This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


Wind vs. Nuclear Energy in the UK: A Question of Scale

Robert Wilson's picture
University of Strathclyde

Robert Wilson is a PhD Student in Mathematical Ecology at the University of Strathclyde.

  • Member since 2018
  • 375 items added with 246,395 views
  • Mar 21, 2013

Occasionally the Google search referrals I get for this blog give me ideas for a post, and today I got a rather topical one. The UK government has just given planning approval for the first new nuclear power plant in almost 20 years, and evidently someone wanted to know the following:

How many windfarms are needed to generate the same amount of power as Hinkley point nuclear plants?

According to the BBC this nuclear power plant will take up 170 hectares of land. For those who don’t understand hectares (and that includes me) this is 1.7 square kilometres, and from the air will look something like this:


So, how big would a wind farm need to be to provide as much electricity as Hinkley C? The plans are to have two 1.6 GW reactors on site. These will likely have capacity factors somewhere between 80 and 90%. So, the average power output from the plant should be at least 2.6 GW. Let’s use the London Array offshore wind farm as a starting basis to see how big a wind farm would need to be to match it. The London Array is a 630 MW wind farm, covering an area of 100 square kilometres just off the English Coast. If we assumed a capacity of 35% (I have not seen projections for its capacity factor, but it is not likely to differ too much from this) then its average output will be just over 0.2 GW. So, if we wanted to scale this up to provide as much power as Hinkley C then we would need a wind farm covering about 1,200 square kilometres, which is just a bit less than the area of Greater London.

This of course does not mean that nuclear power is better than offshore wind. There are a range of other considerations. But it is a nice demonstration that large scale renewables will require a lot of land, or sea.

Robert Wilson's picture
Thank Robert for the Post!
Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.
More posts from this member
Spell checking: Press the CTRL or COMMAND key then click on the underlined misspelled word.
Ivor O'Connor's picture
Ivor O'Connor on Mar 21, 2013
I don't know anything about offshore spacing. On shore though the spacing is about 4 acres to every 4MW turbine. 170 Hectares = 420 acres 420/4 = 105 4MW Turbines or 420MWs at a .35 capacity = 147MW 147/6200=2.37% the energy density Or the inverse 42.17. 42.17*1.7KM = 71.7KM 72KM does not equal 1200KM! Perhaps there is something strange about this offshore wind farm and the spacing they are using? I'm getting the impression you cherry picked numbers to make nuclear look bad. Elon Musk goes around saying if you installed PV on the lands used by nuclear power, not just the plant but the contaminated areas and the areas blocked off for security, there would be no need for the nuclear power plants. It would be interesting to run those numbers too...
Lund's picture
Lund on Mar 21, 2013

First: Offshore parks doesn't take up any land at all.

Second: Horns Rev II (209 MW offshore park, west of Denmark) capacity factor is slightly above 50%, with inefficient 2 MW turbines, that was the standard 10 years ago. From today on, it is fair to count on 60% of the capacity, based on the wind conditions in most of the North Sea, with modern 4-8 MW turbines.

The average capacity factor of nuclear power in the world is 80%, so we need 4,2 GW offshore wind, to match the electricity porduction from Hinkley Point C.

This is 6,7 times as much as London Array, or 670 square kilometres.

Dogger Bank is 17,600 square kilometres..... more than enough area to produce the same amount of electricity as consumed i the whole UK.

Best regards,

Søren Lund


Bill Woods's picture
Bill Woods on Mar 21, 2013

A not-guaranteed-to-be-typical sampling of wind farms suggests that your '4 MW per 4 acres' is out of line.


London Array (UK) : 630 e6 W * [0.35] / 100 e6 m^2 = 2.2 W/m^2
Greater Gabbard (UK) : 504 e6 W * 0.4 / 146 e6 m^2 = 1.4 W/m^2


Clyde (UK) : 350 e6 W * 0.3 / 47 e6 m^2 = 2.2 W/m^2
Shepherds Flat (US) : 845 e6 W * 0.27 / 78 e6 m^2 = 2.9 W/m^2
Bald Hills (Aus) : 104 e6 W * 0.36 / 17.5 e6 m^2 = 2.1 W/m^2

your hypothetical : 4 e6 W * 0.35 / 4*4047 m^2 = 86 W/m^2


Hinkley Point C : 3.2 e9 W * 0.8 / 1.7 e6 m^2 = 1500 W/m^2


Robert Wilson's picture
Robert Wilson on Mar 22, 2013

Bill Woods,

Please check the arithmetic in your comment. I use the London Array as a starting point. It is 100 square kilometres, and averages about 0.22 GW. So, to scale this up to an average of 2.6 GW you just need something 2.6/0.22 bigger, which is roughly 1,200 km^2.



