NESO's Margin Call
NESO recently suffered their very own margin call, so is their Winter Outlook smart or cheating?
Introduction
The fallout from the blackout near miss we had on January 8th continues. We might characterise the events of that day as NESO’s very own Margin Call because they were forced to demonstrate they had the collateral to meet demand when it mattered. NESO (National Energy System Operator) has conducted a kind of post-mortem and put out a report claiming that they had far more operating margin than other analysts like Kathryn Porter have calculated. NESO claimed they had 3,709MW of margin, compared to Porter’s calculation of just 580MW. But NESO’s Winter Outlook boasted of higher operating margins than last year.
These events bring to mind the film Margin Call, where the chairman of the failing bank says:
"There are three ways to make a living in this business: be first, be smarter, or cheat."
Time to walk through how NESO decided that we had plenty of spare generating capacity this winter and work out if they are smart or cheating.
How Does NESO Calculate Operating Margin?
Back in October, NESO released its Winter Outlook that shows expected generation capacity, demand and operating margin as seen in Figure 1.
NESO go through all the potential generating capacity by fuel type and then apply a de-rating factor to recognise that not all generators will be available all the time because of maintenance, unplanned outages, or in the case of intermittent renewables lack of wind or sun. They then compare this de-rated generation capacity to their estimate of peak demand and the difference between them is what they term the de-rated margin. We might think of this as the contingency over and above expected peak demand.
They said of their forecast:
“The current Base Case de-rated margin is 5.2 GW (representing 8.8% of peak cold spell demand). The associated loss of load expectation (LOLE) is below 0.1 hours. This is higher than the 4.4 GW (7.4%) published in the Winter Outlook Report for 2023/24 and remains comparable to the Early View published in June.”
De-Rated Generation Capacity
The methodology looks reasonable until you dig into the detail of their calculations. First, let us take wind. They calculate that there is a total of 30,532MW of installed wind capacity across offshore, onshore and micro-wind generators. They apply a large de-rating factor to this capacity to produce a de-rated capacity of just 4,416MW. The trouble with this calculation is that according to NESO generation data, since the beginning of their forecast period on October 28th 2024, up until January 15th 2025, there have been 832 half-hour periods when wind generation was below this de-rated margin. Moreover, for 23 of the 48 periods that make up the 8th of January, wind was generating less than this de-rated capacity. The minimum wind generation this winter of just over 1,012MW occurred in November, and the minimum so far in January is 1,919MW. It would appear as though the de-rated capacity for wind power ought to be set at zero, like with solar, or certainly no more than 1GW. Now this of course would reduce the de-rated margin identified above from 5.2GW to 0.8-1.8GW, far from a safe level.
Second, let us look at interconnectors. We have several interconnectors to the Continent where we can import power from France (4GW), Netherlands (1GW), Belgium (1GW), Denmark (1.4GW) and Norway (1.4GW), giving a total of 8.8GW. There are other connectors to the island of Ireland, but they are mostly used to export electricity rather than import. According to REMIT data, for January 8th, 1,500MW of the French interconnectors, the entire link to the Netherlands and 728MW of the Danish link were unavailable, bringing maximum total available down to 5.5GW. It should be noted that the Viking Link was brought back at short notice, but the maximum import power was 6,284MW on that day and that coincided with peak demand. Imports were >300MW below the de-rated capacity in NESO’s Winter Outlook of 6,608MW and there was simply no more extra capacity available on the critical day of the 8th of January.
Now turning to biomass, the NESO Winter Outlook groups together biomass and waste incineration, with a combined gross capacity of 7,182MW and a de-rated capacity of 6,038MW. However, the maximum generation from the Biomass and Other categories in the generation data was just 5,138W and that occurred at 17:00, around the time of peak demand. According to REMIT, the Lynemouth and Teesside biomass plants were at least partially out of action with maximum capacity of nearly 700MW between them. It is not clear what extra biomass or waste incineration plants might have been available to take the generation capacity above the de-rated capacity, but as we shall see below NESO is not claiming any extra Biomass capacity was available.
It certainly looks like NESO is playing fast and loose with its de-rated generation capacity. Assuming 4,416MW of de-rated capacity for wind when actual generation can very often fall below that level is insanity. In addition, it is the height of folly to assume that any shortfall in domestic generation can be met with interconnectors from overseas because many countries are likely to be experiencing low wind conditions too with consequent power shortages. Moreover, the actual available Biomass capacity also seems to be well below the de-rated capacity. These capacity shortfalls alone reduce NESO’s estimate of the de-rated margin below 1GW at best and below zero at worst.
NESO Demand Forecast
Now we can turn to the effectiveness of NESO’s demand forecast. In their Winter Outlook they produce two different views of demand. The first is their “credible range” of peak national demand for Winter 2024/25 and the second is their weather-corrected daily peak transmission demand. Generally speaking, the weather corrected daily peak is within the credible range.
We can add up the number of half hour periods in which NESO’s record of transmission system demand is above the upper bound forecast and the weather corrected daily peak and the result is shown in the chart below. The earliest date is October 28th, 2024, and the data covers up to the end of January 15th, 2025. The result is shown in Figure 2.
So far this winter, demand has exceeded weather adjusted peak estimate for the day in question in 259 half-hour periods. Demand has even exceeded the upper bound demand estimate on 96 occasions. Sometimes the amount that Transmission System Demand Exceeds NESO’s forecasts can be very significant indeed as shown in Figure 3.
On November 21st, 2024, demand exceeded the weather adjusted forecast by 4.6GW (3.4GW above upper bound) and on January 11th, demand was some 5GW (2.8GW) above forecast. At 5:30pm on the near blackout day of January 8th, demand exceeded the weather adjusted peak by over 2.5GW.
It certainly looks like NESO is systematically under-estimating peak demand. In fact, sometimes the level of demand is above their forecast by almost as much as NESOs own estimate of the de-rated margin and well above any sensible estimate of margin if they applied proper de-rating factors to wind, interconnectors and biomass.
NESO’s Claims About Margins
In their post-mortem report, NESO claims that there was in fact 3,709MW of available headroom at 16:30 on January 8th as seen in Figure 4. Note the subtle change of language from “margin” to “headroom.” Such moving of the goalposts is usually an indication of cheating. Kathryn Porter has also produced some further analysis that challenges NESO’s claims here.
However, close analysis of NESO’s chart shows that most of this was made up of batteries and what they term “Flex Gen.” Batteries can only discharge for 1-2 hours before needing to recharge. According to their Clean Power 2030 report there was only 1.9GW of battery discharge capacity available at the end of 2023. It is stretching credibility to think about half of this remained available throughout the period given the high prices offered to supply the grid all day.
They do not explain what comprises “Flex Gen” and it certainly does not appear as a category in their Winter Outlook. However, there is up to 1,100MW (869MW de-rated) of Demand Side Response in their calculations. It does seem odd that they would rely on turning off some customers as way to meet demand.
Strangely, they do not seem to include the allegedly spare CCGT power plants on the chart in their calculation of 3.7GW of headroom.
It is also worth noting that in NESO’s Daily Operational Planning Margin Report, in the last forecast they have for January 8th, published on January 6th, they show an availability margin of just 473MW and that was for peak demand of 43,619MW, >3,300MW below peak transmission demand of 46,937MW and >2,200MW peak Net Demand of 45,823MW.
By way of contrast, while the UK was scratching around to find the last MW of generation capacity, France was steadily producing over 50GW from its nuclear fleet (see Figure 2, from Gridwatch). Yes, France was producing much more than our entire domestic generation from just its nuclear power plants.
Conclusions
There are significant weaknesses in the way that NESO calculates the de-rated margin in its Winter Outlook. The de-rated capacity for wind power is clearly in the realm of fantasy because we can clearly see that wind generation regularly falls below their de-rated threshold. The calculations for inter-connectors are also flawed because they are less likely to be available at times of peak demand that coincide with low wind conditions. The de-rated capacity for biomass also looks very suspect.
The demand forecasts are similarly a work of fiction because we can see that actual demand often rises above the upper bound of their “credible range” of peak demand and above their weather corrected peak demand forecast even more often.
If they adjusted the de-rated generation capacity to a more realistic level and revised their demand forecasts upward to recognise the actual level of demand, then their de-rated margin simply evaporates.
There’s another great line in the Margin Call film:
"Speak as you might to a young child or a golden retriever. It wasn't brains that brought me here; I assure you that."
It certainly looks like whoever produces the Winter Outlook was selected on the basis of ideological purity, not brain power. Yet, NESO are using this report to promote false optimism about the state of our electricity generation system. The document is simply junk, so bad that it calls into question the veracity of all NESO output, like the Clean Power 2030 report and their Future Energy Scenarios. Clearly, their processes require a significant overhaul and an injection of realism.
Their latest report, effectively a post-mortem of the January 8th debacle, creates more questions than it answers. On that day, NESO faced its own Margin Call and it was revealed that they are far from smart and look to be cheating.
This Substack now has over 3,300 subscribers. This growing interest has led to me being invited to give a talk to Sacred Cows on 28th January in London entitled “Net Zero: Why the cure is worse than the climate change disease”. Ticket details can be found here, do come along if you can make it
If you enjoyed this article, please share it with your family, friends and colleagues and sign up to receive more content.
In reality I suspect only the engineers operating the system will know how close we came to exceeding capacity and the lights starting to go out. I bet there was major stress in the control centre as ultimately the power for the whole country would have been in the hands of a few people. It would make for a good documentary.
Thank you, David. You and Kathryn Porter are doing a great public service.