Will the weather save us from the energy crisis?

Eh, no

As I have noted here before, the media in Britain — and from what I can tell, in continental Europe also — seem to be trying to sell the idea that the energy crisis is either over already or not as bad as it seems. The most striking thing about this coverage is that it either:

1. Never cites any actual data to makes the case, instead vaguely alluding to a broad collective gnōsis that has realised that the previous fears of a serious energy crisis are now consigned to the past.

2. Some data is cited but it is vague and often attributed to third-party publications by consultancies and/or pressure groups; and when you dig into the third party statistics they are in turn either vague or inaccurate.

Another trope that I have noticed emerge is that we are in for a warm winter and so this should keep the pressure off the grid. Once again, the data and arguments cited tend to be vague or nonexistent. So, let’s remedy that — for the case of the UK, at least.

I am going to use data from 2010 onwards. Global temperatures are rising — albeit it at a very gradual pace — so we need a sample in which there has not been too much of a secular upward trend. Post-2010 fits that bill. We are also going to focus on electricity, not on gas. This is because the government emergency plan aims at curtailing electricity consumption, not gas consumption in a crisis scenario. (Electricity can be turned on and off, but if you turn off the gas you need to repressurise the pipe before reactivating it).

The following scatterplot shows the relationship between electricity demand in the fourth quarter of the year versus the average September-October temperature of that year.

A few things stand out.

1. The relationship between the variables (R2) is weak. Very weak, in fact. It is not actually statistically significant. This means that, even if temperatures were very high, this would not be a reliable guide to electricity demand.

2. The dot highlighted in red is 2021 and, as we can see, that year had a high Sept-Oct temperature and low electricity demand in Q4; this is worth noting as gas imports etc for this year are often compared with last year.

3. The dot highlighted in orange is a forecast for Q4 electricity demand in 2022 based on current temperatures; note that the temperature implies that a fairly average amount of electricity will likely be required.

Here is the same data in a time series chart. Not the final observation for Q4 electrity demand is our forecast.

Two further points stand out here.

1. Sept-Oct in 2022 is certainly above average — so when media people say that they are not lying — but it is not dramatically so, and it is certainly not unprecedented these past 12 years; similar or higher temperatures were observed in 2021, 2014, and 2011.

2. Our forecast of electricity demand for Q4 2022 — for what it is worth, because recall the statistical significance is low — is higher than in both 2021 and 2020; once again, this is worth paying attention to when comparing gas imports etc this year to those two years.

All in all then, these past two autumn months have shown above average temperature. But not dramtically so. However, there is no firm relationship between tempetature and electricity demand. And even if we take the extremely weak relationship that is there, it appears that the UK will need more electricity available this year than we saw in 2021 and 2022.

The weather will not save us.