I am not attempting to make nuclear look bad. In fact I can provide this link to David MacKay's blog. He assumed 2 W/m2 for wind farms, which is roughly what I am assuming. When he did this he was accused of cherry picking to make wind look bad, not nuclear. Fortunately, MacKay has actual data on his side about the energy density of existing UK wind farms, which he makes available.

Bill Woods's picture
Bill Woods on Mar 22, 2013

Yes, I had the same ratio:

London Array (UK) : = 2.2 W/m^2 = 2.2 MW/km^2 = 220 MW / 100 km^2

--> 2.6 GW / 1180 km^2

From context, I'd guess Ivor miswrote, and meant '... to make wind look bad.'


Of course, cost is a more important metric.

Cape Wind (US) :      2.5 e9 USD / (454 e6 W * 0.37) =   14.7 USD/W
London Array (UK) :   1.8 e9 GBP / (630 e6 W * [0.35]) = 12.4 USD/W
Greater Gabbard (UK): 1.5 e9 GBP / (504 e6 W * 0.4 ) =   11.3 USD/W
Clyde (UK) :          500 e6 GBP / (350 e6 W * 0.3 ) =    7.2 USD/W
Shepherds Flat (US) :   2 e9 USD / (845 e6 W * 0.27) =    8.8 USD/W
Bald Hills (Aus) :    250 e6 AUD / (104 e6 W * 0.36) =    6.9 USD/W

Olkiluoto (Fin) :    8 e9 EUR / (1.6 e9 W * 0.85) = 7.6 USD/W
Flamanville (Fr) : 8.5 e9 EUR / (1.6 e9 W * 0.8 ) = 8.6 USD/W
Hinkley Point C :   14 e9 GBP / (3.2 e9 W * 0.8 ) = 8.3 USD/W
Vogtle (US) :       15 e9 USD / (2.2 e9 W * 0.85) = 8.0 USD/W
Summer (US) :       10 e9 USD / (2.2 e9 W * 0.85) = 5.3 USD/W

For reasons Willem lists, that understates the relative value of nuclear vs. wind. On the other hand nuclear has a fuel cost.

Ivor O'Connor's picture
Ivor O'Connor on Mar 23, 2013

Robert Wilson,


I have not looked at your references or Bill Woods. However I have no idea where I got the 4MW/4A so I'm going to assume my information is invalid. I'm new to this and I have got to start making files so I can always double check and refer back to what I think is reality. I'll add your references. Thank you.

Robert Wilson's picture
Robert Wilson on Mar 23, 2013


Could you possibly provide some actual evidence that "it is fair to count on 60% of the capacity, based on the wind conditions in most of the North Sea, with modern 4-8 MW turbines." You have quoted the capacity factor of the *best* performing Danish offshore wind farm. This feels awfully like cherry picking.

Of course wind farms with turbines of 5 and 6 MW exist. How many of them are getting over 60% capacity factors? I am not aware of any.

I can also point to the most large scale evaluation of the UK's offshore resource. This assumed an average load factor of 40% for wind farms with fixed foundations, and 50% for those with floating turbines. Much lower than your rather optimistic projection of capacity factors of 60%.

So, my assumption of 35% is in line with existing evidence. And I also assumed a nuclear power plant capacity factor which is lower than is likely to happen, so my calculation is far from unreasonable.

Ivor O'Connor's picture
Ivor O'Connor on Mar 23, 2013

From two days ago:

"The “capacity factor” or the percentage of time that wind turbines are active has increased, he added. Capacity factors now hit 50 percent, thanks to better site selection and improved turbine efficiency. In the past, capacity factor was closer to 30%. Wind farms in optimal sites in the Midwest can produce power for 5 cents a kilowatt hour, and that’s before federal tax credits, he said."

Does hitting 50% mean it's just a rare case but not common? Or does it mean the new turbines are all getting about 50% if installed in a good site. And aren't all the best sites used first? It would be nice to know but as it stands now it appears he is trying to say we can now expect capacity factors of 50% instead of the old 30%.

Lund's picture
Lund on Mar 23, 2013

Dear Robert, who ask:


Could you possibly provide some actual evidence that "it is fair to count on 60% of the capacity, based on the wind conditions in most of the North Sea, with modern 4-8 MW turbines." You have quoted the capacity factor of the *best* performing Danish offshore wind farm. This feels awfully like cherry picking."

