The Changing Face of UK Power Supply

DECC publish an inventory of UK power generating assets annually in table dukes5_10. I discovered recently that prior years are not deleted but are hidden and easily accessible on the spread sheet for years back to 2004. This post compares 2004 with 2015, allowing the evolution of UK power generation to be examined.

A summary of dispatchable power is shown in Figure 1. While the amount of nuclear, coal and gas generation has changed significantly, the total dispatchable power has changed hardly at all, decreasing by 2.1 GW from 70 to 67.9 GW.

With peak winter demand of the order 55 GW there appears to be ample capacity margin of dispatchable power. 67.9-55 = 13 GW margin = 24%.

Figure 1 Nuclear capacity has declined by 2.64GW, coal capacity has declined by 8.21GW these partly offset by CCGT capacity that has grown by 6.77GW. Import capacity via inter-connectors has grown by 1GW and is included on the assumption that the UK will have access to 3GW of French nuclear. Hydro has added 60MW and biomass 833MW. The details of power stations opening and closing are tabulated towards the end of the post.

What is the Blackout Risk?

A blackout occurs when a large segment of the power distribution network fails leading to the lights going out. There can be 3 principal causes for this happening:

  1. System engineering failure that may be caused by a nuclear power plant tripping out, instability caused by highly variable renewable power or a storm bringing down a power line.
  2. System capacity failure, where demand exceeds supply, and power needs to be cut to a chosen area to protect the remainder of the system.
  3. System fuel failure, where for example the UK runs out of natural gas, leading to a capacity failure.

[1] is a relatively common occurrence. There are grounds to believe that ever higher levels of wind penetration may lead to increased risk of engineering failure in future. The UK has come close to [3] in the recent past as natural gas supplies were strained by a combination of factors such as the closure of the Elgin platform and the Fukushima nuclear accident in Japan. The sport of blackout speculation normally focusses on [2]. Since unforeseen outages are common place, all countries have a capacity margin in place as contingency, and the speculation recently has been that the UK capacity margin is now wafer thin. The data at face value quite simply do not support this contention.

When is the Blackout Risk?

UK power demand follows three cycles of daily, weekly and annual duration. It is when the peaks of these three cycles are coincident that UK meets its annual demand peak. It is at that time the risk of capacity failure is greatest. Demand is always highest at around 6pm, it is always higher Monday to Friday than at the weekend and it is always higher in winter – December to February. Therefore, risk of capacity failure will always be at around 6pm, on a weekday in winter. At all other times, demand does not press the capacity margin. In January this year, peak demand was 53.5 GW at 17:30 on Monday 19th.

Renewables and Averting Blackouts

Figure 2 includes data for solar and wind capacity. Solar we know can never contribute to averting a capacity failure in the UK since it is normally dark across the UK at around 6pm in winter when electricity is needed the most. Wind is a wild card. It is just as likely to be flat calm everywhere as it is to be windy. From 26th September to 4th October wind contributed virtually nothing to UK power supply. Since it cannot be relied upon, dispatchable capacity must be kept in reserve and paid for and this is why there has been virtually no change in dispatchable capacity in the UK since 2004.

Figure 2 Same as Figure 1 but with renewable capacity added.

Renewables, however, do displace gas and coal generation, and in so doing, fuel stores can be preserved during winter reducing the risk of [3] system fuel failure.

The Changing Face of Generation


Figure 3 Since 2004, four of the UK’s ageing Magnox reactors have closed. Wylfa is the only remaining Magnox reactor and is scheduled to close in December of this year. 2.64GW of nuclear capacity has been lost.


Figure 4 Since 2004, six large coal power stations have closed (Cockenzie, Didcot A, Floots point, Ironbridge, Kingsnorth and Tilbury B) with the loss of 8.2GW. Note that Ironbridge has converted to biomass with reduced capacity of 360 MW.

Gas – CCGT

Figure 5 The UK’s large fleet of CCGTs has seen a number of old small units close and some large modern units open. For example Pembroke (2.18GW), Staythorpe C (1.77GW) and West Burton (1.33GW). In total 6.77GW cpacity has been added.


Figure 6 The UK has a large number of small hydro schemes. While several new schemes have opened, two quite large old schemes have closed, Kinlochleven (30MW) and Fort William (62MW). Both schemes were linked to Aluminium smelters and I presume they may re-open to provide domestic electricity supply. The only new large hydro to open is Glendoe (100MW).

Wind and Solar PV

Wind and solar capacity was effectively zero in 2004. Different sources provide rather different figures for 2015 as detailed in Figure 7.

Figure 7 Installed wind and solar capacity in the UK from various data sources.

For the sake of consistency it is the dukes5_10 numbers that are plotted in Figure 2 but they do seem to underestimate reality. The Renewables UK and Renewable Energy Foundation (REF) numbers seem to be closely alined for wind. The latter are based upon subsidy uptake and are likely to be the most reliable. But when it comes to securing electricity supplies at 6 pm on a weekday in winter the difference between 16.3GW (DECC) and 20.4GW (REF) is neither here nor there. On many days each year, the actual generation from these sources will be effectively zero at the time of need.

Concluding Comments

During review, Roger Andrews sent this chart (Figure 8) from National Grid. It correctly assumes that all power stations may not be available at the time of need and that is why a capacity margin is required.

Figure 8 Derated capacity margin for UK power generation from National Grid.

There are four main points to me made from this post:

  1. Dispatchable capacity in the UK has declined 2.1GW since 2004 which is not a material difference since electricity demand has also fallen in that period.
  2. The nameplate capacity margin is approximately 13 GW which seems ample contingency for plant outages. Increasing this margin would involve paying companies to keep higher unused capacity in reserve.
  3. The risk of capacity failure is always around 6pm on a weekday in winter and only lasts for a few hours each day.
  4. 13.5 GW of wind and 7 GW of solar cannot be relied upon to provide any supply at 6 pm on a winter week day when the blackout risk is greatest.
This entry was posted in Energy and tagged , , , , , . Bookmark the permalink.

32 Responses to The Changing Face of UK Power Supply

  1. dereklouden says:

    Great analysis of the changes. The importance of the Interconnectors with France and the Netherlands should be stressed. If these fall over or the throughput is cut at a time of peak demand we’d be up excrement tributary without a convenient means of manual propulsion.

  2. matthew_ says:

    Nice. This is important stuff for a reliable grid.
    Does the UK grid operator publish any of their reports on reserve margin calculation?
    As I understand, calculating reserve margins is a probability calculation, where the grid operator attempts to keep the risk of blackout during peak load below some probability. In this probability calculation even wind power is assigned a factor of “availability” since there is a real probability that wind is contributing power during the peak load event. Some grid operators may assign wind a 0 probability factor as you have done here, but I think this only occurs in places where wind penetration is so low that it makes no difference in the calculation.
    The Norwegian grid operator Statnett has used 5% available wind in their report for Norwegian peak load winter 2011/2012. Sweden’s grid operator Svenska Kraftnät (SK) used 6% available wind in their report for winter 2011/2012. As a comparison, SK uses 90% availability as the correction factor for both gas and nuclear power. For hydro SK does a specific calculation due to the connections between stations on the same rivers, and ends up with an availability for that winter of 85%. If it is interesting to have a link to the Norwegian and Swedish reports I’ll add that, but they are written in Norwegian and Swedish.

    • matthew_ says:

      In the report from SK for the winter 2014/2015, they specify that solar power is credited with 0% since the peak load normally occurs when it is dark. 🙂
      Wind is maintained by SK at 6% availability following a Nordel report and an internal SK evaluation.

    • Euan Mearns says:

      National Grid methodology is here – pdf alert. Correct, they use a probability approach. There is good chance that wind will contribute something during the winter peak demand period. But it is the high probability that on occasions it may not that needs to be planned for.

      • matthew_ says:

        This comment ended up a little on the side of your original post regarding the changes in production capacity over time, and thanks for the link to National Grid. I was surprised at how optimistic they are regarding the probablity that wind will contribute during peak load. They show 23% (Table 16, p.55) which is much higher than the Norwegian and Swedish grid operators. I’m curious how they calculate such a high availability.

      • Willem Post says:

        Wind as a capacity source is derived from a probability calculation for LONG TERM planning purposes.

