Monday, August 3, 2015

Offshore wind farm load factors

I heard that recently-built UK offshore wind farms have higher load factors than old wind farms. I thought it would be interesting to plot the raw data, which is available from the helpful REF website. (The load factor is the ratio of the average output of the wind farm to the maximum possible output, and it is usually expressed as a percentage.) The graph below shows the rolling load factor of the 27 offshore wind farms in the UK, versus their accreditation dates. The size of each point is proportional to the capacity of the wind farm. One of the newest wind farms has a very low rolling load factor - this is Gwynt y Mor, the bottom right point on the graph; Gwynt y Mor is the second biggest offshore wind farm in the world. Presumably this poor load factor is a temporary glitch associated with the commissioning of the wind farm. If we ignore that point as an outlier, and simply compare the average rolling load factors for accreditation dates before and after June 30th 2010, it looks as if the average load factor may have increased from about 33.9% (± 2%) to about 37.4% (± 1%).
Of course, a proper analysis should account for ageing effects and variations in the weather, both of which have been carefully studied by Staffell and Green. They found that "Wind turbines ... lose 1.6 ± 0.2% of their output per year" (which implies, for a load factor of 35%, a reduction in load factor of 0.56% per year - a trend that I have shown by the green straight line in the next graph).

So, has there been a technological improvement in offshore wind turbines that has boosted load factors? Having been involved in innovation support while I worked at DECC, my prior expectation was that the answer to this question could easily be "yes". But actually it looks like it is possible to account for the apparent trend in the data by a simple "ageing" hypothesis: perhaps the newer wind farms are better just because they have aged less?

7 comments:

Anonymous said...

Is the aging a reduction in individual turbines' output or simply an increase in the annual amount of maintenance downtime required averaged across a farm?

Bishop Hill said...

Presumably one could equally conclude that the turbines are ageing faster than Staffell and Green suspected but that there has been some technological advance. I would have thought this was a more plausible explanation than no tech advance at all.

David MacKay FRS said...

@Anonymous: The Staffell-Green paper didn't resolve whether the ageing is caused by changes in the turbines or because of maintenance downtime. This is the sort of information that wind farm owners will have at their fingertips, I am sure, but (even though the public have actually paid everything) such data are never published, sadly!

@BishopHill - yes, what you said is conceivably true; and it is possible that some but not all of the new wind farms are indeed much better, and that we would be able to see an improvement signal if we could disaggregate the data by wind-turbine technology. Sorry, I just did a quick and dirty one-hour job on this, to see if there was any stand-out signal in the data.

pawb1 said...

The performance of wind farms is very dependent on their location. This applies to both onshore and offshore wind farms. For example the rolling average load factor for onshore wind farms in England is 25%, in Wales it's 25%, in Scotland it's 28% and in NI it's 31%. Within England there are big regional differences (22% to 26%). Has the location (and hence wind resource) been factored into the offshore data analysis?

James said...

Another hypothesis would be that size matters. It looks to me like the bigger producers also have higher load factors (outside of the anomolous one in 2014). Of course, the bigger ones also tend to be the most recent, so all very difficult to interpret. You would need data from a previous point in time to disentangle the effects, but your aging hypothesis looks like the cleanest explanation to me.

James

Anonymous said...

There could be, at least 2 factors at work here. The reduction in aerodynamic performance by the lack of regular cleaning of the blades is one reason and the fact that the earlier offshore farms were not offshore at all, they're close inshore and relatively inexpensive to build, whereas the later farms are further offshore and in cleaner wind.

BillG said...

Pawb1 is right.

With the abandonment of many Round 3 projects in more exposed/distant waters, especially off the West Coast ( e.g. Celtic Array and Atlantic Array) it seems that offshore development is increasingly being concentrated on easier, closer, more economic East Coast sites and extensions to existing schemes with lower capacity factors due to wind resource.

This is probably more than balanced for operators by less wear and tear and downtime due to access difficulties.