Psted by ADMIN on behalf of our Australian friend.
Dear Dave,
In relation to the PNAS paper recently distributed, much has been made of a few theoretical studies on the geographic dispersal of windfarms and their alleged ability to back one another up. In Australia we have over 30 windfarms spread over many hundreds of kilometres. (I had the signal honour of carrying out the technical feasibility study for our very first.)
These charts show the aggregate data from 18 of the largest windfarms spread from Adelaide, to Melbourne, to Hobart to Canberra. The first chart shows the actual aggregate output from yesterday, the second the same aggregate output normalised to 100MW to remove any bias due to size.
Yesterday was not very special weatherwise but the strong correlation between outputs can be readily seen which results in a minimum generation less than 16% of maximum generation for the day. The normalised situation is slightly worse at less than 15% of normalised maximum. The contention that there is always wind blowing somewhere is shown to be
false.
So who is the author? and where are those charts? Hugh Sharman
The average capacity factor for Australia seems to be in the range 30-35%. So this article’s cherrypicked data by an anonymous author is that generation on one day was about half the average. Big whoop. That’s hardly ground-breaking news, and has nothing to to do with the article’s title or closing sentence: if the wind wasn’t blowing, then the amount generated would have been zero; instead, it was about half the annual average. So what?
Dear Andrew, sorry, wrong comparison (and I think you know it). From a grid control point of view, the comparison that matters is % of max. rated capacity, not the average. The writer has ‘been kind’ and only related it to the normalised maximum FOR THAT INDIVIDUAL DAY, which will be a LOWER nominal % swing.
The real, bad news is that this is an incontrovertible 85% WHOLE-fleet swing over a huge geographic area (far larger than UK – 1000s of miles) within a single day. The frequency with which this occurs is irrelevant – the WORST case has to be designed for by the Grid Operator re. backup systems. It’s only a single day, so other days might be even WORSE. Ramp RATE is very important too. By rough scaling off the diagram,that steep up-ramp looks like +85% in 5 hours, i.e. 17%/hour. Applying that to UK’s 33 GW, ‘kindly’ assuming % swing no WORSE in a much smaller area, gives a ramp rate of 5.6 GW/h, not the 2 GW/h that was being ‘bandied -about’ a few days ago. And don’t forget Paul-F Bach’s real West Danish (?2005) storm high-speed trip incident; they lost the whole fleet in just a few hours,and only back-up from LARGER neighbours with lots of rapid response hydro (i.e NOT the UK case) via Interconnectors ‘kept their grid afloat’ (just), at only c.20% wind penetration.
Regards, Chris.