Is Wind Power Reliable? Capacity Credit of Wind Energy

Is Wind Power Reliable


The following is a commentary on David Milborrow’s article in “New Power UK/Issue 1/February 2009”.


As David says, you would not design a thermal power generating system which did not have built in reserve. He has answered his comment about those letter writers being unconcerned (or unaware) that there was a shortage of nuclear output during the cold snap in early January.   “Why should they have been concerned/aware?”   The system is designed for this.


David quotes research to back up his assertion that wind has a capacity credit and the reserve margin of conventional plant may be reduced.   He indicates that based on statistical analysis this reduction might be 7 GW if 40 GW of wind power was to be installed.  More than the reserve available to the National Grid during the cold spell in the first week of January 2009.  He overlooks that two of the researchers, Grubb and the Oxford Economic Research Associates (OERA), quote periods bereft of wind.  For example:-    


  • Grubb says there are on average 6 hours per month during the winter and twice this amount during summer when the wind is low or non existent.
  • OERA say that there will be 23 one hour periods in the year when the output from wind turbines will be less than 10% of the rated output when demand is between 90% and 100% of peak demand. 


If, when preparing his Table 2, David had studied the wind speeds over the whole of the UK he might have drawn a very different conclusion.   His table quotes the recorded output of 1288 MW of wind turbines all of which are located in Scotland.   I do not have the record for the 6th of January 2009 but do have the record for the 7th January.   On this date, at 17.00 hours,, recorded a wind energy output of 140 MW.   It was a cold day, like it was on the 6th of January, and the remainder of the UK was experiencing calm low wind speed conditions.  The electricity demand at 17.00 hrs was 58.9 GW, within 3% of peak simultaneous demand    If a calculation based on all the onshore Met Office wind speed observations and on the offshore wind speed observations for 17.00 hours had been carried out, it would have found that if 40GW of wind power had been evenly distributed around the UK and the seas surrounding the UK the average output would have been 2.46 GW, 6.15% of the installed capacity.  The installed capacity for this calculation in Scotland was assumed to be 9.0 GW.


When broken down into regions and the installed capacities are redistributed as shown in the following table the average is similar.   Assuming an installed capacity of 1.29 GW (1288 MW) the output for Scotland would have theoretically been 0.13 GW (134 MW), very close to the output recorded by    It is also noticeable that the theoretical output from a large array of wind turbines located in the North Sea off the coast Lincolnshire and East Anglia would have been zero. 


If you then consider that, 15% of the 8000 to 10000 turbines required to generate 40 GW might be out of action for repair and >75% of the remainder would be net consumers of electricity, keeping their onboard systems alive, and the losses in transmission might be in the order of 25%, then the output to meet demand would be considerably less than the value estimated and be almost negligible as far as the National Grid was concerned.




Installed Capacity GW

Theoretical output GW






England, Wales and N Ireland [1]




North Sea East of Murray Firth [2]




North Sea off Yorkshire/ Lincolnshire/East Anglia [2]




Irish Sea




South Coast and S and W coasts of Wales [2]




Totals [3]





Notes to the Table: –

* Installed capacity monitored by

[1] The capacity assumed is 11 GW less that recorded by

[2] The totals are 29 GW.

[3] The total of 40GW is the total for onshore and offshore wind discussed in the Carbon Trust Report – Offshore Wind.  They assumed 11 GW onshore and 29 GW offshore.

[4] To allow for wind shear with height the Met office onshore wind speeds were increased by 35% and the xcweather offshore wind speeds were increased by 25%.   The detailed calculations can be obtained from                              

My studies have shown that a high proportion of recent simultaneous maximum demands on the National Grid have coincided with periods of low wind speeds, not windy conditions as reported by Palutikof et al. and some of these periods have lasted for more than 24 continuous hours.


David then says that no backup for wind is required because that backup already exists in the form of conventional power generation.    Obviously this is the case.   If there was no conventional backup and the UK relied solely on wind power large parts of the country would be without power for considerable parts of the year and the UK would have to build a conventional power generation system with the same margin of reserve as exists today.


David then states that in normal circumstances 26 GW of wind power, a 20% wind energy penetration, only displaces 5 GW of thermal plant.  That is, 20% wind energy penetration only reduces the requirement for conventional generation by 11%!




If any myths are to be debunked they are not those of the wind sceptics.


Post Script


Investment in wind energy is a luxury.  It is like investing in a second parallel electricity generating system.   The first is a conventional system where the reserve margin over demand is 25%.   The second is a wind powered system where the reserve margin is 125% and is wholly provided by the first system.   The main benefit of the second system is that it saves fossil fuels/carbon emissions but not significantly.      


