The Cost of Energy Storage

I taught my students that intermittent renewable electricity (wind and solar) was third class compared with dispatchable fossil fuels (first class) and baseload nuclear power (second class). But that renewables may be turned into a first class electricity source with the development of affordable grid-scale storage. There are two important qualifiers to this statement and those are 1) affordable and 2) grid-scale. By grid scale I mean electricity storage that could power a medium sized town for a day or longer, or every night when the sun is down.

In this short post I want to begin chronicling new storage projects as they are announced for future reference and begin with three very different approaches.

Three Reasons Oncor’s Energy Storage Proposal Is a Game Changer

The first case is Oncor Energy’s plan to install vast amounts of battery storage distributed through the Texas grid.

Early last week, Texas transmission and distribution company Oncor announced a proposal to install 5,000 megawatts of battery energy storage on the Texas grid. The words “game-changing” get thrown around a lot about energy storage projects—usually prematurely. But in this case I think there are some clear reasons why Oncor’s proposed deal could be a game-changing development for grid battery energy storage:


Oncor’s proposal calls for installing thousands of battery systems ranging from the size of a fridge to a dumpster around the state with a combined power capacity of 5,000 megawatts and a combined energy storage capacity of 15,000 megawatt-hours. These numbers sound big, but what do they mean, really?

The vital statistics:

Deliverable load: 5,000 MW
Storage capacity: 15,000 MWh
Cost: $5.2 billion
Normalised costs: $346,667 / MWh or $1,040,000 / MW

The article also says that the Texas grid has a capacity to deliver around 69,000 MW. Assuming average load of 70% translates to daily consumption of 1,159,200 MWh. So $5.2 billion buys 18.6 minutes of storage. Of course that storage can be used every day if the batteries can be-recharged. Reading the article it is evident that this project is more about grid stabilisation than storing electricity from renewables. Perhaps it might have been better to not destabilise the grid in the first place?

Renewable energy plan hinges on huge Utah caverns

The second is for a truly gigantic compressed air energy storage (CAES) facility in a salt dome in Utah to be built by Magnum Energy.

A proposal to export twice as much Wyoming wind power to Los Angeles as the amount of electricity generated by the Hoover Dam includes an engineering feat even more massive than that famous structure: Four chambers, each approaching the size of the Empire State Building, would be carved from an underground salt deposit to hold huge volumes of compressed air.

The caverns in central Utah would serve as a kind of massive battery on a scale never before seen, helping to overcome the fact that — even in Wyoming — wind doesn’t blow all the time.

Air would be pumped into the caverns when power demand is low and wind is high, typically at night. During times of increased demand, the compressed air would be released to drive turbines and feed power to markets in far-away Southern California.


The air would be pumped into four caverns, each 1,300-feet high and 290-feet wide and capable of holding enough air to generate 60,000 megawatt-hours of electricity through turbines at the surface.

The vital statistics

Deliverable load: ? MW
Storage capacity: 60,000 MWh
Cost: $1.5 billion
Normalised costs: $25,000 / MWh

CAES is only a small part of this project that begins with a large wind farm in Wyoming, power lines to the CAES site in Utah and more power lines from Utah to Southern California. Total price tag of $8 billion. I can’t find information quickly on the deliverable load and hope maybe a commenter might help 😉

The Coire Glas pumped storage scheme – a massive but puny beast

The third is Scottish and Southern Energy’s plan to build a huge pumped hydro storage scheme at Coire Glas that I reported on last year.

First new large scale pumped storage scheme to be developed in UK for over 30 years

Consent for the Coire Glas scheme was granted in December 2013 however, despite the obvious benefits that pumped storage offers, making a Final Investment Decision to progress the Coire Glas scheme will require overcoming a number of commercial and regulatory challenges. These include changes in the existing transmission charging regime for pumped storage and a satisfactory and supportive long-term public policy and regulatory framework. Therefore any final investment decision is unlikely before 2015 at the earliest.

The vital statistics

Deliverable load: 600 MW
Storage capacity: 30,000 MWh
Cost: $1.26 billion / £800 million
Normalised costs: $42,000 / MWh or $2,100,000 / MW

In my earlier post I summarised Coire Glas thus:

The idea is to pump water into the reservoir when it is windy. The UK wind carpet recently produced 6GW peak output and so let’s assume that 3 of those 6GW were used to pump water into Coire Glas and other such schemes, and 3GW got fed directly onto the grid. If we are to have a renewables based system that can run independently of fossil fuel back up then it needs the stamina to survive a 7 day lull in the wind. So what we need to know is the amount of storage for 3GW of supply to run continuously for 7 days. This also assumes that we had 7 days producing 6GW of wind beforehand to fill the reservoirs – and we are still light years away from achieving that!

3GW * 24 hours * 7 days = 504 GWh of storage

That is 17 times greater than Coire Glas and 3 GW is only about 5% of UK peak demand. Coire Glas, therefore, is simply window dressing in efforts to “Green” UK power supply with pylons, turbines and dams.


The Utah CAES and Coire Glas pumped hydro schemes have in common vast scale and huge capacity but in the case of the latter, impotent to address the problem it is designed to address. The UK certainly does not have sites to build 17 let alone hundreds of facilities like this. I imagine the same will be true for the Utah CAES scheme. I do not know how many caverns can be safely excavated into this salt dome.

The Texas battery scheme is very different. Utah and Coire Glas are deigned for “stamina” and moderate energy output with time. The Texas battery has a huge amount of muscle but zero stamina. Here’s how the costs compare:

Texas battery: $346,667 / MWh
Utah CAES: $25,000 / MWh
Coire Glas pumped hydro: $42,000 / MWh

I wonder if any will ever be built?

This entry was posted in Energy and tagged , , , , . Bookmark the permalink.

29 Responses to The Cost of Energy Storage

  1. bobski2014 says:

    I’m not an engineer, mathematician or scientist. So, a question. Am I right to think that Coire Glas would represent 5/17 % of UK peak demand – 0.29% ? If so, perhaps it would be a POLITICAL solution to something? It seems a good examplar of UK energy “policy” – in itself something of a misnomer!

    • Euan Mearns says:

      UK peak demand is around 55,000 MW, 6pm, week day in January. So 600/55,000 = 1.1% of peak demand. Storage like this makes a lot of sense for peak management. And it is easy to say “we are storing wind to satisfy peak”. That may get rid of some peaking gas plants – fine!

      But to try and use storage like this to span a long lull in the wind is at present impossible.

      And there is an environmental cost. This scheme is in the Great Glen – Loch Lochy I think. When they are pumping there is risk that river runs dry. And when producing river runs flood. SSE do of course have careful management strategy in place, but water management really does limit how much of this can be deployed along the Great Glen.

  2. Joe Public says:

    Hi Euan.

    As usual, an interesting insight into a ‘peripheral’ aspect of the energy market.

  3. Sam Taylor says:

    Do you have any idea of any of the efficiencies involved? Most of my reading on the subject seems to indicate that storage significantly impacts the EROI of all renewable sources, and probably brings them down to a fairly uneconomic level.

    • Roberto says:

      Yes, you are right, a recent article/study by Weissbach et al published in Energy Policy shows that by considering the need for storage, if the penetration increases, the ‘buffered’ EROI of both PV and wind as well drop to levels lower than the minimum for a technology to be viable.


    • Hugh Sharman says:

      Sam, the efficiency of the Texas “mass roll-out” system will probably be in excess of 80% if based, as I believe it will be, on lithium ion. On the other hand, it may need complete replacement within five years, as the deep cycle life of lithium batteries is of the order 3000.

      The Coir Glas scheme will probably have an efficiency of 80%, which is fairly normal for new pumped hydro.

      The efficiency of the two (only) CAES plants in existence is famously well below 40%. Details of the proposed Utah salt cavern scheme are not easy to find. But the main reason for the dearth of CAES developments since the early 1990s is this simply awful round-trip efficiency.

      I hasten to add that if stochastic power sources continue to be built, there will come a point in their growth, already very close in UK and Ireland (especially) when the lack of electricity storage will force a halt to more wind power being added. Already, I am told on pretty good authority, curtailment of the new turbines now being built is expected to be as high as 8% of potential output.

      I also hasten to add that Euan has provided a good snapshot of a fast developing situation. Electricity storage has some serious players who are already beginning to enter the market while some players, not yet in it, will make it very big indeed.

      The existing electricity systems are filled with over-capacity and wasted resources that can be reduced by realistically priced electricity storage

    • Euan Mearns says:

      Batteries and pumped hydro are very energy efficient – though for batteries, the energy used in manufacture would have to be factored in. I thinkTexas are planning to use lead acid.

      It was Chris Nelder who posted the link to the Utah CAES project and I asked him about the efficiency and he either didn’t know or refused to answer. 40% is a joke. This is a part of the harsh reality of renewables carrying the economic and energy cost of intermittency.

      But you know it doesn’t really matter. All you have to do is build twice as many windmills.

  4. “There are two important qualifiers to this statement and those are 1) affordable and 2) grid-scale.” Appreciate your emphasizing this!

  5. Ed says:

    I’m keeping an eye on organic mega flow batteries. Huge potential.

    • Euan Mearns says:

      So its time to buy rhubarb futures 😉

      • Ed says:

        Ha ha. Interestingly the compound used in the batteries, even though it is almost identical to that found in rhubarb, is being produced from crude oil. Just another example of how everything within the energy sector is totally dependant on fossil fuels for its manufacture, be it renewables, nuclear, batteries or whatever.

  6. If we accept that the purpose of Oncor’s batteries is to make use of intermittent power that would otherwise get “spilt” then we can view them as a generation source with a cost of ~$1,000/kW and a capacity factor of ~10% (3 hours/day times 80% efficiency), This gives an LCOE of about 25c/kWh assuming a 5-year battery life, which isn’t too bad, but I suspect that if the plan goes ahead the $5.2 billion capital cost will be found to be optimistic. Initial capital cost estimates almost always are.

    Another question is who pays the costs. If Texas regulators approve Oncor’s plan my understanding is that Oncor’s consumers will pay whether the batteries deliver the promised benefits or not, but I’m willing to be corrected on this.

  7. Graeme No.3 says:

    There is a problem in that compressing the air generates heat, which has to be removed, and expansion requires heat. Storage methods are not yet in operation. may be of interest.

    Huntsdorf compressed storage in Germany has been in operation for grid stabilisation for about 20 years, but note that there are many more pumped water schemes.

    Another problem in Germany is that many of the pumped storage schemes aren’t making money because solar power on sunny days cuts in during peak demand, and reduces their revenue.

  8. Luís says:

    Hi Euan, I think what you called “normalised costs” are not really normalised (or levelised) because you don’t take lifetime into account. The difference between hydro and LiON should be much wider.

    These are some of the technologies I am following:

    EOs zinc-air
    Cost: 160 $/kWh
    Lifetime: 30 years
    In commercial grade testing since 2013; presently in use by GDF Suez. Full commercial production expected for 2015.

    Aquion sodiu-ion
    Cost: 250 $/kWh
    Lifetime: 10 years
    Several commercial trials on-going; accepting limited commercial orders since late 2013. Siemens has acquired an unknown number of these batteries to integrate in micro-grids.

    • Olav says:

      Coire Glas pumped hydro: $42,000 / MWh
      Deliverable load: 600 MW
      Storage capacity: 30,000 MWh
      Cost: $1.26 billion / £800 million
      Normalised costs: $42,000 / MWh or $2,100,000 / MW or $ 2100/ KW Storage loss zero as rainfall is equal or larger than evaporation

      My sun charged thermal storage for cooking
      Deliverable load: 0,0025 MW Thermal at full charge. Boils 1 l water in 200 sek. Storage capacity: 0,003 MWh Thermal, Cost $ 400, Normalized cost $ 133 / KW Thermal. Storage loss 0,4% an hour. Still able to boil potatoes 48 h after full charge. (took then 90 minutes). S 133 a KWh for one cycle = 1 day. As storage lasts 50+ years equ 18300++ cycles then storage cost is 133/18300 = 0,0073– $ a KWh

      • Euan Mearns says:

        Olav, I’m intrigued to learn a bit more “sun charged thermal storage” system.

        • Olav says:

          Hi Euan
          I may see a business case in this as the goal is being able to provide cooking need for a family for $1000 material cost (collector and storage)
          Durability is very long. A microfinance over 5 years should make it cheaper than unsubsidized LPG, after that cooking is free for many years.
          Collecting biomass for cooking is free but avability may be scarce and then it has its smoke issues.
          I will for some time forward keep the details for myself, Have to make some tracking to better test it in Western Norway in Winter.
          About storage loss above it is 4% not 0.4%. 0.4% loss only applies when storage temp is at low end (150 C) just being able to get water to boil.
          Actually you do not need 48 h storage capability. Making a hot breakfast from yesterday sun is a good beginning. Backup is eat cold or use the old solution.

          Besides this which I believe has a game change possibility, I have 3 solar water solutions.
          One purchased system for heating floor in an apartment. Works year round but controllers has failed 3 times. It works year round but solar is insufficient in winter so a
          3 KW electric heater is kicking in if water in tank goes below 30 C

          Another solution I made myself is routing cold tap water thru 80m (equ 80l) black plastic tube. It spiraled on 2 hinged panels with insulated back side and poly plate as glazing above. It is facing due West
          Max temperature in summer approaches 90 C and without use it still holds 60 C at 23 in evening. When hot water is used is hot water coming into the tank. Hot water use
          from noon to midnight is preferred. In the morning is water in tube 25 C
          System is under normal water pressure so pumps and controlled are not needed. I use a compressor to empty the system in late October.
          Has been working for 3 seasons now just have to move 2 valves and connect 2 hoses late March.

          I have a small digged down 8000 litre (3m x 1.7m x 1.4m) concrete inside tiled pool on the outside deck. It has 5 cm insulation on bottom and 10cm on sides. Top is covered with strong poly plate
          for security and insulation. Got it finished in late August. Have a pump that takes water from bottom through 400m 20mm black hose (similar solution as above) and then thru filter.
          The Pump takes 250W when sun shines. Takes (1000 l/h) water from bottom thru the system end emptied at top pool 5 Deg hotter. By Mid September after a week with good weather it reached 35 C in pool. I have to
          run pump more seldom in summer.