The only factor that gives Horns Rev II the best capacity factor of the danish offshore farms, is that this is the wind farm i Denmark installed on the location with the best wind conditions, so far. Only Horns Rev I & II are installed in the North Sea. All the other danish offshore wind farms are installed in the inner danish seas, where the mean wind speeds are lower.

We don't know the exact capacity factor of London Arrow yet, but the wind map shows that the mean wind speeds at London Array are similar to Horns Rev, if the hub hights are the same:


- and the first wind speed report for London Array [2] I found on Google, also says that the mean wind speed is 9,94 m/s in 100 m, which is within 1% off the mean wind speed of Horns Rev II.


As London Array's turbines is a newer, larger, more efficient design than the turbines on Horns Rev II (Siemens SWT 2.3 vs SWT 3.6), and most important, with higher hub hights, we should expect that London Array will get as least the same capacity factor as Horns Rev II, which means slightly more than 50%.

The same type of turbines (SWT 3.6) are now beeing installed on the new offshore farm Anholt Havvindmøllepark. Though wind speeds are substantially lower by Anholt, both Siemens and Dong Energy expect the capacity factor to be close to Horns Rev II.

Mean wind speeds 100 m above the sea surface by Horns Rev II, is 10 m/s. Using this as reference, you can calculate the capacity factor of the much more efficient Vestas V112 3.0 MW Offshore turbine [3], if installed on the same position.


In Vestas' data sheet [3], on page 15, you will find the AEP figures for this turbine, and see that the nominal annual power production by 10 m/s is 15.826 MWh per year, which means a capacity factor of 60,2% - which means 59,8%, including two service days a year.

Several Vestas V112 3.0 MW has already proved higher capacity factors than Horns Rev II....onshore!!!

But Vestas V112 3.0 MW is still only a part of the beginning, and a very small offshore turbine, compared with the turbines to be installed on offshore farms in the near future. Siemens has already installed af few of their newest SWT 6.0, and Vestas V164 8.0 MW is coming soon. Wether they are more efficint than V112 in comparable wind speeds, time will show, but the fact is that their rotors reach higher and faster wind streams than the V112, whereever they are installed, and just because of this, we should expect even higher capacity factors.

Take this into account, and imagin these turbines installed on locations like Dogger Bank, where the mean wind speeds are substantially higher than on Horns Rev II (see link [1]).

Thats why I assume it's fair to count on 60% capacity factors in the near future, and why you are far off, when you assume only 35% in your post.

Even Samsø Havmøllpark, a near shore wind farm with 2.3 MW turbines, installed in the inner danish sea 2002-2003, has a capacity factor of 43%, with wind speeds far lower than any off shore position in the North Sea.

Best Regards,

Søren Lund

- who despise cherry picking, and only wants to bring some facts into your otherwise very good debate.


NB; Anyone participating in this debate, should make themselves a quick self study of all the danish wind turbines, offshore as well as onshore, by downloading this spreadsheet from the danish Energy Authority "Energistyrelsen" 's website:

The spreadsheet shows commisioning and decomissioning dates, generator capacity, rotor size, position, and exact production data year by year, for every wind turbine in denmark.

You can f.x. easily compare the capacity factors of new onshore turbines with the older onshore turbines, and see that 35-50% is now a pretty normal capacity factor - on land!

You can also see the first production data from the first turbines on Anholt Havvindmøllepark, but yet not enough data to calculate the exact capacity factor.

(I can help with danish-english translation, if necessary)


Robert Wilson's picture
Robert Wilson on Mar 23, 2013


I can provide further justification for my assumption of a capacity factor of 35% for the London Array by linking to a non-technical summary of the wind farm from 2005. This says the eventual 1 GW wind farm will produce 3.1 GWh per year, which works out at a capacity factor of about 36%. 

Also this more recent brochure promoting the London Array says that the Load Factor is expected to be 39%. Slightly higher than what I assumed, but I also assumed the EDF reactors would have lower capacity factors than expected.

Now, it may be true that the London Array will have a capacity factor of 60%. However if this is true then they will have done an amazing job of ripping off the UK tax payer. The subsidies have been set up on the assumption that offshore wind capacity factors will be something like 35-40%, and not 60%.

So, overall I suspect that its output will not be 50% higher than expected.

Lund's picture
Lund on Mar 23, 2013

Dear Robert,

Whatever the reasons are, underestimating the capacity factor seems to be a trend.

Horns Rev II was estimated by Dong Energy to be 45%. The actual capacity factor, proved by the statistics from Energistyrelsen, is 50,9%. Samsø Havmøllepark was estimated to be 38%. The actual capacity factor is 43%. Etc, etc.

