Sunday, 3 February 2013

Are wind turbines merely icons of ecological faith?

Simon Hoggart is probably not alone in believing that wind turbines 'are nearly hopeless as a means of generating energy' and their main purpose is to 'demonstrate ecological faith' [1]. Yet just this week the Danish Wind Association reported that wind energy is now generating 30% of their electricity[2]. If they can do it so could we, if we got our act together. The problem is that arguments against wind are repeated by anti-wind lobbyists, whereas arguments for wind are repeated by pro-wind campaigners, with very little criticism on either side. Recently a study was reported by the Renewable Energy Foundation (renowned for anti-wind lobbying [3]) claiming that the usual lifetime of wind turbines is much less than the 20-25 years normally factored into cost and profit calculations [4]. This was reported uncritically by the Telegraph [5] and directly refuted by industry practitioners and DECC [6]. I decided to take a look.

It is often said the wind turbines are useless because they have a low capacity factor: usually only 25-30%. What this means is the on average they only generate a quarter to a third of their maximum power. Obviously they only generate maximum power in strong winds, so, yes wind turbines have a low capacity factor. However this is factored into their costs.

Wind turbine power curve from Vesta. The x axis is wind speed in m/s. In this case maximum power is at 12m/s (27 mph) but this turbine will start generating as low as 3 m/s (8mph). Some turbines will be better at lower wind speeds, some at higher wind speeds - you need to choose one suitable for your location.

Other power stations do not run at full blast all the time either, partly due to maintenance time but also because of varying demand and other factors. The equivalent for capacity factor for conventional power stations is called load factor; in 2011, the average load factor for nuclear power stations was 66% (up 7% on 2010) and for gas combined cycle power stations it was 48%. This is a record low and it was due to over capacity and high gas price compared to coal [7]. It is very hard to plan an efficient power grid when fuel prices vary from month to month whereas it takes years to install new power stations. At least we can know the price of wind in advance.

Average capacity factor for UK onshore wind turbines since 2000 [8] - this has varied a little due mainly to wind levels (black line) - 2010 was a very poor year.

The cost of wind power is critically dependent on lifetime because it is all up front: the fuel is free and maintenance costs are also low. In December, the Renewable Energy Foundation published a study by Prof. Gordon Hughes (an economist and long time anti-wind campaigner) which claims that wind turbine lifetimes are much lower than is normally assumed - only 10-15 years rather than 20-25 [4]. The reason given is that the capacity factor declines rapidly with age due to mechanical problems; for onshore wind farms the survey reports the initial capacity factor (normalised for wind availability) as 25% dropping to 15% for 10 year old plant.

The appendix of this study shows an interesting chart plotting capacity factor against wind turbine age. This shows no trend whatsoever.
Chart from [2] showing no trend whatsoever between turbine age and capacity factor (The solid green boxes mark the range covered by 50% of the sample points, the long thin bars cover 80% of the range and outlying points are marked as small blobs).

However, Hughes looked deeper, working to separate the effect of age from that of differences in wind speed by location and over time. He did this without using wind speed data at all - simply by using statistical techniques. I quote: Wind availability is treated as an unobserved factor whose contribution to monthly variations in wind output is estimated by a series of fixed period (month/year) effects in a statistical model of monthly output for all wind farms. Given that the observed trend in capacity factor is zilch, a significant age related deterioration must be matched by a considerable increase in wind availability which is completely absent from the statistics reported at the regional level. I dare say some people would trust a statistical derivation over a direct measurement, but it seems bonkers to me.

Not satisfied with merely claiming that wind turbines deteriorate to the point of uselessness in a fraction of their expected lifetime, Hughes goes on to tell us that new wind farms have lower initial capacity factors now than they used to because the best wind farm sites were taken up early: in 2000 new UK onshore wind farms started with capacity factor 26% but by 2011 this was down to 16%. If this were true, then with the deteriorating performance of older wind turbines combined with the poor capacity factor for new wind farms there would have to be a clear effect in the measured trend - and yet as we have seen there is no overall change either in capacity factors or windiness.

Now even if some of the effect Hughes reports is genuine, that still does not prove that wind turbines don't last longer than 15 years. The study goes back 18 years and over that time technology has improved considerably. I would expect older wind turbines to wear out more quickly than new ones because the initial gearing technology was very poor (and noisy). Studies in Denmark have shown that new generations of wind turbine technology do indeed last longer with less maintenance than older ones [9]. Also if new wind farms have lower capacity factors than older ones that could be because better technology allows the turbines to make better use of higher wind speeds, hence the average power generated will be a smaller fraction of the maximum.

I should add that I am a great fan of wind turbines so I hope you have been exercising your critical faculties while reading this post.

[1] Simon Hoggart, in the Guardian 2/Feb/2013 As I have said before, these structures are essentially religious in purpose, designed to demonstrate ecological faith, because as a means of generating energy they are nearly hopeless.'
[2] Wind Energy Hits Over 30% of Denmark's Electricity Consumption At End of 2012 Cleantechnica 2013
[3] Will the Real Renewable Energy Foundation please stand up Guardian 2011
[4] Wear and Tear Hits Wind Farm Output and Economic Lifetime Renewable Energy Foundation 2012
[5] Wind farm turbines wear sooner than expected says study Telegraph 2012
[6] Study on turbine lifespan just more anti-wind propoganda EWEA
[7] Digest UK Energy Statistics 2012
[8] Regional Renewable Statisitics: Historic regional Statistics DECC
[9] Operational and maintenance costs for wind turbines Wind Measurement International


  1. What is the definition of "last"? For example, the lifetime of a nuclear reactor is limited by the wear on the reactor vessel. It is usually uneconomic to replace because the entire structure must be torn down around it.

    Wind turbines require regular maintenance but the entire turbine can be rebuilt with a crane. If you change the generator, which may account for a significant part of the cost of an individual turbine, is that still the same turbine? I doubt your statement that the entire cost is upfront, it is still moving electrical machinery and there's plenty of people who make a living maintaining wind turbines.

    A real study would analyze the amortized cost per kWh including maintenance and overhaul. It is also improper to look at a wind farm, you must analyze a wind turbine individually as when turbines fail they are naturally replaced.

    1. According to Wind Measurement International, the source I cited above, the components of a wind turbine are designed to last 20 years and at the end of that time it can be cheaper to refurbish it with some new parts than to start again. The cost of maintenance of older turbines was around 3% of original cost per year - for newer turbines this has come down. Recently Bloomberg reported a continuing trend of lower O&M costs for wind turbines,