The Future Of Energy Storage On Both Sides Of The Meter

Noor III Solar Tower of the Ouarzazate Power Station, at dusk. (Credit: Marc Lacoste)

That energy storage is a hot topic is hardly a surprise to anyone these days. Even so, energy storage can take a lot of different forms, some of which are more relevant to the utility provider (like grid-level storage), while others are relevant to business and home owners (e.g. whole-house storage), and yet other technologies live in this tense zone between utility and personal interest, such as (electric) vehicle-to-grid.

For utilities a lot of noise is being made about shiny new technologies, such as hydrogen-based storage, while home- and business owners are pondering on the benefits of relying solely on the utility’s generosity with feed-in tariffs, versus charging a big battery from the solar panels on the roof and using the produced power themselves. Ultimately the questions here are which technologies will indeed live up to their promises, and which a home owner may want to invest in.

Dealing With A Changing Grid

Bath County Pumped Storage Station (Credit: CHA)
Bath County Pumped Storage Station (Credit: CHA)

For many decades the electrical grid has been a relatively straightforward system with dispatchable generators and consumers, with the former producing when required and the latter consuming and a bit of on-grid storage to smooth over the bumps and valleys. The biggest change that has been occurring over the past decades has been the introduction of electrical generators that only produce electricity when weather and other conditions are just right.

The result of this has been that the peaks got a lot taller and the valleys a lot deeper. While curtailment of unnecessary (excess) power would be a potential solution, a popular idea is to use as much of this excess power as possible at a later point by storing it somehow. At this point in time, most grid-level energy storage is provided by compressed air energy storage (CAES) and pumped hydro storage (PHS). What these two technologies have in common is that they both have a relatively low energy density by volume, but make up by that through having a lot of volume.

In the case of the second-largest PHS system, the Bath County Pumped Storage Station in Virginia, its upper reservoir has a capacity of 43,000,000 m3 (35,599 acre-ft), which provides it with enough gravity potential to provide power for 11 hours at around 3 GW, for a total of 24 GWh. With a round-trip efficiency of 79%, it provides important storage for PJM Interconnection, to buffer energy and take the stress off transmission interconnects between parts of the grid.

For CAES there are not nearly as many installations as there are of PHS installations, mostly due to the complexity of finding subterranean caverns with the right (airtight) properties. Because of this, there are only two CAES installations in operation in the world today: the McIntosh, Alabama (USA) and Huntorf, Elsfleth (Germany) CAES plants. The former has a capacity of 2,860 MWh, the latter 870 MWh.

As a result of the diabatic process used with these CAES plants, the compression process creates waste heat, and subsequent expansion requires input of heat. In the case of these existing plants, natural gas plants are used for this, which in the case of the McIntosh CAES plant results in an overall system efficiency of 27%, up to over 40% when combined with energy recovery mechanisms.

For PHS the number of potential new sites where such storage sites could be economically constructed aren’t too many, which has led to the focus on new technologies that could provide PHS-like storage, without the logistical and situational complications.

Hydrogen Storage

The so-called ‘hydrogen economy‘ has its roots in the 1970s, when the term was coined by John Bockris during a talk at General Motors. The main driving idea behind it is that hydrogen can be used to decarbonize many aspects of industrial processes, as well as to create ammonia (NH3), which forms an essential element of fertilizer. It’s also been suggested to use ammonia as an intermediate form before converting back to hydrogen.

At this point in time natural gas (NG, mostly fossil methane) provides most of the hydrogen and ammonia used. As a result, the use of excess electricity to create hydrogen via electrolysis has been suggested as a source instead. None of this has been developed into large-scale projects yet, however. This comes alongside trials to mix hydrogen into NG, up to a certain percentage. Large amounts of hydrogen in NG infrastructure can lead to issues such as metal embrittlement, due to the hydrogen diffusing through pipe walls.

To store hydrogen, for later use (e.g. time-shifting large amounts of energy), currently mature technologies are compression and liquefying. Complications here are the immediate energy loss from electrolysis (~20-30%), the losses of compression or cooling down the hydrogen gas, the leakage from storage containers, and if conversion back to electricity is desired, the 40-60% efficiency of a hydrogen fuel cell (HFC).

Everything taken together, the round-trip efficiency of a hydrogen-based storage system for the grid would be between 18 – 46%, according to studies. This would be a major drop in efficiency compared to PHS and CAES systems, while adding more complexity and the potential hazards of handling cryogenic liquids and highly inflammable gases. This makes such a system a very hard sell. When the hydrogen is to be used immediately, the question is also whether it cannot be produced more efficiently through e.g. radiolysis.

Thermal Energy Storage

Overview of the thermal energy transfer in the Natrium reactor design. (Source: TerraPower)

Comparatively, thermal energy storage (TES) is significantly more efficient and straightforward. It’s seeing significant use in the form of geothermal heat pumps which can cool and heat buildings rather efficiently. In an industrial setting, molten salt storage is a common type of sensible heat storage, the concept of which is behind storage heaters, which many private homes use.

With concentrated solar power (CSP) plants, molten salt is often used to store the heat from the solar rays, after which the heated salt can be used to generate steam for use with a conventional steam turbine generator. This is also the operating principle behind the TerraPower Natrium Generation IV reactor, which uses the heat from nuclear fission to heat the salt. Because of the slow heat loss, and high efficiency of such a heat transfer system, it can be used to generate electricity or heat, even after storing the molten salt and not using its thermal energy for a week.

A 400 MWh capacity system would need a tank of about 9 m tall and 24 m in diameter, using nothing but conventional materials that would be safe to handle (when not hot). Compared to the complexities of PHS, CAES and especially hydrogen-based systems, TES could end up playing a big role for energy storage. Perhaps unsurprisingly, the Ouarzazate CSP plant in Morocco is the largest energy storage plant on account of its molten salt storage at 3,005 MWh.