Wether it's a matter of simple conservatve estimation, or literally an intention to mislead the politicians, I can't say, but Horns Rev II was granted a garanteed price of 51,8 øre/kWh in 50.000 full load hours, because this was the price it needed to pay back the installation costs of 3,5 billion DKK in 50.000 full load hours.

So is the pricing practice, so the developer could actually have an interest in underestimating the capacity factor, overestimating the construction costs, and so on.

Since the bid on Horns Rev II, UK started to develop offshore farms very aggressively, which caused that there was only one developer, Dong Energy, with available construction capacity to bid on Anholt Havmøllepark in 2010.

The installation costs was then raised to 10 billion DKK (tripple the costs of Horns Rev II though the capacity is only double), and "won" the bid on a price of 105 øre/kWh garanteed in 50.000 full load hours.

This started a huge debate in Denmark, but Anders Eldrup from Dong Energy defended the extreme price hike saying; "We can get even higher prices in the UK!" ;-))

The Horns Rev II was according to Dong Energy build exactly on the budgeted 3,5 billion DKK, and I believe that Anholt Havvindmøllepark will cost Dong Energy exactly the budgeted 10 billion DKK too, because the whole margin to the tru costs will be paid to A2SEA, a very profitable company....owned by Dong Energy. ;-)

So much for the developers estimates, and moreover, it dosen't make much sense to base your numbers on estimates from 2005, and even put them lower than this, when there are so many facts from actual working wind farms and so much development have been done on new wind turbines in the latest few years.

The new turbines are tested according to very specific standards, and it only takes a few months of testing to know how the turbines will actually perform in specific wind conditions. These data together with the known wind speed data on each location are essential for the developers to make decisions.

Thats why you can trust these data and make much better estimates, and thats why I wrote on already in 2009 that Horns Rev II would be 50% rather than 45%.

I don't say London Array will reach 60%, but as the wind speed is the same as on Horns Reef, and the turbines are newer, larger and more efficient, it should be able to produce as minimum the same capacity factor as Horns Rev II.

The 60% is my assumption of all the coming wind farms, with much larger and even more efficient turbines on even windier locations in the North Sea.

Best Regards,

Søren Lund


Robert Wilson's picture
Robert Wilson on Mar 23, 2013


Again, this feels overly anecdotal. And I must remind you that the plural of anecdote is not data.

Also, the life capacity factor of Horns Rev 2 appears to be 48.4%, not 50.9% as you state. A single wind farm having a capacity factor 7% higher than expected is not something than be generalized from. Such differences are very possibly accountable simply from above average wind conditions.

Even at 50% we are still looking at a capacity factor of 20% higher than what DONG is claiming in their own marketing literature about the London Array. And one wind farm out performing DONG's claim capacity factor by 7% over a three year period is not overly convincing evidence that the London Array will outperform it by over 20%.

Lund's picture
Lund on Mar 23, 2013

Dear Robert, who says:

""Again, this feels overly anecdotal. And I must remind you that the plural of anecdote is not data.""

Neither the capacity factor of London Array, Anholt Havvindmøllepark or other wind farms connected in the future, has any factual production data or capacity factor yet, so neither yours or mine estimates on these wind farms are data - but that doesn't mean that your estimate of 35% is quite well supported by the facts as mine.

The data from the wind turbine manufactures data sheets, the 10 years mean wind speeds, collected by satelites, as well as all the production data I gave you from, are all real data, which supports my estimate - not yours.

Robert says: ""Also, the life capacity factor of Horns Rev 2 appears to be 48.4%, not 50.9% as you state.""

Thats wrong!

You can find the commisioning date and the production data of the 91 wind turbines on the lines 1822 to 1912 in the spreadsheet:

In the years 2009 to 2012, the 209,3 MW wind farm produced respectively [kWh]:




- which immideiately gives the capacity factors:



But then you can read in the spread sheet that the turbines are connected through the year 2009, and that 14 of the turbines was not connected until December 2009. These turbines was still in their run-in period the first two months of 2010, which means that the farm was not fully operational until March 2010.

In other words, you can not use the production data of 2009 to state any capacity factor, and you can only use the data from 2010 with some subject, while the years 2011 and 2012 are fully representative for the wind farms capacity factor.

The average of 2011 and 2012 is 50,9%, or 49,5% if you include the year 2010.

Robert says:

""A single wind farm having a capacity factor 7% higher than expected is not something than be generalized from.""

