the time it wasn’t open. But it has now been running for over a year.

What has happened? The wind farm was expected to have a capacity factor of 39%, if my memory serves me correctly. Capacity factors for UK wind farms are calculated by the Renewable Energy Foundation, using Ofgem data. While it is true that REF is itself an anti-wind farm organisation, their data and capacity factor calculations have never been shown to be wrong, and are thought to be reliable.

They tell us that the capacity factor of the London Array has been 40.2% since it opened, which is pretty much in line with expectations. It’s total capacity is 630 MW, therefore its average output is currently 253.6 MW. In other words it is delivering a power density of 2.53 W m−2 , much more than 1 W m−2 .

Now, this does not prove that the work of Adams and Keith is incorrect, or that the saturation effects they discuss will not prove serious issues; it only shows that it has not come into play at the London Array.

The London Array itself was supposed to see a further expansion, which would have provided additional evidence to test these saturation effects. However, the supposed threat of the turbines to migrating birds put an end to that. So, we may need to wait a while yet to see a wind farm much bigger than 100 square kilometres.

Power density is a topic I continue to return to.  Unfortunately many researchers, policy makers and environmentalists pay too little attention to it. Hopefully this will change next year with the publication of Vaclav Smil’s book on the subject, which is certain to be provocative and informative.


3 thoughts on “Do wind farms produce less than expected?

    Alan N said:
    November 22, 2014 at 2:44 pm

    Another test would be to compare output from the London Array with that from the smaller but nearby Thanet offshore wind farm. The difference is quite dramatic – a capacity factor of 40.2% for London Array, and 32.2% for Thanet over the same period. Monthly data on the Variable Pitch website shows a greater capacity factor at London Array for each of the 12 months. Of course there are other variables that may be at work – London Array is further out to sea, and has larger turbines (3.6MW rather than 3MW) – but these differences must be indicative that Adams and Keith’s limit to wind power production is at an area of greater than 100 square kilometres.

    Unfortunately, if the downscaling of large offshore wind projects continues as it has been doing recently, it may be some time before we get evidence from wind farms of larger than 100 square kilometres.


      jmdesp said:
      November 25, 2014 at 11:49 am

      Thanet is only one specific example. I personally did notice that the projects that get very high load factors, like above 50%, seem to be very small scale. Alpha Ventus for example has really excellent results, but is only 30 MW. Doing a systematic study would be useful.

      Also it would be interesting to study if the on-off behavior denounced by one study for Robin Rigg (either have the full power production, or be near zero) exists also elsewhere. This could explain that large scale plant aren’t very attractive, because going in a very short time from 2 or 300 MW to almost zero isn’t easy to integrate on the grid, maybe it’s better to have more small scale sites which aren’t as correlated.


    G Kochanski said:
    January 19, 2015 at 5:10 pm

    +jmdesp: Looking at different size wind farms runs into a sampling problem. There may be places that have really high wind power density but they are small places. (An example would be a wind farm on a high ridge; you can’t build a huge wind farm on a ridge.)

    So, you’d expect that small wind farms would have higher power densities because they’d be sited in the best spots. Large wind farms end up with (more or less) average spots.


Comments are closed.