Keeping Things Simple

Even if PHS and CAES capacity is unlikely to be expanded significantly any more, there would seem to be still a number of technologies that utility providers can look at to expand storage capacity, even without resorting to electrochemical batteries. This also gives some hints as to what might make sense for private home and business owners.

Since on-grid storage is unlikely to see massive expansion, there is little incentive for grid operators to motivate the feed-in of intermittent power from solar and wind power. When there is the option of using locally installed PV solar panels as well solar water heating systems (sometimes combined in the same panel) to charge up a battery installed inside the same building and heat up water, the costs of this system are highly predictable, and compensated by the electricity (and natural gas) not used from the utility provider.

With the many uncertainties in the energy market and the current world economy it’s hard to say what the coming years will bring, but sticking with proven systems, and aiming for local consumption for small producers might be just the ticket.

[Heading image: Noor III Solar Tower of the Ouarzazate Power Station, at dusk. (Credit: Marc Lacoste) ]

81 thoughts on “The Future Of Energy Storage On Both Sides Of The Meter

  1. “The result of this has been that the peaks got a lot taller and the valleys a lot deeper.”

    Let’s have look at what the situation actually looks like.

    >In a new modeling study in the journal Patterns, Lall and Columbia Ph.D. student Yash Amonkar show that solar and wind potential vary widely over days and weeks, not to mention months to years. They focused on Texas

    >Drawing on 70 years of historic wind and solar-power data, the researchers built an AI model to predict the probability of a network-scale “drought,” when daily production of renewables fell below a target threshold. Under a threshold set at the 30th percentile, when roughly a third of all days are low-production days, the researchers found that Texas could face a daily energy drought for up to four months straight. (…) Batteries would be unable to compensate for a drought of this length

    >”In a fully renewable world, we would need to develop nuclear fuel or hydrogen fuel, or carbon recycling, or add much more capacity for generating renewables, if we want to avoid burning fossil fuels,”

    1. How much energy and cost does it take to prepare and maintain storage for spent nuclear fuel? What is a practical way to store spent fuel?

      These questions has been asked over and over for decades and there is still no answer. At first we were told that nuclear energy would be “too cheap to meter” and now we are getting surcharges on our electric bills. Whatever answer you get from the nuclear advocates is a lie. They have been lying to us continuously for decades, why should they stop?

      1. That was one guy in the 50’s who made a prediction in a press conference that “It is not too much to expect that our children will enjoy in their homes electrical energy too cheap to meter”. Anti-nuclear propagandists have been clinging on to that statement as if it was the official promise of the Atomic Energy Commission – in other words, a big strawman.

        Deep borehole disposal is the cheapest and safest option, but blocked by politics and protests.

        >Whatever answer you get from the nuclear advocates is a lie.

        If you refuse all answers, then you obviously win, right?

        1. Ironically, it also became true in places.

          I have lived in apartments where the electricity was billed by a fixed rate and not metered, because it would have cost more to send someone out to read the meter. The electric bill was simply rolled in the rent, and the meter was read only if they suspected foul play, like someone boiling moonshine in their bathroom with the free power.

          1. Installing a new meter or upgrading service amperage requires the whole structure be brought up to modern code.

            Meters are read by computer, for a good long time now. But rewiring a building is a huge expense.

            They might have to rewire to put each apartment on separate circuits anyhow.

            Then there is the guy who finds out the hall lights are on his circuit…and makes a federal case of it.

            Always turn off your breakers and see if any common area circuits are affected. You never know when the landlord will decide it’s ‘fine’ to run the common laundry room off one apartment.

          1. Really doesn’t matter how big they are, its still peanuts compared to the world as a whole, or the area that can practically be linked to a common power network to cover any outage…

            Also Hail won’t break solar panels easily, there is a reason solar used for outdoor PV arrays are thick and heavy… They can take a beating from the elements unharmed (not saying hail can never damage them, but they are built to take it, so it would have to be very very impressive hail to really break one, and rather hefty to cause any harm that leads to early but not instant failures). And in similar vein wind turbines in storm setup – with the prop feathered and locked are supposed to be able to take stupendous windspeeds unharmed, so most of them in the path of a hurricane would be fine, as its only that narrow centre track that is actually possessing wind powerful enough to perhaps break them, or throw something heavy into them – still a big enough area but again peanuts compared to the full extent of the vortex…

            Also your core paper assumes using energy exactly as we do now and throwing electric stores at any deficit, which is generally considered a backwards way to think of renewables, it just isn’t the way to get the best of them, making life more efficient (and as you note at the end adding vaster renewable supply) change the dynamic hugely, build and run the system in the way its good for, which at worst is minor changes to how folks live and work and there isn’t ‘Energy Droughts’ significant enough to matter in the same way your drinking water supply can run a little short, hosepipe bans every so often perhaps, but assuming a some investment in maintaining the system and a little competence in the engineers and forecasters its a tiny blip that makes not much difference…

          2. >or the area that can practically be linked to a common power network to cover any outage…

            I don’t think we will agree on “practical”, and besides the distributed approach is actually more fragile due to its interdependence, and leads to even bigger disasters when it inevitably breaks.

            >which is generally considered a backwards way to think of renewables

            The “forwards” way is rationing energy when there’s not enough production and curtailing output when there’s too much in the lack of adequate storage capacity, which is basically operating in a fail state and pretending it’s fine.

          3. >which at worst is minor changes to how folks live and work

            These minor changes imply huge political and social changes to implement, which has certain side effects.

            In the 70’s during the oil crisis, you could buy gasoline for your car every other day depending on whether your license plate was odd or even. That was a temporary emergency measure that required the state to control commerce. When such emergency measures become the normal state of society, it implies that the state is now authorized to control commerce to that extent indefinitely…

            …which is a big reason why certain parties would absolutely prefer to normalize such behavior.

          4. Everything is already ‘rationed’ by cost as it stands anyway! Gas prices are up this month, but we know its temporary pain because of x so we won’t run our foundry at all till its resolved, as lots of overtime to clear the work is still cheaper than the energy cost. Or electric is cheaper at night so we do all our energy intensive operations on the cheap rate – NOTHING NEW at all in shifting use patterns to supply…

            And that is all that is really needed, relative minor shifting of some loads to when there is surplus cheap energy, ideally with improvements in efficiency in general – like actually insulating your homes, not using 4000W halogen bulb when the 40W LED/CF bulb can do the job just fine, not leaving every device ‘on’ in the office, better standby power consumption etc…

          5. >Everything is already ‘rationed’ by cost as it stands anyway!

            If you fail to see the fundamental difference between market prices and state economic planning through rationing, it’s no wonder why you don’t get the point.

          6. My point is you don’t need a state imposed rationing, the people and business ration themselves based on the cost, exactly as happens now. NOTHING NEED change in a mostly renewable grid. Plus if you as the state want to get to proper state imposed rationing you can do so no matter where your energy is sourced… At least till the revolution comes because you are being arseholes keeping the people down for no good reason, and if there is a good reason people will just carry on as best they can.

          7. Look, the point is that substituting demand side flexibility for supply side reliability doesn’t actually solve the problem. It just switches one issue for another.

            If you let the suppliers get away with little or no reserves – no reliability – then people do indeed start rationing their own energy use because they very often can’t afford it. Since not all demand is flexible, the average power cost for consumers goes up while the quality of service goes down – and they have to buy the storage technologies without the economies of scale, which also costs more. The result is called energy poverty.

            Here we’re talking about shortfalls lasting weeks. Your average storage boiler runs out of heat in a day or two. The part where the state has to start rationing energy is when the rich people buy all the available energy off the market to heat their boilers, and the poor people are left showering with cold water.

          8. Folks ration not because ‘energy costs too much to buy’ but because they can make their money go further, make more profit, live better by not buying lots of it while its expensive – in the same way every time rumors of a fuel cost increase circulate everyone goes and fills up the car right now, even if they have lots of fuel because they expect the next time to cost them more – its not that they needed it or couldn’t afford it after the usually pretty small increase, just that it gets them one more beer at the pub on friday or whatever – that little extra money to spend on something they want or a little more saved for future needs…

          9. >not because ‘energy costs too much to buy’ but because they can make their money go further

            That’s the same thing in different words. “Too much” doesn’t mean you’d go absolutely broke. If you want cold beer, but the electricity to run the fridge eats into your beer budget, that is saying the electricity costs too much.

          10. Besides, since energy is fundamentally used to run the entire economy, not only do you pay more directly, your purchasing power drops because other people have to ask higher prices to do all the stuff you’re buying.

            If you’re paying 10% of your income for energy, and energy prices rise by +20%, the direct cost would seem to be insignificant because you’re not paying 12% your income on energy – big deal. You’re still left with 88% of your money – or so it seems.

            Where earlier you were left with 90 cents on the dollar to spend elsewhere, after the general price hike your remaining purchasing power would be equivalent to 73 cents. You would have to buy approximately 1/5th less of everything else, and that is not insignificant – especially for the poor who are already having to compromise.

          11. Energy runs the whole economy, fair enough comment, but as Renewables means vast amounts of ‘free – no really take it all off our hands right now’ energy at times, and is on average actually very cheap to build for its energy output so the cost per watt should end up really being about the same at worst to fossil fuel – which therefore means a vast amount of the driving force of the economy isn’t paying as much for energy, your purchasing power likely goes UP! As the average cost has almost certainly dropped, even ignoring the peak times giving the stuff away moments, and those moments are definitely very very very very cheap – heck folks have got actively paid to charge their EV and House battery by the grid before as its cheaper than modulating the other power stations!!!

          12. > Renewables means vast amounts of ‘free – no really take it all off our hands right now’ energy

            It’s not free. It’s just given away because governments are handing out subsidies for it – forcing other people to pay instead. The cost of energy is the cost of the generator divided by the total amount produced in its lifetime – regardless of who pays and who doesn’t.

            Producing it is not free, and building the infrastructure to utilize such random energy costs more money, making it actually more expensive.

          13. > got actively paid to charge their EV and House battery by the grid before as its cheaper than modulating the other power stations

            Germany, right? Yeah, and that’s because there’s a legal right-of-way for renewable power, so other power stations HAVE to modulate instead of just shutting down the wind turbines. The government pays for every kWh of renewable power sold to the public and forces the public to buy it, so the producers can make maximum profit.

            The UK does it the other way: if there’s too much power and not enough buyers, the wind turbines shut down, and the government pays the owners for however much they did NOT produce.

            It’s just… mad.

          14. Dude the Germans are crazy for solar power not because they are stupid greenies but because they hate Putin. The practical alternatives are wood heat or sending money to a Rusky bastard.

            Yes, they pay the highest electric rates in the developed world, to deny some money to the worlds biggest enema nozzle.

            They do a lot of wood heat there too, for the same reason. Better than giving more money to the god damn Russians.

            There are ‘stupid greenies’ in Germany (they have a political party), just not that many. I’m talking about Germany in general, broad brush.

            Ref: Son of Germans, visit family there fairly frequently.

          15. Dude in almost all the developed world fossil fuels still have massive massive subsidies, and renenewables often don’t get anything much, you just don’t notice because any new money for renewable is NEW subsidy, where legacy propping up and aiding of fossil fuel exploration, extraction, refining (etc) to make the important bits cheaper has been done probably for longer than any of us have lived, i expect it happened even in the earliest days of industrial revolution’s rapid growth, its certainly been happening alot longer than I’ve lived…

            As for paid to take electric away I know that has happened in many places, not just Germany…

            And really the excess spikes of renewables are ‘free’ – you needed to build enough to meet the minimum output targets, but that means heaps and heaps of power excess that is effectively at no added cost, its an economic magic trick really, as you had to spend that much to build them but often enough they give you back way way more than they are expected to, and all renewables if they are if not the cheapest to build for output are damn close to it…

      1. > what happens if there’s a hurricane that rips a quarter of the wind farms to the ground, with a hailstorm that breaks a significant number of solar panels?

