Skip to content

Can Batteries Solve The Intermittency Problem?

May 29, 2018


By Paul Homewood



The wind turbines outside Stirling Castle (Getty)

There seems to have been a certain amount of interest in adding battery storage systems to wind and solar farms recently. Needless to say, the owners are not attempting to cure the problem of intermittency. Instead there appear to be two main reasons:

1) Storage systems are eligible for subsidy under the Capacity Market and also the National Grid’s Supplemental Balancing Reserve, both designed to buy in standby capacity.

2) Power produced during times of surplus, say on windy or sunny days, would tend to have a lower economic value. By using it to recharge batteries, the stored power could be sold at a higher price at times of shortage or high demand.


Hywind Scotland, the floating offshore wind farm is a good example of the latter, with plans to add 1 MW of storage.

We often hear about the capacity of battery systems, but rarely how long they can supply this power. In other words, the amount of energy stored.

Hywind tell us that their 1 MW will supply 1.3 MWh, meaning it can only run at full power for just over an hour.

Elon Musk’s heavily hyped Tesla system in South Australia has a similar ration, 100 MW and 129 MWh.

These sort of systems may be OK for balancing small, short term spikes in the grid, but are totally useless for dealing with the problem of intermittency. So, just how much storage would we need for that?


I have analysed the official wind power in the UK for February 2018. As you can see, it is highly variable, ranging from 369 MW to 10313 MW. The theoretical installed capacity is 18100 MW, and the average in the month was 5300 MW.




If there was no other form of backup power, and we had to rely solely on batteries, how much storage would we need to continually supply that average of 5300 MW?

There were two periods during February, when wind output was seriously reduced for lengthy spells. During the second spell, output was below average from 16th to 23rd February, averaging 2870 MW, just over half the average.

To supply this shortfall, a battery storage system would need to have stored 405 GWh, which is equivalent to about three days of average output.

To put this figure into perspective, Musk’s Australian battery stores 129 MWh, so we would need 3140 of them. At an estimated £125 million each, we would be looking at a cost of £392 bn, and they would all likely need replacing every ten years.

And all this just to back up wind farms that only supplies about 15% of the UK’s electricity.


I have only looked at one month, and it may well be that even more storage would be needed at other times of year.


To put the figures into perspective, Hywind Scotland, used as an example above, has installed wind turbine capacity of 30 MW, probably giving an average of about 12 MW.

Instead of the 1.3 MWh battery it is installing, Hywind would need 665 of them, to be able to ensure continuous average output.

If all wind farms were legally obliged to install three days of storage, there would not be many left!


This exercise shows above all why the dream of running the country on 100% renewable energy is just that, an impossible dream.

  1. Ben Vorlich permalink
    May 29, 2018 6:07 pm

    That picture of Stirling Castle and the Braes of Doune eyesore on the hills behind it causes my heart to sink every time I see it. Before my mother died that view always raised my spirits as I knew I was nearly back at my childhood home.

    The knowledge that the energy produced by said eyesore is expensive and unreliable is just salt in the wounds.

    Trying to explain to disciples of so called renewable energy all the problems outline in this article is met by a “does not compute” stare.

    • Sheri permalink
      May 29, 2018 11:23 pm

      That was my first thought—what a horrible background for such a beautiful castle. Wind developers care nothing about the planet, history or beauty. They are greed incarnate. I should have known a subsidy was involved in the batteries. None of these vultures does anything without appropriating someone else’s money.

  2. AZ1971 permalink
    May 29, 2018 6:41 pm

    Math—the bane of alarmists pushing 100% renewable (or even some portion less than that.)

  3. mikewaite permalink
    May 29, 2018 6:52 pm

    Please tell me that the picture of Stirling Castle is not true , but a photoshop of what could happen. Surely no true Scot would even consider polluting the view from the National Wallace Monument . Having visited Stirling in happier days it saddens me , and I am English..

    • Ben Vorlich permalink
      May 30, 2018 7:05 am

      It has a telephoto lens effect but Braes of Doune is clearly visible in the background view of the castle from many approaches to Stirling from Glasgow and Edinburgh. Fortunately last time I visited there were no turbines in the view from the Dunblane Bridge of Allan stretch of road where the castle stands out from the Carse.

  4. Gamecock permalink
    May 29, 2018 6:58 pm

    It’s worse than we thought!

    ‘If all wind farms were legally obliged to install three days of storage, there would not be many left!’

    You’re not comprehending. You would also need the GENERATION CAPACITY to charge them! Not just infinite money for the batteries, but another infinite (!) more for the additional production.

    You can only store excess.

    Wind/solar is crap for electricity production. Why would they be good for generating backup power for that which needs backup? It is intellectually vacant.

    • AngryScot permalink
      May 31, 2018 6:45 am

      The only use for them IMHO is to power reverse osmosis plants to convert sea- to freshwater. Then we could dismantle all of the revolting things and transport them to where they are needed most!

  5. Gamecock permalink
    May 29, 2018 6:59 pm

    Can Batteries Solve The Intermittency Problem?

    Betteridge’s Law applies (No!).

    • AngryScot permalink
      May 31, 2018 6:45 am

      No but Thorium can…

      • Gamecock permalink
        June 1, 2018 1:38 pm

        Thorium can what? Trick people into bringing back a 50 year old conspiracy theory?

  6. Ian permalink
    May 29, 2018 7:00 pm

    This link’s about a site or sites in Northern Ireland. Thought it could do with some exposure: (I think I got it from the Euan Mearns site).

    • May 29, 2018 7:13 pm

      Thanks Ian.

      Are you referring to its banning?

      I noticed it yesterday, and the site, which is useful and I belive run by a renewable enthusiast, applies to the whole Uk

      • Ian permalink
        May 29, 2018 7:33 pm

        Yes Paul. A case of “they don’t like it up ’em”?

      • May 29, 2018 7:41 pm

        I have used the information at Variable Pitch for many years and have frequently been in touch with the site’s owner. As far as I know, all the data at Variable Pitch is publicly available (e.g. from Ofgem, the relevant local authority etc), so I doubt that the Irish trougher can stop his data from being published.

  7. Joe Public permalink
    May 29, 2018 7:01 pm

    Battery systems are ‘sold’ to the public as ‘storage, but for many (most?) the primary function is actually voltage control / grid stabilisation – simply because there’s now too much volatile intermittents! (Public attitudes to intermittents / batteries would undoubtebly be different if these relevant facts were known to them – facts the usual suspects within the MSM deliberately mislead-by-omission on)

    The Tesla / Hornsdale ‘Big South Australian Battery’ (AKA BSAB) is a prime example.

    Every man and his dog reported it is “100 MW and 129 MWh” (no offence Paul), but of the 100MW/129MWh battery capacity, around 70MW of capacity is contracted to the South Australian government to provide grid stability and system security. Only the remaining 30MW of capacity will provide three hours storage availability.

    It’ll enable its owners to load-shift energy from their adjacent Hornsdale wind farm, and in turn will allow them both to avoid potential curtailment enforcement and also to take advantage of high peak prices in the electricity market.

  8. May 29, 2018 7:22 pm

    Batteries are better than nothing perhaps, but only at the margins e.g. taking a bit of the strain at peak times, or when some part of the grid drops out for any reason.

  9. Turbotom permalink
    May 29, 2018 7:23 pm

    Could someone with greater knowledge of these things than me explain how Battery storage systems are synchronised to the grid? I assume the batteries deliver dc current which needs to converted to ac and then synchronised to the grid frequency – do they therefore rely on the buffer of a stable grid to be able to do this?

  10. May 29, 2018 7:28 pm

    Batteries can be used to help fill the intermittency gap by operating as reserve or can be used to support frequency stability by acting as response. In both roles the batteries are pretty useless and are also high emitters of CO2 (but you won’t find the developers saying that – according to their lies and propaganda, batteries are “low carbon”.