          But the solar water is not so special the cooking solution is the big thing…

          • Euan Mearns says:

            Hi Olav, I have designed a solar clothes drier that I hope to retail for $500. I may post some details later. Genuinely interested in these solar gathering / hot water stores. But always left perplexed as to why they are so difficult / expensive to install.

          • A C Osborn says:

            Roger, based on my simple experiments I doubt if those troughs gather as much energy as a focused Parbolic mirror, but are a decent compromise.

          • They have parabolic mirrors too. (Maricopa Solar, Arizona)

        • A C Osborn says:

          Euan, a 4 inch Parabolic Torch reflector can heat a 1/4″ piece of steel 1″ long (Allen Bolt) where the bulb would normally be to a temperature of 450 C in a few minutes at mid day in Wales.
          I see no reason why this could not be used to heat a copper pipe with water flowing through it.
          If a 4 inch mirror can do this imagine what a 12″ or 16″ Search Light mirror could do.
          The one major snag is that it has to track the sun, so a solar powered tracking motor would be needed.
          Once focus of the sun is lost (DWIR only) the heating disappears completely.
          This is obviously the basis of a Solar Still or oven.
          I like Olav’s idea of the black panel to warm water, just place a black refuse sack in the sun to see how hot it gets.
          This could be used as a method of pre-heating the water for his other methods.
          They used to advertise Solar Heating (not electicity) quite a bit in the past.

          • AC: This is how the Andasol CSP plant in Spain works. The absorber pipes the heliostats focus on contain a synthetic oil that gets heated to 400C, hot enough to generate steam to drive a turbine.

          • Graeme No.3 says:

            Some years ago a University (Sydney, I think) was working on a non tracking solar concentrator. Unfortunately I cannot find any reference.
            The cross section of the trough was semi-spiral in shape, almost like a spiral ammonite shell for the first 300 degrees, if you know what I am trying to mean. It focussed all incoming rays onto an off-set evacuated tube and achieved highish temperatures with oil pumped through ~ 250℃.
            The point being that once in place it did not have to be moved, even for seasonal solar altitude variations. May have succumbed to dust problems.

            Or there is the compound parabolic type see

  9. Luís says:

    Just another thought: actually, the figure of 350 $/kWh for the Texas system seems too low to be LiON. Is there any direct reference to the technology they will use?

    • Euan Mearns says:

      Luis, thanks for links. The costs of systems obviously need to be adjusted for a number of variables. The CAES system looks like it will waste 60% of the input energy – I guess there will not me much of a market for low grade heat in the Utah desert – they should build a hot water pipeline to Seattle 😉

      I re-read the Texas battery article and can’t see reference to what they will use, but I seem to recall reading somewhere it was Pb acid. It is also evident that Oncor as a “wires” company are not allowed to generate. So this has as much to do with them wanting to muscle in on the generators patch as anything else.

  10. ed says:

    Olav. You should set up a blog !! Your home made energy experimentations/solutions are really interesting.

    • Olav says:

      Can’t make a blog but here is the least costly way of saving several KWh a day.
      A bathroom may have electric cables in floor. I belive they are rated to 1200W
      They probably draw 300W in average year round. That adds up to more than 2000 KWh a year

      There are several costly and maitainance burdensome solution to use warm water drain in
      shower to heat incomming water. I was checking the avability of those solutions and was recommended not to do it because they easily clogged up and the cleaning job after was…..

      I prefere the bath tube rather than the shower so then…..
      A really good bath uses 100 l water and the difference between 40C and 20C equ 2,3 KWh which goes straight down the drain when the plug was pulled,what a waste of energy….
      Leving the water there without any precautions till next day is not a recommended as
      steam from the water fills the room and that is not good.

      But: The edges on a bath tube is exactly at same level all the way around.
      I took a left over peace of polylate and put it over the bath tube and then…
      No steam at all and the water was still at 20 deg the next day when I pulled
      the plug. The flushing then was still easy
      The effect is as to have a 100 W heater on contineously.
      Have not switched the electric cables on yet. But The snag is cold floor and low WAF.
      The house has another bath room with shower so I was permitted to do so
      Anyway solutions without sideeffects are seldom effective

      WAF = Wife Acceping Factor

      • Raff says:

        Have you tried a rug or carpet? Or Slippers?

        • Olav says:

          No: That will not work. You need the stiffness and It will be soaked and moisture enters the room. A plywood sheet with plastic fastened on underside will work but it is heavier.
          Best is Polycarbonate 3 mm thick from a green house supplier.
          I used 6mm thick and the insulation is too good. Getting the heat out of the water and into the room in less than 24h is preferable

Comments are closed.