A remarkable coincidence

Four national institutions have failed to model the 2050 energy system correctly, and all of them in ways that lead to understatement of the costs of Net Zero.

Over the weekend, the Sunday Telegraph reported that the Climate Change Committee has got its energy system modelling wrong. The revelation was made by Sir Christopher Llewellyn Smith, the lead author of the recent Royal Society report on electricity storage, in remarks made at a seminar at Oxford.

According to Sir Christopher, the Climate Change Committee’s estimates of the costs of Net Zero are fundamentally flawed because they have only modelled isolated years. As he pointed out in the seminar, low-wind years can happen back to back, which means that the Climate Change Committee need twice as much storage capacity as they thought. As a result, they have underestimated the costs.

However, the Sunday Telegraph didn’t mention that it’s not just the Climate Change Committee that has made this mistake. In the same seminar, Sir Christopher pointed out that the National Infrastructure Commission has done the same thing, despite being warned of the problem of clusters of low-wind years. So they too will have underestimated the costs.

The National Infrastructure Assessment…is also based on one year…they were told by the Met Office ‘you can get extreme events’…it’s not enough to look at one. They looked at one, so they got the answer wrong. The Met Office are really angry, because they told them ‘don’t do it’, but they did it.

I can also reveal that National Grid ESO, in its Future Energy Scenarios, has done the same thing. I wrote to the NGESO team to ask how they did things, and was told that their models are prepared using weather conditions in 2013, which they describe as an “average year”. They are starting to run tests against low-wind conditions (so-called ‘dunkelflautes’), but back-to-back wind droughts don’t seem to be on their radar yet:

The generation provided from renewables, as well as the demand profile, is typically based on an average weather year (2013).

For FES23, we also conducted an initial piece of analysis looking at abnormal weather conditions (resulting in abnormal supply and demand patterns), the results of which can be found in our FES23 publication under the title Dunkelflaute Period. We took a period of extreme weather, in this case between Jan-Feb 1985, and applied it to our Consumer Transformation scenario in 2050, to look at how the system would respond to a sustained period low renewable output…

We are planning on looking at abnormal supply and exceptional demand in more detail going forward as well as the effects of more extreme weather.

That means that they too will have underestimated the cost of Net Zero.

The Royal Society is to be congratulated for clarifying the problem. However, it turns out that their own modelling is fundamentally flawed too. That’s because, while they model 37 years of different wind speeds, they assume that electricity demand is always the same. Sir Christopher has admitted that this is not correct, in a podcast broadcast last year. As he put it then:

And now I confess something that is a bit of a weakness in our report. We’ve got this model of one year of demand…based in the weather in 2018…We simply repeat that 37 times.

This is clearly wrong, because in 2050 it is imagined that we will all heat our homes with electric heat pumps. Electricity demand will therefore be much higher in cold years than in mild ones, and if we have back to back cold years, we are going to need much more storage.

So, four well funded national institutions have failed to model the 2050 correctly, and all of them in ways that low-balls the cost of Net Zero. That’s a remarkable coincidence, and one that should probably raise alarm bells about the extent of the rot in the British establishment.

Appendix

Film of Sir Christopher’s comments can be seen in the video below.