    A good explanation of the role of batteries (and DGs and gas generators) acting in reserve or response can be found in the booklet produced by CPRE Wiltshire. It can be downloaded at

    • Jack Broughton permalink
      May 29, 2018 8:12 pm

      Thanks for this link, a nice straight forward report.
      It had not occurred to me that the charging cycle is taking power from the grid’s power-mix, rather than just from the unreliable generator, as is always presented in the meja. So, storage has a significant CO2 emission penalty (compared with using the live-power).

      This is a bit like the wood burning CO2 not being counted as it might be “renewed” eventually, possibly.

      Rather than just making a killing from the totally unfair subsidies, they will be able to sell the power at peak demand times at even bigger profit margin: even beats curtailment charges!

      • It doesn't add up... permalink
        June 1, 2018 1:39 pm

        It also has an efficiency penalty. The South Australian battery eatsabout 20% of its input on losses.

  11. May 29, 2018 7:38 pm

    A few more of these and there wouldn’t be an ‘intermittency problem’…

    SSE is to build a new $467m combined cycle gas turbine (CCGT) power plant in Lincolnshire

  12. mikeo28 permalink
    May 29, 2018 8:36 pm

    I am in Australia and have the capability plus the data to analyse what happens on our grid. I am concentrating on 2017 for my analysis. We have about 40,000 km of grid with something like 400 electricity generators connected to it. Despite the wild statements in the mainstream press the fact is about 5.8% of our energy comes from wind. We have a plate capacity of 4679 MW of wind generators dispatching energy to the grid. They geographically occupy an area of 2,250,000 km² and physically about 2000 km². Our present government is hellbent on building a very large pumped storage facility. It is to be 350 GW hours if it is ever built. I have modelled the idea that using it you could get steady power from our existing wind installation. I did this by hypothetically dedicating the wind output to the pumped storage and then only drawing energy from it. I have accounted for a 30% loss for recharge pumping and if you reach a point the dams are full there is nothing you can do with excess electricity. Currently our wind installation varies from -4 kW to 3656 kW over 2017 and dispatch about 11.5 TW hours to the grid. This is 27.7% of the plate capacity. So what happens if I apply my model in order to get a constant power? There is a significant reduction in output you can only run such an arrangement at a capacity factor of 16.5%. The storage gradually declines from the beginning of the year when hypothetically it is fully charged until it reaches nearly zero about the middle of the year which are our colder months. There is a distinct pattern that we get less energy from wind during times when it is cold. From that point this model shows recovery returning to fully charged near the end of the year. The optimal appears to be 450 GW hours from which a capacity factor of 17.5% could be achieved. It is not possible once you introduce energy storage to do better than that because of the losses previously explained. I have been working for some time to present this in a digestible form to the public. There are 105,120 five-minute data points in a year even if I aggregate the entire wind installation into one. The cost is at least $15 billion for something that can be replaced by a generator of less than 2 GW.

  13. swan101 permalink
    May 29, 2018 9:27 pm

    Reblogged this on UPPER SONACHAN WIND FARM and commented:
    A large dose of commonsense needing to be widely circulated…..

  14. markl permalink
    May 29, 2018 9:51 pm

    More proof that the green blob doesn’t understand energy needs and associated issues.

  15. May 29, 2018 9:57 pm

    I agree that temporary energy storage is a poor substitute for dispatchable power when energy intermittency of days or weeks are involved. But there are circumstances where demand is out of sink with generation where several hours of storage can make a lot of sense. I know that the Tennessee Valley Authority uses pumped hydro to store power from nuclear plants at night when demand is low and release that water through turbines during the day to generate power when demand is high. I could see this also work in a state like California where demand is highest in the hot, dry summer months when solar generation is highest but peak demand is in the evening while peak generation is in the middle of the afternoon. Pumped hydro in that circumstance could even become part of the water delivery system where it’s moved over mountain ranges anyway, it just takes a bit more elevated storage than what is already there to level out demand with generation. If storage is needed on a much larger scale for longer periods, more traditional hydroelectric could provide a much larger buffer but environmentalists are as fond of hydroelectric as they are of nuclear.