David‘s article does not consider the cost of wind energy compared to conventional power generation.   He refers the reader to the Carbon Trust Report – Offshore Wind Power. This publication makes the assumption that the capacity credit of wind allows 6 GW of conventional power to be retired.   As a consequence, their calculations significantly under estimate the real additional cost of electricity when 40 GW of wind power is added to the UK generating capacity [1].



[1]   Critique of the Carbon Trust Report Offshore Wind Power: big challenge, big opportunity   by D Stephens posted on the Claverton Website





Posted by D S Stephens

Please note there is an extended learned discussion continuing on the topic of this article on Claverton Email Service

See –


  1. Reply

    Denis, where do your figures of 25% transmission losses and 15% unavailability come from?

    In the region/capacity table, is there a formatting problem: could it be that the percentage figure shown for Scotland of 0.10 is actually 10%?

    As you’ll know, a 20% wind penetration can reduce GHG emissions from electrical generation by over 20%, as it’s the high-carbon, dirty coal that can get taken offline first; so is it that 20% isn’t significant to you, or have you assumed some other mechanism in place?

    And from the POV of reserve capacity, there’s the big question of what capacity of interconnectors the system has available: 29GW offshore wind looks like at least 29GW of potential interconnector capacity, using the North Sea Ring model. And how much on-call capacity from grid-connected gensets? Do biomass CHP plants count as “conventional capacity”? How much responsive demand is there in this particular system that can be postponed?

    After all, as a lot of that peak power capacity is only needed a few hours each year, building fossil power stations for it seems a bit daft – much cheaper to use a combination of interconnectors, gensets, pumped hydro, a smart grid with responsive loads, etc.

    As ever, in questions of intermittency, these partial spreadsheets are amusing diversions, but don’t really tell you how the system would behave: that’s what system models are for.

    It would be wonderful if wind power were a luxury, and we didn’t need to decarbonise, and there was no global warming. But hey, that’s reality, and a large wind turbine fleet for Britain is crucial and inevitable.

  2. Reply

    Received from Dr Mark Barrett:-

    Dear all

    Apologies for the nth broadcast of this.

    These materials show how renewable generation with back-up, storage and transmission allow for reliable electricity services. More detailed work is required, as always, but I think it clear that existing facilities plus new fossil/biomass back-up generation/storage/transmission as required will be adequate. As the modelling shows (and as Dave A says), keep fossil for low capacity factor output as required until storage/transmission replace it. CCS is not so low in carbon (about 75% reduction per kWh and we need ~100% in elec to leave room for aviation etc,) and nuclear is incompatible with high renewable and will shut it out for decades because it has to be baseload for technical and economic reasons. (This is quite apart from the security risks, ethics, politics, economics etc. of these options.), We should start the transition to 100% renewables now without recourse to the last gasps of finite fuelled options.

    Note the importance of heat storage for load management. Currently about 70 GWe of connected heat storage (10 GW off peak heating, 60 GW water heating) and this will increasingly come into play as elec replaces gas.

    Slide 73 of this summarises currently (~2005!) available back up capacity with public and private generation.

    Renewable electricity system: Feasibility of a high renewable electricity system
    Barrett, M. 2007, A Renewable Electricity System for the UK. In Renewable Energy and the Grid: The Challenge of Variability, Boyle, G., London: Earthscan. ISBN-13: 978-1-84407-418-1 (hardback).

    Best wishes


    Dr Mark Barrett, Principal RCUK Academic Research Fellow
    Energy Institute, University College London
    Room 227, Wilkins Building, North Cloister
    Gower St, London WC1E 6BT
    Site :

    Skype: MarkAlexBarrett (Mark Barrett)

  3. Reply

    From David Millborrow on the Claverton mailing list:-


    It looks as though I did not place sufficient emphasis on the need to examine several years worth of data in order to derive reliable estimates of capacity credit, see Dale, Milborrow, Slark and Strbac, 2004, Energy Policy, Volume 32, Pages 1949-1956. (I think it is posted on Claverton).

    That paper states, “analysis of a five-year time series of wind speeds…… do not provide any statistical evidence for wind variations at peak being substantially different to those at other demand levels for similar times of the year.”

    In Energy Policy, Volume 35, Pages 112-127, Graham Sinden implies we were perhaps being a bit cautious………. “there is an increased probability of high wind power output…. during periods of high electricity demand; this translates into a capacity factor during peak electricity demand hours that is around one-third higher than the annual average.”

    Sinden’s conclusion ties in nicely with all the other estimates of capacity credit at low wind energy penetrations that were the basis of figure 1 in the my latest paper. (Graham, do correct me if I am misquoting you!)

    Best regards,

    David Millborrow

  4. Reply

    Please note, there is a lengthy and learned discussion thread on hte topic of this article ongoing on the Claverton Mailing list:

    See –

  5. Reply

    Nice site i like it

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these <abbr title="HyperText Markup Language">HTML</abbr> tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>