        Then again, what happens when there’s a crisis in the middle east that halts oil shipment through the Persian Gulf for a while? Or a crisis in eastern Europe that shuts down half the regions gas pipelines?

        Nah, On second thought, those kinds of scenarios are too unrealistic to worry about.

        1. That’s why we have strategic petroleum and gas reserves in pretty much every country.

          Disruptions in oil trading routes are relatively quickly bypassed. Same for gas: the EU has actually replaced about half the missing gas from Russia by increasing LNG imports by ship. Development of the LNG infrastructure has been in the making since 2014 – for obvious reasons – and it’s rapidly growing.

          Meanwhile, to replace a significant portion of your renewables infrastructure after a major disruption would take billions and last several years and, if not for the petroleum/gas reserves, you would have nothing to go by in the mean while. It is paramount that any fully renewable power system has an equivalent of a strategic reserve or else it’s just a social collapse waiting to happen.

          1. As the very nature of renewables is to be very very distributed anything that can really take lots of them out without depopulating or destroying all the infrastructure so there is no need for the power anymore anyway is damn nearly impossible…

            You don’t need massive reserves as nothing short of armegeddon can actually knock down enough of the distributed renewables to be more than an inconvenience of small impact, and at the end of days nobody cares if there is 6 months of fuel in bunkers somewhere, too busy…

          2. > nothing short of armegeddon can actually knock down enough of the distributed renewables to be more than an inconvenience of small impact

            Quite the opposite. The large distributed system is fragile because it depends on most every part being present and functional to operate. There’s been studies on how much of such infrastructure you have to disrupt to bring it all down, and its alarmingly little. System efficiency and reliability are at odds because redundancy costs money.

            The EU synchronized grid is an example of exactly what we’re talking about. January 2021 was the last near-miss of a major EU-wide system blackout. The synchronized grid split into two non-synchronized regions that could no longer trade power, but this prevented the spread of a cascade failure and everything resumed to normal.

          3. The problem is that electric grids have to balance input and output at all times, so sudden disturbances can escalate and set off something that resembles an epileptic seizure in the grid.

            Very large scale distribution requires very high capacity power corridors, which means there is now long distance point-to-point links that carry dozens or hundreds of gigawatts of power, and the sudden failure of one such corridor deprives entire countries of power a thousand miles away, which makes the failure spread in the blink of an eye.

            The alternative to that is the US style “bucket chain” grid where each area transmits power to the next in line, but this is not really the sort of distributed system we’re talking about because the total transmission capacity from one side to the other is limited by supply and demand in the middle. You don’t have the sort of geographical distribution to support the “it’s always windy somewhere” operation.

          4. But no one unit or even cluster of renewables being knocked out has anything like the impact of a single powerstation having issues Dude – a single failure of anything in a fossil fuel and nuclear powered grid system is orders of magnitude more impactful as its individually responsible for massive more, where the very nature of renewables in being so damn distributed is any problem in one generator, substation, transmission line is a much much much smaller fraction of the whole!

            And it really doesn’t matter when things go wrong if all the grids end up for a while acting independently, the power is still on life goes on pretty much as it always has, and basically nobody noticed… Or even if there is a blackout – as loosing power for a while isn’t a big deal, it happens to us all once in a while large company or homeowner for a variety of reasons, all that matters is its recoverable, as as you can’t actually kill a huge distributed network of renewables properly its going to be recoverable in short order…

          5. >or even cluster of renewables being knocked out has anything like the impact of a single powerstation having issues

            Yes it does. The grid failure that almost hit Europe in 2021 was about just 2.2 GW of power suddenly going missing because of the synchronization loss, which was caught by Italy dropping 1,000 MW of industry offline and France bringing 1,200 MW of emergency reserves up. Mind, such emergency reserves wouldn’t exist in a no-storage only-disperse situation with renewable power.

            We’re talking about a scale of to 1-2 large wind farms. Meanwhile, a hurricane can be a quarter of the size of Texas, and wipe out or force the closure of multiple wind farms at once across the state, and adjacent states as the storm proceeds further inland.

          6. >a single failure of anything in a fossil fuel and nuclear powered grid system is orders of magnitude more impactful

            No it isn’t, because such systems are built with dispatchable spare capacity. See:


            >”Most power systems are designed so that, under normal conditions, the operating reserve is always at least the capacity of the largest supplier plus a fraction of the peak load.”

          7. Something like a hurricane is of no damn surprise, the grid will be ready for it all the turbines locked off so most of them in the path will be ready to roll once the wind dies down to usable levels again, its not a shock to the grid at all…

            And you always build in reserves/margins to any system, even the most budget cut under invested system that actually needs to work will have to have some, just because the methods are going to be slightly different to how it is done right now on primarily fossil fuels makes no odds at all to the ability of such a system to function perfectly well… Also it takes meaningful time to bring up reserves even with the fast gas power stations just because the capacity is there doesn’t mean it can be instantly deployed in time to correct for everything…

            In your hypothetical Hurricane the absolute worst case is deliberate brown outs, and that is stupidly unlikely as the winds miles and miles out but around a hurricane are damn certain to be there, and be there at a nice predictable consistent output for the duration, so you are more likely to be in oversupply of WIND power despite shutting down everything in the dangerzone, as every other windfarm quite probably in the whole damn nation will have a nice stiff breeze to work on…. At least assuming you actually have the distributed network built.

          8. > you can’t actually kill a huge distributed network of renewables properly its going to be recoverable in short order

            Here’s the contradiction: renewables don’t work reliably enough as they are in small areas – small being the size of Texas. You solve this issue by introducing a larger network the size of multiple states, or the size of continental Europe etc. which is all needed to average out variations in any individual area. But, the system only works as a whole system, which makes it inherently sensitive to small disruptions, and it won’t recover until those local disruptions are resolved.

            It’s a house of cards. It makes small localized problems become big system-wide problems, and recovering from those is made all the more difficult and costly.

          9. No doubt at some point my other comment will make it through the moderation system, but in short dude pfft…
            EVERY SYSTEM that actually matters gets operational reserves in some way or other, even if its vastly underfunded and should get more when its important you will have reserves! And even when you have them it doesn’t prevent all possible failures… Just because the methods are somewhat different from currently does not negate renewable based systems from being just as reliable.

          10. >EVERY SYSTEM that actually matters gets operational reserves in some way or other,

            Then you should look at California, and how they’re having blackouts exactly because they didn’t build enough operational reserves while forcing more renewable power online by government policy. Reason being that the required reserves (i.e. batteries) simply cost too much, and too little money was allocated too late. In effect, there is not enough operational reserve in the grid and it has become unstable to the point of occasionally breaking down.

          11. The problem in California is preventing bush fire during the really hot and dry weather of recent years isn’t it? Nothing to do with the grid actually being incapable if you don’t care about the risks of burning down the world around you…

            Also its California perhaps the most bonkers place in the developed world where everything must be labeled as potentially causing cancer, useful (and pretty damn safe) industrial chemicals can’t be imported, etc – perhaps not the best case for you to pick as the population there seem to want this chaos by their own choices…

    2. And in other news:

      >Porsche and its international partners have started building the factory that will produce a new synthetic fuel starting in 2022. Located in southern Chile, the plant will make fuel for race cars

      >electrolysers split water into oxygen and hydrogen using wind power, hence why the plant is in Chile; it’s located in one of the most reliably windy parts of the world. CO2 is then filtered from the air and combined with the hydrogen to produce synthetic methanol (…) When it arrives in Europe, where it will be distributed by ExxonMobil, it could cost about €2 per liter depending on taxes.

      It’s well to note that due to current world situation, gasoline already costs €2 per liter including taxes. The point however was that hydrogen – as much as it is touted as the next economy – is not actually the most optimal way to store renewable energy in the long term, and fortunately we don’t have to use it. We can make every sort of chemicals other than hydrogen and ammonia.

      1. And mind, here the energy efficiency of producing a combustible fuel has absolutely no relevance as long as the end product cheap enough to buy.

        That is of course, as long as you’re not patching up poor economic performance with state subsidies to hide the cost.

    3. I’m from Europe and I though the Texas grid, being an independant power grid, is basically the reason for these droughts/outages. Local influences have a huge effect.

      By creating a grid over a large area, like the European power grid, you can minimise these power droughts. Off course, they will never be completely gone, but the impact of local events is severly reduced.

      So Texas doesn’t seems to be a good example to draw conclusions from.

      1. It is, but it’s also very big. On the map, it would cover Germany and all the small neighbors, and a third of France.

        With large area interconnects, you’re also struggling with local effects. It’s not just simply going to average out when one side of your grid is in a completely different climate than the other, and the transmission capacity between them is limited (not to mention inefficient). That’s part of the reason why the DESERTEC plan, which still leans on backup from coal power, involves everything from Norway to Gibraltar, North Africa and parts of Middle East – assuming superconductive ultra-long DC cables connecting it all together.

        It’s also a fragile network which is subject to natural disasters and hostile nations, wars and corrupt regimes along the way. That is why planning for energy independence over an area roughly the size of Texas, even in Europe, would be the smart idea.

        1. That’s the DESERTEC system map we’re talking about, if you want it to work with the same reliability as the present European synchronous system:

          Of course, none of the countries in the system could separate even for a moment without collapsing the entire house of cards. As one of the authors in the study commented: “We won’t solve the problem by building a larger network.”

      2. Texas being an independent grid also has less to do with the recent outages. There was power available from west and south. For the eastern interconnects, they had the exact same weather problems and not enough power was available across the state borders.

        Texas has grid interconnects to other US states and Mexico – it just can’t trade effectively within the US without triggering some legal clauses over interstate sales, which would bring the entire ERCOT system under FERC regulation.

        The US has this funny system because the federal government is not allowed to mess with the states’ internal trade matters. So, it has created a bunch of independent regulatory bodies like the FERC which operate to mess with the states’ internal matters without direction or review from the congress and without any democratic oversight from the public. They collect their operating costs directly from the entities they regulate, which is problematic. Since it’s “hands off” regulation, it is technically constitutional – but it’s essentially a very bad idea and Texas is well wise to stay out of it.

  2. The simplest storage vessel for CAES is likely to use a Tunnel Boring Machine to make a few parallel tunnels, slurry fill the gaps and line the tunnels with steel to make them airtight. But this would require fairly geologically stable conditions and a decently strong rock like granite, while also being built at a semi large depth.

    A 4 meter dia tunnel spaced 35 meters center to center beneath 80 meters of granite storing 160 bar would yield a safety factor of about 5, while storing 282 kWh per meter of tunnel. Or about 36500 $ worth of lithium batteries per meter of tunnel.

    Drilling 4-8 meters per day on average in granite isn’t unrealistic for a typical TBM, so that puts the budget at 140-290 k$ per day for the tunnel to be cost competitive against lithium batteries. Except, some of the tunneling cost will have to be spent on the compressors/motors to actually use the tunnel as energy storage.

    However, the main problem I see with most CAES installation ideas is that the compression/decompression is too rapid, speed isn’t desired when working with compressed air, at least as far as efficiency is concerned. I also suspect near isothermal operation to be more practical.

    But CAES has the advantage of not technically requiring any fancy materials for the energy storage itself, and when the tunnel is built one shouldn’t have to do all too much maintenance on it over time, perhaps reline it every other decade. And preferably a storage facility should have more than just 1 tunnel, as to provide redundancy for maintenance and inspection reasons. (though, inspecting it with a robot is likely more practical.)

    Though, not every location on earth has suitable geology for this approach, but a fair few places do have suitable stable rock. Else one can look into high pressure tubing instead, might also be sufficiently cost effective.

    1. Forgot to add that the “282 kWh per meter” is when one discharges it at 100% efficiency, realistically we might only reach 85% if doing things really slowly, so we could get upwards of 239 kWh per meter, likely less. So this is only equivalent to about 30900 $ of lithium batteries per meter.

    2. Depending on depth you can probably get away with almost any geology, as the shear mass of stuff around it is more than sufficient to take the pressure’s load, you just need something less porous as a lining to hold the air in and tough enough to take the weight from above when there is no pressure…

      That is all dependent on just how deep, how much pressure etc, but there isn’t really anywhere outside of the very active earthquake zones such a tunnel couldn’t be made to work. But as tunnel boring is stupendously expensive upfront I can’t see it being used much, even if the long term durability of the system makes it a good investment…

    3. The simplest compressed air storage is a big bladder anchored deep under the ocean. The hose leading down to it absorbs heat from the sea water on the way back up, so you don’t need to bank the heat; the sea absorbs and returns it. A similar alternative is a big rigid tank, similarly anchored, that you pump air into, pushing the water out the bottom. Either way, water pressure pushes the air back up.

      An equally simple system is a spherical tank, also anchored to the sea bottom, with a pump at the bottom. For storage, you pump the water out, leaving water vapor inside; and let it back in, through a turbine, for power. Deep under the sea, such a tank can store a remarkably large amount of energy. The only connection back to the surface is the power cable.

      The compelling advantage of hydrogen and ammonia synthesis is that, once your local tanks are full, you can continue producing fuel for sale. I.e., the expensive synthesis equipment is idle only when you are drawing down your tanks. There will never be a shortage of demand for your excess. And, if your tanks run low, you can ship in more. The advantage of synthetic fuel systems is that the tankage is extremely cheap, and furthermore portable. Equatorial countries will produce for export to higher latitudes in winter.

      Liquifying air is extremely mature technology, so already very efficient. In practice, this would be liqufied nitrogen. To extract power you let ambient air warm it up, and send it through a turbine. A cold-nitrogen turbine needs much less maintenance than a steam turbine. This method works even at small scale. Again, liquid nitrogen is industrially valuable, so excess power goes into liquid nitrogen for sale.

      The key to understanding storage for renewables is that wind and solar generating capacity has become so cheap that absolute round-trip efficiency doesn’t matter very much; to cover the difference, you just build out more panels. There is no penalty for building out too much generating capacity, because any excess produces fuel for sale.

  3. standard railway equipment can be used for energy storage
    like this: electric locomotive goes up a long hill,when there is excess power,when power is needed it runs down the hill useing
    the motor/generator to return energy to the grid.
    everything is off the shelf.
    cogged rails would allow a steeper grade to be used and result in
    a more compact system.
    round trip efficiency is going to be similar to pumped water.

    1. If we have a train loaded with 80 tons per car, and a total of 200 cars, making for a nearly 5 km long train. And driving it up a 1 km elevation difference, then it equates to 43 555 kWh of energy at 100% system efficiency.

      Then there is the question of rolling resistance, this is usually quite low for trains.
      There is also air resistance, fairly small at low speeds.

      But one might rather argue that better train scheduling could result in more storage, since if freight trains overall go downhill during peak demand hours, and uphill during low demand hours, then this is likely going to result in a larger amount of energy storage than a dedicated train. After all, there is hundreds of freight trains out and about all day long.

      But yes, the idea of using trains for energy storage isn’t all that wild and has been tested in practice.

      Though, I suspect it is about on par with flywheel energy storage on a day to day level. All though, one can leave a set of freight cars on a mountain for the next season, but here it might be more interesting to move large blocks of stone on these trains and make a few trips instead.

      1. Grades over 2% are relatively rare, 1% is more typical, which means it would take a train 100 kilometers to drive up 1 km.

        One of the reasons you don’t have very steep grades is because it becomes mighty difficult to stop a heavy train at high speeds downhill if it just keeps on picking up more speed. If the wheels start to slip, the dynamic friction becomes a fraction of the normal rolling friction and there are several cases where this has lead to a runaway train and a big wreck.

    1. That’s a cool thing, but the requirement for 2,000 C heat source is a seriously limiting factor. It’s difficult to get things that hot, to keep them that hot, and not many materials even stay solid at those temperatures.

      The proposed idea of highly insulated graphite is a bit silly. First of all, graphite is not cheap and it’s very inefficient to make it out of conversion from regular carbon. Secondly, hot graphite burns extremely well in air.

  4. All big Coporate problems. Stick to Independant.
    A home owner can have it all. Battery rolling around as transportation can supply the home when needed or everyday.
    A home owner can easily switch to the best tech to hit the market. Dinosaurs can not help. Corporate Problems lead to big government regulation that is bad for everyone.

  5. Where are you going to get the 5 years worth of storage needed to survive the _next_ volcanic winter? Think about the demand when the sun power stays at lower than winter levels for multiple years. It will be very cold so you need more heating, and you can’t grow crops in the field so you either have famines, strip the oceans of what fish remain in them, or you move to indoor agriculture under grow lights etc. on a massive scale.

    1. In that scenario, there are no combination of existing food-stores and reconfigurable-to-in-time farming techniques to avoid everyone* starving to death long before needing to worry about energy storage. Amongst other issues.

      1. All the more reason to depend on nuclear and geothermal for the energy for EVERYTHING including food production. Growing food outside in the fields is VERY convenient compared to underground in illuminated tunnels but those tunnels are much more controlled environments and there are many benefits with that….No pesticides, 24/7/365 continuous growth, no hail-storms/other-destructive-weather, no droughts. Besides all those plants make an underground bunker much more pleasant anyway….good cultural preparation for living offworld too.

      2. Yes that additional problem needs to be addressed concurrently, as well as the wars that will break out, civil and multinational, if it is not addressed fully. However you have to get the energy issue fixed first as the food production issues has that as a key requirement. Depending on the timing and location some people may start freezing to death before the food runs out.The bottom line is that humanity is living on borrowed time and worrying about all of the wrong things, like lemmings running toward a cliff while looking at the sky to check for eagles.

  6. Nuclear reactors store years worth of reserve energy right there in the production facility. They work where the land is too flat for gravity storage and on cold dark windless nights. Apart from deep geothermal, it’s the only super-reliable non-depleatable energy source. Eventually some form of nuclear is going to outperform all other options to the point of them becoming mere niches. (nice niches, but niches never-the-less) I’d only want to be counting on Nuclear and Geothermal if we ever get another A.D. 536, for instance. We could be in for worse years too. Go nuclear or perish is the wise path. We can work out the logistics for making it safe, reliable, and environmentally appropriate. For the most part, we already have. We simply have to have the will to carry them out.

    1. Cost of building nukes is several times that of renewables + storage. Cost of just *operating existing nukes* just matches current renewables + storage. But cost of those is still in free fall. During all the years you are building your nuke, you are also paying through the nose for coal, and the price of the competition is falling. Most likely the nuke gets cancelled before it is finished. Even if finished, it will never be able to produce economically competitive power.

      1. >Cost of building nukes is several times that of renewables + storage.

        Not really. The latest EPR reactor that started this year, built in Olkiluoto Finland, sells power profitably around €40-50 per MWh despite the huge budget-overruns and delays. The original estimate for the producer price was to be €19 per MWh but that was never reached.

        Unsubsidized wind (LCOE) without storage ranges from €25 to €52 per MWh depending on location. With batteries, assuming €150/kWh and 3,000 cycle average life, you have to add another €50 per MWh. The batteries alone cost more than nuclear power.

        Then adjust for efficiency losses, say -20% and you arrive at somewhere between €90 – 127 per MWh which makes the renewable wind between 2-3x more expensive than a modern nuclear power station – even when the power station in question was a total hack-job in terms of project planning.

        >But cost of those is still in free fall.

        Not really. The technology has become cheaper, but for example with solar power it’s difficult to push prices down further because the module price is no longer the dominant factor. The actual modules account for about 10-15% of total system cost. 2/3rds of the price per watt is tax, permits, land, financial, labor, and logistics costs. The cost development trend has leveled off since about 2010.

        1. Though of course you can adjust the estimates down by using the renewable power directly, so nuclear power would be roughly the same cost.

          But, as you build up more renewable energy, the portion that has to be stored in batteries starts to increase sharply. Around 20-30% grid penetration levels the peaks start to rise permanently above the demand, because renewable power has a high peak-to-average output ratio between 5:1 to 8:1 for various renewable power options. This means, when the average output is closer to 100% of your demand, the majority of your total energy production comes from the peaks and has to be stockpiled into batteries.

          If not for batteries, you would have to double or triple the number of generators and toss the excess energy, which means doubling or tripling the power prices. It’s a balancing game between whether you want to pay for over-provisioning or batteries, and how much of each.

      1. Yeah, only, it’s much easier and cheaper to make a big reservoir on top of a hill, than to carve an equally large and deep hole in the other direction.

        Also what with the habit of nature to fill such holes with ground water, so as you pump it out it gets filled back in.

  7. Nuclear is to renewables as Space-X space-tech is to just watching old Star Trek episodes. All of those things are cool but some are more relevant to practical human needs than others. We need to chose the most reliable options so we live long and prosper.

  8. One interesting fact is energy can be stored in the air. Of course, the energy usually travels. Light travels from the sun to the Earth and can be considered stored in transmission. Similarly, sound can be stored in the air, albeit with some loss to surrounding molecules. Yet if sound is resonated with enough input energy, the sound can be considered stored in the pathway of it’s resonance.

  9. Maya, your columns in energy storage are great. For this column I had hoped to be read more about the consumer side of the meter. I know that you already wrote about V2G and concluded that V2H was the most viable application of BEVs, but I’d like to hear more about V2H. In particular, what’s the analysis for homes with local generation (i.e. solar) and a BEV where the homeowner wishes to maximize grid independence both for grid outages and for avoiding paying the grid for power when the local generation is off (night)? I’m not talking about complete independence, but rather minimization. As more utilities kill favorable net metering terms, this has an effect on solar cost effectiveness.

    Questions include: does nightly self-use of the BEV battery present the same duty cycle problems as V2G? How significantly does this type of approach increase the required generation capacity for the average home? In what parts of the country do grid pricing and sunlight hours make this most favorable? What’s the state of bidirectional BEV charging standardization? For a homeowner looking to use their BEV to avoid outages, how does the total cost compare to a traditional natural gas generator?

    1. Lithium battery chemistry is a ridiculous complex thing to try an figure out a true management plan for best lifespan, keep it too full it dies faster, get it too hot it dies, empty it too much or too fast it dies etc. So using a vehicle to power your home might actually work out increasing the lifespan of the battery if you otherwise would have left it fully charged (somewhat depending on how aggressive on the chemistry your particular EV brand is).

      However most houses won’t draw more than a couple hundred watts over night, many not even that. In those cases using your vehicle to power the house overnight is such a low draw and tiny discharge it really won’t do much to the battery at all – put in context the entire night of power probably wouldn’t take you a mile (at least not the mile you initially accelerated in). And with the shear capacity of a BEV I would think you could easily go off grid with it as your primary batter if you wanted, not that I would suggest doing so, you would require a reasonable number of solar panels and a small home battery for keeping the fridge (etc) on when the car isn’t there (as I assume you want a car, not just a battery).

      On the subject of avoiding outages a BEV being purely a solid immutable energy store it can only help you as long as the battery lasts or in the rolling brownout – there is no way of topping it up for a long duration outage the way you can with a generator, if that is your priority you want one of the plug in hybrids – a pretty big battery, enough for almost everyone’s daily trips, may even do a few days between charge but with a (usually) little, efficient petrol generator on board for those times the battery isn’t big enough – like powering your house for days. Usually I’d say they are a bad idea compared to just getting the bigger battery EV, as why haul around the backup generator and fuel tank full of fuel that will probably go bad before you use it 100% of the time, but there are users and usecases it is worth it.

      IFF we assume your outages are only ever a single day (or anything well inside the scope of the batteries) I would suggest the BEV is far cheaper and way more reliable than a gas generator you probably haven’t fired up in eons, and quite likely haven’t done the maintenance on – human nature being what it is most folks don’t look after their backup systems, though some do of course. Which means it is going to run terribly, if it runs at all, and even if you do treat it with great care to be sure your backup is actually reliable that costs you extra all the time to maintain something you really shouldn’t need…

      Where with your BEV you were using it anyway, keeping it maintained anyway, even if using to to cover the outages took away of year of usable battery capacity, which it shouldn’t come even remotely close to doing, its still just a small share of the costs of the replacement, as that would have to have done anyway because you need your transportation, it just made it happen a touch early.

      So the cost of using a BEV over a gas generator is whatever the price differential between units of Electric and bottles of gas work out at (I’d bet Electric is cheaper, certainly is than bottled gas here – so actually saving money at this point, but that is going to vary hugely globally I would think), added to the extra maintenance cost the generator requires and whatever the share of lost battery life works out at, which if we assume a 10-15 year life (which seems to be very plausible for most users) and a reasonable couple of months worth of added wear and tear from using it as the backup power source its only costing you in the ratio (Lost lifespan – say 2 months):(To original max lifespan in months) – so something like 1:60 of the new battery or vehicle costs can be considered its share of just lost lifespan, which may well be entirely offset by the cost of gas bottle and doing the maintenance on the generator over those years…

  10. Texas Interconnection doesn’t have direct AC ties to the other interconnections. All of their inter-ties are DC back to back converters (converts Texas power to DC, then inverts it back to AC in sync with the other interconnection). These inter-ties are routinely used to trade energy inside/outside Texas Interconnection, but they only have limited power transfer capabilities. During the winter event in Texas, they had to shed ~10,000 MW of load, but the inter-ties couldn’t have supplied anywhere near enough to make that unnecessary (maybe ~1MW or so at most).

    The lack of AC interconnections within Texas Interconnection (contained entirely in Texas) means that the FERC (Federal Energy Regulatory Commission) doesn’t regulate their electricity *market*. Bulk Power System reliability is still under FERC jurisdiction via the 2005 Energy Policy Act. FERC regulates interstate electric markets themselves, and they delegate electric reliability to NERC (North American Electric Reliabiliity Corporation). Ultimately, in the US, FERC will approve any fines/standards NERC creates. NERC also covers parts of Canada and Mexico (7 provinces + a piece of Baja Mexico) beyond FERC’s jurisdiction. TRE (Texas Reliability Entity) is a member of NERC (which includes all of ERCOT).

    FERC is a federal agency of the US government that is run independently. This means that only 3 commissioners (of 5) can be of the same party. The President (and Congress) nominate and approve FERC commissioners. Their decisions are subject to challenge in federal court. This approach is completely constitutional (not “technically”).

    FERC’s budget is approved by Congress and is recovered from costs they impose on utilities (allowed under legislation). NERC’s budget is approved by FERC and is recovered from costs they impose on utilities (an annual assessment dependent on size of utility).

    Assessed penalties on utilities increase neither FERC nor NERC’s budget. Those penalties just reduce the annual assessment for all of the other utilities. There is no incentive for either entity to either excessively limit or excessively fine entities – it’s a zero-sum equation.

    1. >This approach is completely constitutional (not “technically”).

      The point is that the feds have no constitutional right to butt in states’ internal matters. Setting up independent regulatory agencies is a loophole that allows the federal government to mess with the states’ internal matters anyways, but in consequence they have created something which is worse.

      The criticism is that NERC/FERC are simply rubber stamps for approving things like gas pipelines, subject to regulatory capture and corruption since they were created with only nominal oversight by the public. It’s kinda like when the FCC became AT&T’s rubber stamp for doing whatever they wanted.

      1. ERCOT isn’t regulated by FERC. It’s regulated by PUC of Texas. Texas themselves has joined as a NERC member. As I have previously stated, this is completely constitutional both for the federal as well as Texas governments.

        FERC commissioners are appointed by your elected representatives, if you are an American. The public gets to see and comment on everything they do. In my experience, they tend to look more at reliability and markets, what they’re supposed to, instead of say, environmental aspects, which they aren’t supposed to. That’s a subject of debate within the government as to who does that and with what authority. EPA tends to get more involved with the environmental aspects through their comments. Those FERC orders get quite specific on what FERC wants, and it’s usually not utility friendly what changes they ask for.

        This agency model is just how our executive branch works. You’re free to disagree with the principle, but it’s not an end run around the constitution. Our government is a republic that the constitution gives these agencies authority to operate. The agencies only enforce the laws that Congress creates, and the federal courts are the place to argue otherwise (which happens a lot).

        There are probably other ways to do it, but this is the way federal rule making has coalesced. Are you suggesting the public vote on every federal rule somehow? You might be voting a lot with all the work going on… Just vote on less rules/do less work? Awesome! Now you’re letting industry really do what they want. This is what you wanted, right?

        In my experience, FERC staffers and commissioners are not corrupted or regulatory captured. I would argue they use quite a heavy hand on utilities. Your experience may differ.

        Also, only FERC handles interstate gas pipelines, not NERC. Gas is actually handled by a separate office at the FERC than electricity. That can change if folks want it enough…like I said, they only do what Congress allows. If you’re that adamant, work to change the law.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.