  16. Tristan Dodson permalink
    May 29, 2018 10:13 pm

    Grid scale Li-ion batteries aren’t used to ensure wind generation gives continuous supply. That would be pointless because electricity demand isn’t constant. They are used to provide frequency response i.e. very quickly respond to sudden changes in demand, while slower response generators or sources of demand can be ramped up or down accordingly. So essentially you have successfully calculated that something ridiculous is indeed ridiculous. Well done.

    • May 30, 2018 5:38 am

      Did anyone claim that a 1MW battery is grid scale? You are wrong in claiming that these batteries are not used for reserve, but only for response. Developers do use them for reserve, to fill the intermittency gap. At least we agree that it is ridiculous, but then in the UK we have a ridiculous energy policy.

      • Tristan Dodson permalink
        May 30, 2018 9:07 am

        The entire purpose of this post is to calculate the size of the battery needed to smooth out all of the UK’s wind generation. This clearly is ‘grid scale’. But the basis for this is that these batteries are used only for reserve. This is not true. Their core revenue stream is providing grid services such as frequency response.

        While it may be interesting to know how much battery capacity is needed to provide constant power supply, it holds no value and cannot be used as an argument against battery storage.

    • May 30, 2018 8:59 am


      Perhaps you had better tell the BBC then!

      Or maybe Claire Perry:

      • May 30, 2018 9:55 am

        I have examined lots of applications for “battery barns”. All of them quote statements from the Government, Ofgem and National Grid explaining the need for storing energy when there is too much wind or solar power and reinjecting it into the grid when demand is high and generation from wind or solar is low. In other words the purpose of the batteries is reserve, not response, as explained in the CPRE Wiltshire report I linked to above.

        I have not seen a single application for batteries to provide response.

    • Joe Public permalink
      May 30, 2018 9:23 am

      Tristan – see my comment of May 29, 2018 7:01 pm, above.

      Batteries are ‘sold’ to the public for their storage capability; many of the public would be aghast that they’d be unnecessary if it wasn’t for the high penetration of intermittents.

  17. May 30, 2018 12:38 am

    Interesting article!

    I made a kind of same quick analyze for the Netherlands last year. In 2015 we had a installed capacity of wind and solar of 4.906 MW. The average power production thru the year was 918 MW. To use these 918 MW yearly round we would need a storage capacity of roughly 1000 GWh, that’s about 76 million of Tesla Powerwalls!

    Then I scaled it up to run fully on wind and solar (if the demand would be the same, what is hardly to believe because of the electrification of transport, heating etc). Then we would need a storage capacity of almost 9000 GWh, 637 million Tesla Powerwalls.

    For al those complainers who say that this is stupid to calculated or pointless because batteries aren’t designed for that, Yes I know, but it’s to show how intermitted and unreliable the wind and solar are and the big scale/size of any kind of storage is needed!

    The most effective way to store the energy is by the use of pumped hydro, but to give some indication of what is theoretical possible in Europe for storage as in pumped hydro is 6924GWh! This is not even enough for only a small country like the Netherlands. And this is what is in theory is possible, but absolute not likely that is ever will be build!

    And although I wrote it in Dutch, the numbers are international 🙂

    • It doesn't add up... permalink
      June 1, 2018 1:45 pm

      The Netherlands has zero pumped storage potential. It’s flat!

      The UK is little better. Our biggest site Dinorwig can store just 9 GWh. About 1% of daily demand.

  18. pochas94 permalink
    May 31, 2018 2:22 am

    There’s no stopping people whose mission is to save the world.

  19. Robin Guenier permalink
    May 31, 2018 7:38 am

    During this morning’s rush hour (in the UK), wind and solar power were contributing only 5.5% to the National Grid demand – an example of the intermittency problem. But CCGT, nuclear, biomass and the French IC “solved” it by contributing 92%.

Comments are closed.

%d bloggers like this: