National Grid says that the country has the electricity generating capacity to meet the average maximum need over the course of the UK winter. But this calculation critically depends on the reliability of power stations as well as an accurate assessment of the true generating capacity of each plant. This article looks at National Grid’s assumptions on power station availability over the next months and casts a somewhat surprised eye on its apparent errors, particularly in calculating the likely output from nuclear stations.

These mistakes – if they are mistakes - may not matter. The Grid has introduced new payments for cutting electricity demand, meaning that the spare capacity margin is around 3.4 GW or 6% of maximum expected demand in the average year. However what I believe may be its errors over nuclear power reduce this number by at 50% at the very least. It seems strange that the business at the centre of the electricity industry in this country appears to be substantially over-optimistic in its assessment of power supply. It seems it has ignored evidence published by EdF that its nuclear power stations cannot possibly reach the output that the Grid projects over the winter months.

Last year, National Grid estimated that the average availability of electricity generators would be 79.4% of rated capacity over winter 2013/14. The figures ranged from a low 25% for wind (for obvious meteorological reasons) to 97% for pumped storage plants. For plants subject to the possibility of mechanical or other failure, such as coal power stations, the number tends to be between 80 and 90%.

This year, even in the face of strong, repeated and growing evidence of declining mechanical performance of our ageing power stations, National Grid has increased its estimate of the reliability of the main types of power station, coal, gas and nuclear. Across all power plants, the expected availability rises from 79.4% to 81.8%. Perhaps this seems a small change. However it raises the amount of capacity the Grid expects to be ready to meet peak winter demand by about 1.7 GW. This is half the buffer that the Grid says will be available on the day of highest demand in the average winter. When margins are tight, apparently small changes really matter.

The errors in the nuclear forecast

Perhaps most strikingly, National Grid has raised its assessment of the nuclear fleet’s availability, and by more than any other major type of power station. It predicts that 90% of the UK nuclear capacity will be working at the point of maximum demand, up from 84% last year. In the face of repeated unplanned shut downs at EdF’s plants this year, I can think of absolutely no reason for this enhanced optimism. And, indeed, National Grid’s cheery forecast is not shared by Ofgem, which held its estimate at 81% availability, in its report in mid-summer. The Ofgem document actually predates the unplanned closures at Hartlepool and Heysham 1 that started a couple of months ago and I doubt Ofgem would be as optimistic today.

I looked at the performance of the UK’s nuclear fleet from early December to mid-February this year. Only for a couple of days did it actually achieve the 90% output that National Grid – based on information from operator EdF – suggested it will for 2014/2015. Average performance was 81% of potential, in line with Ofgem’s more conservative forecasts for this winter and last.

As I write this, only 3 of EdF’s nuclear generating units out of 16 (in eight power stations on seven sites) are working to their full rated capacity. A further 4 are operating at 20% below maximum power as a precaution. Sizewell (one station but two turbine units) is on a planned refuelling stop. 2 other units are suffering from mechanical faults and 4 are being inspected for a possible problem in their boiler units and will return to operation between now and the end of December, although at a lower output than previously. Another plant is returning to full power after refuelling.

The current state of the UK’s nuclear power stations as at 29^th October 2014

Station	Unit 1	Unit 2	Station output as percentage of maximum
Dungeness	Rising to full load	Out. Boiler pump failure	40%
Hartlepool	Out. Boiler inspection	Out. Boiler inspection	0%
Heysham 1	Out. Boiler inspection	Out. Boiler inspection	0%
Heysham 2	OK	OK	100+%
Hinkley Point B	OK but 20% power reduction	OK but 20% power reduction	80%
Hunsterston	OK but 20% power reduction	Out. Investigation of bearing vibration	40%
Sizewell	Out. Refuelling	Out Refuelling	0%
Torness	OK	OK	100+%

Wylfa (NOT EdF)	OK		100%?

The total nuclear output, including from Wylfa which is not owned by EdF, is currently (18.00 on October 29^th 2014) around 4.5 GW, or less than 50% of potential capacity. Only 3 stations (and I cannot even be sure about Wylfa) are working to full capacity). It certainly seems that National Grid is unrealistic in thinking that 90% of nuclear power will be available at the moment of peak need, which typically happens about seven weeks from today in mid-December.

In fact, we already know that 90% is actually not achievable. The total rated capacity of UK nuclear is – according to National Grid – about 9.6 Gigawatts. (Both EdF itself and Ofgem give lower figures, and National Grid surely should have noticed this, although the differences are small). 90% of the National Grid figure is slightly more than 8.6 Gigawatts. But, according to EdF’s own public statements, 8.6 GW is unattainable at any point this winter.

Heysham 1, unit 1, is said by EdF to be out until the end of December, past the point of likely peak demand. This reduces maximum output by about 0.6 GW. (As it returns to service, the second unit at Hinkley Point B moves offline, cutting power by almost 0.5 GW. So even if peak demand occurs in January, there won’t be additional capacity to meet it).
The other unit at Heysham and the two units at Hartlepool are subject to a 20% restriction on output when they return to service at some point during November or December. This cuts maximum output by just under 0.5 GW
The working power stations at Hinkley Point and Hunterston are also subject to precautionary power reductions of about 20%. This reduces potential output by about 0.5 GW.

In total, EdF’s fleet can only produce a maximum of 1.6 GW less than their rated output, or about 8.0 GW. This means that the availability of UK nuclear during winter 2014/2015 can only be 85% of the maximum potential, much less than the central National Grid assumption of 90%. This is before any additional mechanical or electrical problems. The reality is that nuclear output at critical times is, if recent experience is any guide, likely to be little more than 7 Gigawatts.

This reduces the UK’s spare capacity at winter peak by about 1.6 Gigawatts, cutting the safety margin by about 50%. A more conservative view of the reliability of gas and coal power stations would have an effect similar in size. If these numbers are correct, National Grid is being too optimistic in its Winter Outlook and the true position is that a typical winter will bring the UK far closer to power cuts than the company admits. A colder than average winter will make the UK’s position worse.

National Grid hasn’t responded to my written questions on Tuesday afternoon about the overstatement of nuclear availability and other issues.