One of the most frequent criticisms levelled at wind power is variability. That is, when the wind drops (or blows too hard) the windmills stop spinning and you get no power. To begin with, Oswald simulates the output rises and falls that might result from a lot of windfarms distributed around the UK by using Met Office archived data from different points up and down the land. Many wind advocates have argued that with enough windfarms, widely enough distributed, you would get more reliable power output as some windmills would always have wind.

Oswald's analysis says this isn't true, with calm conditions across pretty much all the UK being fairly regular events.

Analysis from 1996 to 2005 shows similar results: large, rapid, and frequent changes of power output being common occurrences ... any national power system has to manage under the worst case conditions likely to occur ... These are not extreme cases, whose frequency is so low as to render the events negligible. Rather, these are representative ...

If the government succeeds in building its mighty 25 gigawatts of wind base by 2020, according to Oswald's Met Office data-based model its output will dip to pretty much nothing fairly routinely.

In his view, all this means that - certainly in a 2020 timeframe - the only feasible backup for the planned 25-gig wind base will be good old gas turbines. These would have to be built even if pumped storage existed, to deal with long-duration calms; and the expense of a triplefold wind, gas and pumped storage solution would be ridiculous. At present, gas turbine installations provide much of the grid's ability to deal with demand changes through the day.

The trouble is, according to Oswald, that human demand variance is predictable and smooth compared to wind output variance. Coping with the sudden ups and downs of wind is going to mean a lot more gas turbines - ones which will be thrashed especially hard as wind output surges up and down, and which will be fired up for less of the time.

Even though the hardware will be cheap and tough, it will break often under such treatment; meaning increased maintenance costs and a need for even more backup plants to cover busted backup plants. Thus, the scheme overall will be more expensive than the current gas sector. And since people won't want to thrash expensive, efficient combined-cycle kit like this, less fuel-efficient gear will be used - emitting more carbon than people now assume.

In essence, wind plans aren't actually wind plans, according to Oswald. They're gas plans with windfarms used to reduce the amount of gas actually burned in the plants. But he thinks the assumptions now made on costs and emissions reductions to be anticipated are unduly optimistic.