        It has nothing to do with dispatch value, which is zero, or near zero.

        In Texas, with over 13000 MW of wind turbine capacity, about 6.8% is counted for long term planning purposes.

  3. Günter Weber says:

    I am surprised that interconnector capacitiy did not change much in the last 10 years. Are there significant upgrades ongoing/planned for the near future?

    • Leo Smth says:

      Inter-connectors are privately funded and must relay on arbitraging the electricity prices across the continent. They are not there to balance the grid – they are there to make money.

      There is a lot of capital and quite high maintenance costs in them.

      Also, when its bad demand wise in the UK, its bad everywhere we connect to. So they dont actually help that much.

      I remember some years back our coal plants running flat out in Feb, exporting to France…

      • Roger Andrews says:
        • Willem Post says:

          That is a lovely graph.

          The UK’s growing dependence on FRENCH nuclear, AND that imported nuclear energy likely had a lower cost than the REAL cost of wind energy.

          Is there a message here?

          • I think the message is clear. If a bitterly cold, windless high pressure system settles over Europe this winter the UK may find itself exporting 3GW of British gas rather than importing 3GW of French nuclear.

          • Euan Mearns says:

            Correction! The UK may find itself exporting 3GW of Qatari gas, either as gas or electricity.

      • Günter Weber says:


        when I see the big share of gas-fired plants in UK, I assume that production costs of electricity are significantly higher than in France and the Benelux countries (or Germany). There gas is driven out of the market. Are interconnectors such expensive that they offset this price difference? Or is their availability much lower than that of a conventional power plant?

  4. Math Geurts says:

    The new challenge. 2023: UK without coal, Germany without nuclear, France already now, in the winter, some times importing from Germany.

    • Euan Mearns says:

      Thanks for the link. This is BSFC stuff!

      Energy secretary Amber Rudd is reported to be planning to announce the coal phase out before the start of the UN climate summit in Paris next month. Under the proposals, Britain’s ten coal-fired power stations would either have convert to alternative fuels such as woodchips, to fit carbon capture and storage equipment or close, said The Times.

      So the options are:

      1) Burn virgin forest
      2) Make electricity so expensive no one will use it
      3) Shut up shop and go back to the pre-industrial era

      • Willem Post says:

        The UK imports wood pellets from the US. This is a very energy intensive activity that causes grossly excessive CO2 emissions/ kWh delivered to the UK grid.

        The emissions are measured from harvesting the wood in the US South, producing the pellets, transporting them to plants in the UK.

        Excerpt from this article:


        A 2013 study, published in Environmental Research Letters, analyzed the CO2 equivalent emissions of exporting wood pellets from the US Southwest to the UK.

        A breakdown of the biomass lifecycle, according to GHG emissions, is as follows:

        See Table 4, which shows 5 of the 7 CO2 emissions components.

        – Pellet production accounts for about 48%

        – Shipping the pellets across the Atlantic Ocean accounts for about 31%

        – Burning the pellets accounts for about 10%*

        * Emissions due to combustion are about 1.8 kg of CO2/kg of pellets.

        That means the A to Z process of getting wood from the forest, turning it into pellets, transporting the pellets from the US to power plants in the Uk, and burning the pellets, would release about 1.8/0.1 = 18 kg of CO2/kg of pellets.

        If the power production is at an efficiency of 30%, then 7,750 Btu/lb of pellets x 2.2 lb/kg x 0.30/(3,413 Btu/kWh) = 1.5 kWh/kg of pellets would be produced, or 18/1.5 = 12 kg of CO2/kWh for the A to Z process, if CO2 sequestering by regrowth would be ignored.

        EVENTUALLY, 100% sequestering would, at the very most, offset 2 of the 12 kg!!! Such an environmentally harmful way of having the UK, Germany, etc., meet their EU CO2 obligations should not even be allowed to exist by EU rules, and the US should not be aiding and abetting. However, some folks are making money.

        This is a far worse boondoggle than the US corn-to-ethanol program, which, on an A to Z basis, is about CO2-emission neutral, but is derided by the EU.

        The US Southeast exported to Europe about 1,650,000 ton and 3,250,000 ton of wood pellets in 2012 and 2013, respectively; likely 5,7 million ton in 2015.

        See URL, with photos, regarding the unsustainable clear cutting of US Southeast forests to enable Germany, UK, etc., to meet the EU CO2 emissions standards, because the EU declared biomass emissions to be CO2-free!! Germany, the UK, etc., are co-firing the pellets in their coal-fired power plants!

        In the US Southeast many forests are managed. It takes about 20 – 25 years from harvest to harvest; in Maine about 35 – 40 years. One may wonder how long it would take to deplete the soil to significantly affect crop yields. If 3,250,000 ton were exported in 2013 (a lot more was harvested but not exported), that would be 1,300,000 cords/yr of wood being cut from a given area, and a same area being planted that has just been cut, etc. That means about 20 – 25 such areas are in various growth phases at any point in time; more area if more tonnage is exported.

  5. Leo Smth says:

    There is a slight correlation between wind and demand. Unfortunately it is the wrong way round. The coldest times are often when a continental high pressure system reduces wind power to zero.

    This is not always the time of maximum heat demand though, as the worst conditions for heating are very cold with a moderate wind. That strips heat out of building more than anything else.

  6. Gaznotprom says:

    It’s about de-industrialisation (of the West). Our Coal fired are literally demolished, any useful parts exported…

  7. Nick Perrin says:

    Euan: 1) Are STOR capacities included anywhere?
    2) Could you comment on the impact of Eggborough and Longannet closures – loss of over 4 GW due soon. Thanks for a detailed analysis. Nick.

    • Euan Mearns says:

      Nick, I had to look up STOR capacities – short term operating reserve. Dukes5_10 has many GW of small generators which may come into that category. I haven’t included them.

      Wylfa to close this year and Eggborough and Longannet next year. Another 4.5 GW out the window. How I suspect this is covered is by mothballed CCGTs. I don’t know how much of my 13 GW margin is mothballed CCGTs. In their analysis, National Grid are probably working with a lower gross margin number than 13 GW and so what happens in the next couple of years is that we see mothballed capacity coming back on line keeping the effective gross margin constant. And then we hope that gas supplies hold up, but the UK does have some large new gas production coming on. Elgin is re-drilling wells and Lagan Tormor should be on soon when they get the gas plant finished.

      Longannet closure is more important for Scotland – our last baseload FF generator. Peterhead is held as reserve capacity.

  8. PhilH says:

    One factor that is only mentioned in passing in the first conclusion is that the UK’s electricity demand has been falling since about 2005, by about 1% a year, both in total amount and in peak demand (see third figure in So, if this trend continues through 2015, the peak demand in winter 2015/16 would be about 500MW less than winter 2014/15, and about 5000MW less than winter 2004/05. Fig 1 above shows that the dispatchable capacity, exc interconnectors, has only fallen by about 3000MW, so the capacity margin ought to have improved since then, though that doesn’t seem to be what’s generally concluded in recent commentaries on the situation.

    • Euan Mearns says:

      Phil, It has become a new national sport to speculate about blackouts. I did write about this but then removed all references to it in order to keep the post factual. Electricity demand is falling for four main reasons 1) improved efficiency, 2) higher price, 3) recession, 4) distributed generation.

      The latter is solar PV + small generators that are not metered by BM reports but which are seen as negative demand.

      Why National Grid and Ofgem now like to speculate about blackouts when it should be their responsibility to ensure that they don’t happen is an interesting question. It is also true that the anti-renewables lobby enjoys this speculation along with freezing winters and excess winter deaths.

  9. Derek G Birkett says:

    Derated capacities have been assessed on an historical basis. Most coal and nuclear have reached an age when the ‘bathtub’ curve of plant incidence failure is upon us. Capacity credit for wind has been assessed at 23% giving a lift of just over 3GW for available capacity assessment. There is a ‘cushion’ of 2.5GW of balancing reserve not included in this assessment that is not yet finalised by National Grid
    Nuclear cannot really be considered as dispatchable power in an operational sense. Coal can store a year’s supply of fuel at site and is better suited than gas for coping with intermittence. Germany has a programme for 25GW of fossil-fired plant in progress, mainly coal. The GB grid could not function without fossil-fired generation.

    • Euan Mearns says:

      Derek, it sounds like you know what you are talking about, but I disagree with a number of things you say:

      Most coal and nuclear have reached an age when the ‘bathtub’ curve of plant incidence failure is upon us.

      I thought that EDF taking over our nukes from British Energy had transformed their reliability.

      Capacity credit for wind has been assessed at 23% giving a lift of just over 3GW for available capacity assessment.

      What is the probability that wind will deliver effectively zero power 6 pm on a weekday in winter? The probability may not be 1.0 for a given year, but its certainly 1.0 over a period of 2 to 3 years. Wind is currently churning out 0.7 GW. I’m not sure exactly what your statement means, but I don’t think wind has been above 3 GW for three weeks.

      Nuclear cannot really be considered as dispatchable power in an operational sense.

      Nuclear is of course dispatchable, its just not flexible like coal and esp gas. I agree with you comment about coal and the security it provides, but the UK I gather is planning to shut down all coal within a decade.

      • Willem Post says:


        Some of the French reactors are old-following.
        It is beyond rational for some folks to say nuclear is not dispatchable.
        Tell that to the French which have had over 75% of electricity on their grid for about 3 decades.
        Only in RE la-la-land is it not dispatchable.

        As I noted earlier, total installed wind turbines, MW, has a capacity value, MW, for LONG TERM PLANNING purposes. In Texas it is 6.8% of installed.

        The installed wind turbines have zero capacity value for dispatch.

  10. manicbeancounter says:

    I know I am being a bit pedantic, but 4 nuclear power stations and 1 coal power stations have earlier build dates in 2015 than in 2004.

  11. Derek G Birkett says:

    Thankyou for responding. I was not too clear with the 23% capacity credit for wind. My point being with any peak demand asessment especially in winter cyclonic conditions the 23% figure is much too high for an intermittent source. At the beginning of the decade with the cold winters, peak demands reached 60GW on five occasions when the wind contribution was between 2.5% and 5.5% of grid metered wind capacity. The question to ask with any probability assessment are the consequences if it fails to materialise. Again this was my point with the incidence of plant failure reaching the end of its days. I am a retired grid control engineer.

    • Euan Mearns says:

      Derek, after replying to your initial comment I Googled a bit and found your submission to Scottish Parliament consultation. I grew up in Angus, have lived in Aberdeen much of my life but view Perthshire as my home – Taybank in Dunkeld is my local. Visited the Pitlochry dam hundreds of times. 15 MW output (from memory) equivalent of 3 large off shore turbines. Which is more valuable? And which does greater environmental harm? The puffin or the smolt grinding machine 😉

      Experienced commentary is highly valued here!


  12. Nick Perrin says:

    Euan, Thanks for your evaluation of STOR. I have downloaded the Dukes 5 data and need to study myself. I am surprised you conclude there is ‘many GW of small generators..’ What makes up STOR? I assume its from diesel generator sets grouped in strategic places plus public enterprises standby generators (NHS etc.) Wikipedia ( has them sized up to 2 MW so some thousands are going to be needed to get into GWs. Do you think this is the case?.

    PhilH A quick read of the link you gave talks of reductions in demand being also due to solar coming on line as well as true reductions in demand. I use Gridwatch figures and the demand is also abated by solar estimates but I see the reduction (overall) in demand since early 2012 as being 5%,5.6 % and 1 % (2014). My point here is also the peak proportion that is derived from coal, being 48%,47%,42% and 36% early this year is falling steadily. This is the reliable baseload.
    Being made up by a growing proportion of solar and wind leaves us exposed gradually more and more to the weather!

    Sorry to go on. I wonder if Gridwatch can now include things like STOR and the range of mothballed units being fired up.

    Great website and lots of insight – thanks Euan.


  13. clivere says:

    Richard North covered STOR a few times for example here

  14. Nick Perrin says:

    Thanks Clive,
    Though 2013 is not long ago I did not remember the post given in your URL – So this triggers more thoughts on the subject of STOR. I really feel like giving up now . You can throw money at problems and solutions can be found. It is also very significant if all these generators can be remotely controlled as groups as and when required. Local planning authorities seem to have been very ‘understanding’.

    Thanks again,


Comments are closed.