In your own post above, you generalise ""how many wind farms are needed..." based on very early estimates of one single wind farm, which doesn't even have any actual data of the capacity factor.

My only intention is to provide facts to your otherwise very good debate. The readers can then make their own estimates.

Best Regards,

Søren Lund



Nathan Wilson's picture
Nathan Wilson on Mar 23, 2013

The Texas and China CF values seem low.  Do they include curtailment due to delayed grid expansion that these locations have experienced? (as I recall, at the end of 2012, something like one quarter of China's wind farm weren't connected to the grid).

I K's picture
I K on Mar 23, 2013

Hope this helps:

The USA fleet of just over 100 reactors achieves an average 92% capacity factor. That means half the reactors do better than that. One would hope a brand new reactor built today would be better than the average reactor in the USA built 40 years ago. So taking 92% is probably a good idea and the real figure for a new reactor should be higher. (Potentially as high as 98% Capacity Factor is possible)

The two EPRs at Hinckley would thus achieve: 1.6 x 0.92 x 24 x 365 = 25,790 GWh of output per year.

Comparing a wind farm is more tricky. The builders tend to assume 38-40% for offshore and some offshore sites have now been in operation and roughly confirm that (some years they produce less some years a bit more) but lets go with 40%.

You would therefore need an offshore wind farm with a nameplate of 7.36GW running at 40% CF to match the energy output of the new proposed dual reactor station at Hinckley.

But importantly the nuclear reactors output steady predictable power which means if you build them you can turn off 3.6GW of old coal or gas stations. If you go with the wind farm you need to keep that old coal and gas station there is no two ways about it. Other than cost, those stations even when not producing power for the grid will consume energy themselves. Not a lot but still in the 10s of MW of power

This sounds fairly pro nuclear, but IMO nuclear is only workable and a success if you plan to build at least 20 reactors. That way the first two or three will cost you an arm and a leg, the next 2-3 will be too expensive to compete with gas, the two-three after maybe can compete with gas and whatever you build on top of that will likely be cheaper than even coal. So build a few and its going to be a disaster (a la Finland) but build a lot and it will be a success (eg USA and France).

Lund's picture
Lund on Mar 23, 2013

BTW: your link to "" is right about the data from 2012, and describes correctly how the capacity factor is calculated. And it links to the same statistics I gave you before.

But they doesn't say how they figure out the "life capacity factor", and I dont know such figures from Energinet or Energistyrelsen.

According to Energinet's and Energistyrelsen's data, the production data of both 20 and 25 years old wind turbines in Denmark doesn't show any decline in capacity factor, related to their age.

How ever "energynumbers,info" got it, it can't be anything else but estimates, but at least they could have taken into account that the wind resources in 2012 was only 96% af a normal indexed wind year in the region, which means that if the average wind resourses of the coming 22 years are equal to the previous years since 1979, the capacity factor of Horns Rev II will be clearly higher through the lifetime than in the previous 3 years.


Robert Wilson's picture
Robert Wilson on Mar 24, 2013


Can you please think before you comment? This is the site I linked to when I quoted the capacity factors for Horns Rev 2, you are simply giving me information my comment makes clear I am already aware of.

I K's picture
I K on Mar 26, 2013

You can install 10 x Hinckley nuclear power stations and get 32GW nuclear at 95% capacity factor = 266TWh. This 32GW of nuclear would integrate into the grid. We would also need 23 GW of CCGT operating at just 32% capacity factor. All the connections exist and the 23GW of CCGT exist. It would just be a matter of replacing some of our old coal plants with new reactors.

The wind alternative is to install 76GW of offshore wind @40% capacity factor and have 55GW of CCGT backup. That means we would need to build about 30GW of new CCGTs. We would need to connect about 150 wind farms with 21,000 turbines to a AC-DC station and then run 150 DC cables ashore and convert back into AC and connect to new or existing transformers. None of the above in itself is a deal breaker its all possible.

The problem comes when your 76GW of wind is outputting 80% of capacity or 61GW when demand is say 31GW. The difference of 30GW is a near impossible figure to utilise. At the very best we could perhaps export (why would we want that) 3GW to France and the Netherlands but that still leaves us with 28GW of unusable wind power. The only choice would be to turn 30GW of the wind farms off until demand picked up or wind slowed.

 As a result we would not get the assumed 266TWh from wind, as we would have to stop the turbines fairly frequently.

 This would not be a big problem if wind got cheap enough so it was producing under ~£40/MWh unsubsidised but it is a problem if it is producing at ~£100/MWh subsidised.

Lund's picture
Lund on Apr 1, 2013

Good work, David!


Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »