Andrew Montford: Does the Climate Change Committee understand the energy storage problem?

By Paul Homewood

 

 

Yesterday, I reported that four national institutions – the Climate Change Committee (CCC), the National Infrastructure Commission, National Grid, and the Royal Society – have got their energy system modelling wrong and have thus underestimated the cost of Net Zero.
Last night, the CCC’s Chief Executive, Chris Stark put out a long Twitter threadaddressing these issues. But while it’s dressed up as a rebuttal, it’s nothing of the sort. In fact, it’s a masterpiece of bureaucratic obfuscation.

Recall firstly that this blew up when the Sunday Telegraph reported Sir Christopher Llewellyn Smith’s criticisms of the CCC’s energy system modelling: they had failed to look at the possibility of back-to-back low wind years. This meant that they underestimated the amount of hydrogen storage the system would need, and thus the costs involved.
There are 24 tweets in Stark’s thread. On number 10, we get this:
“We could certainly look further at a sequence of years. We are hoping we can do this in later work.”
Clearly then, Stark accepts Sir Christopher’s central point. He would have had to, of course, because he had already done so in correspondence with the Sunday Telegraph’s Ed Malnick, who reported in his article:
“…in response to further questions from this newspaper, the [CCC] admitted that its original recommendations in 2019 about the feasibility of meeting the 2050 net zero target, were also based on just one year’s worth of weather data.”
And since the CCC had the underlying modelling for the 2019 Net Zero report dragged out of them under FOI, we can see in the model itself that only one year’s worth of data is analysed!
But while Stark has to accept the point, in true bureaucratic fashion, he dresses it up so that it appears to be a rebuttal:
* quote tweeting someone saying that the Royal Society’s criticisms are misleading
* calling the Sunday Telegraph piece “nasty” (it isn’t) but not linking to it
* multiple tweets describing the (incorrect) modelling that they did
* claiming to have made a strong rebuttal.
* saying “there’s nothing ‘right or wrong’ here.
* calling it a “silly story”
* etcetera.
Stark introduces a 2023 report, for which he says they looked at five different years of weather data, so he is once again accepting Llewellyn Smith’s central criticism, namely that they haven’t looked at back to back low wind years and will thus have got the costs wrong.
He also says:
“we modelled two sensitivities looking specifically at the impact of low-wind periods (‘wind droughts’) up to 30 days. An understanding of these extremes is essential to system design (although its impact on the overall net zero transition shouldn’t be exaggerated).”
This appears to betray an alarming misunderstanding of the issue. A period of a few weeks with little or no renewable generation (usually referred to as a “dunkelflaute”) is a secondary problem. Dunkelflautes are typically a couple of weeks long, but even one lasting 30 days would only reduce annual output by 10% or so. In simple terms, it would mean that we would need 10% of annual demand in the store at the start of the year.*
I use the term wind “drought” to refer to years in which wind is low over the whole year. In 2021, for example, annual wind output was down 20% or more. To get through a year like that, we’d need 20% of demand in the store. To survive back-to-back wind drought years, we’d need to store 40% of demand (and to have a commensurately larger generation fleet so that we can quickly refill it). Thus the costs will be grossly understated.
That Stark appears not to understand this, even after Llewellyn Smith has explained it to him, should be a cause for concern.
It may be, of course, that bringing dunkelflautes into the thread is just part of his efforts to obfuscate his admission of failure, but we need to be clear. So, does Chris Stark accept that back-to-back wind droughts mean more storage, more generation equipment and higher costs, or doesn’t he?
We need to know.