The Great Grid Stabilizer

The weird world of energy storage

In 2022 a strange machine was installed in Moneypoint, Ireland: A giant flywheel, weighing 130 tons, spinning faster & slower. Why?

The machine is a Synchronous Condenser, among the world’s biggest flywheels, and its purpose is a bit unconventional: It adds inertia to an electric grid filled to the brim with on-&-off variable renewable generators, especially wind power.

Spin-up, spin-down…

On a conventional electric grid, turbines in nuclear or fossil fired power plants spin generators at a very specific speed, to produce the characteristic 50 Hertz signal in all our electrical appliances: 50 Hertz means 50 up/ down or positive/ negative cycles a second. Everything has to be synchronised and work together, creating a heartbeat for the whole nation, too fast for you to notice. It’s maintained, with impressive precision, between 49.9 and 50.1 Hz, all the time. 

Professional musicians aren’t this synchronised. The electrical grid shames them all without fanfare, keeping the beat.

Of course, sometimes something goes wrong: A substation shorts out, a power generator dips or it’s half time in the All-Ireland hurling final and everyone puts the kettle on simultaneously. Whatever, something has happened to wrench the grid out of perfect balance, and now there’s more energy being pulled out of those generators than is being supplied. In a fossil or nuclear-powered grid this isn’t too much of a issue, because all those vast turbines, discs and shafts resist the pull of deceleration, just enough to give you time to react.

This is grid inertia.

On a grid dominated by variable renewable energy sources like wind or solar, however, things get a lot trickier: There is a lack of huge spinning turbine shafts to provide passive inertia and so it must be sought some other way: Batteries are a possibility, as is our gigantic Moneypoint flywheel.

And even more sinister, sources like the wind have a mind of their own, and can suddenly surge or decline without warning, increasing the very demand for grid inertia that they strip from a network.

And, of course, sometimes the wind does not blow at all.

As variable renewable power grows, so does the demand for storage, and this is a world of vast variety and crushing compromise.

Let's step in, and see some of the ways it's done…

1: King Hydro.

In the Wicklow Mountains, south of Dublin, a curiously oval lake sits at the very top of a mountain that it has no business being on. No rain could keep such a thing full at this height, and the ground running up to it is smooth and domed, like something outside of nature. It's a hidden beauty. A mountaintop secret.

A tower rises in the middle of the lake. Far below, machinery churns.

Here, we store energy with water, cashing a cheque writ by gravity itself.

Pumped hydro.

We'll start with the big boys: The storage systems that must hold unspent power for days, week, even months against those times where the sun, it-don't-shine and the wind, it-don't-blow.

The biggest of these is pumped hydro storage, which is also the biggest form of grid storage worldwide by stored energy capacity. It involves pumping water uphill to a storage reservoir when power is cheap & abundant, and letting it flow back downhill through a turbine when it's scarce and expensive. All you need is room for a big enough reservoir and lots of vertical displacement. Water, as long as it doesn't evaporate or become clogged with algae, can be held pretty much indefinitely like this, making pumped hydro storage one of the longest period storage options available. It's also pretty cheap per stored megawatt-hour and can release high power in a few tens of seconds notice.

Not quite nimble enough for second-by-second grid stabilizing, but very handy when the wind surges and falls. It may stop you needing to spool up a gas turbine peaker plant to fill the gaps.

It has drawbacks, however: For one, its specific energy per pumped kilogram is very low, so pumped hydro stations need to be massive to be useful on a national level: Water that falls one hundred metres releases a stored potential energy of 981 thousand Joules per ton of water, which is enough to boil a kettle for ten minutes.

So you need a lot of tons.

It also, inevitably, loses energy to friction and electrical losses in the process. Pumped hydro has a round-trip efficiency of 76%-85% when storing grid electricity, so between 15% and 24% of the energy is lost. Not bad, not great.

And obviously it can't be built just anywhere. If you've a hilltop reservoir or a deep unused mine cavern, great. If not, you're fresh out of luck.

What other options have we?

2: Squashing Air.

Compressed Air Energy Storage, or CAES, is one of the oldest storage systems around and can be reasonably simple. It was used to power mining vehicles and perform electrical power smoothing in the 19th century! It can also store for long-ish periods of time, at a much higher energy concentration than pumped hydro, and can be built more or less anywhere.

So why isn't it everywhere? The simple answer is: Appalling inefficiency, and that's all down to one nuisance factor.

Heating.

The goal of CAES is to store energy by squishing air using pumps, which can then be expanded later through turbines to release the energy. Most CAES systems aim to compress air to 70 bar, or 70 atmospheres of pressure. So far so simple, but when you compress air you also heat it, to a brutal 900 kelvin if you compress it to the 70 bar region, and that extra heat will then happily conduct and convect itself out of your storage reservoir and undo most of your gains, bringing your round trip efficiency to a disappointing 40%, meaning that most of your stored oomph is oozing away as waste heat.

How do we improve this?

One way, in ‘Advanced Adiabatic CAES’ is by using a separate circuit to store the heat generated through compression, using a heat exchanger medium and either crushed rock or molten salt to store the thermal energy. This can then be injected into the expanding compressed air through heat exchanging during the generation phase. In such a way, round trip efficiency can be raised to a more useful 70%, albeit at the cost of extra complexity. This also somewhat caps the maximum time for storage, as retained thermal energy is leakier than compressed air and you slowly lose it over time.

Another option, ‘Supercritical CAES’, makes use of the supercritical phase of air, where extremely high pressures are coupled with cryogenic conditions, to improve the efficiency of compression and expansion. A supercritical fluid is neither fully liquid or gas but sits in an in-between zone where compressibility varies with its pressure and temperature. This is exploitable for greater storage efficiency, but requires cryogenic cooling of compressed air and storage in liquid form, with released heat during compression stored in a separate thermal reservoir.

Inevitably complicated, supercritical CAES can in theory add additional efficiency to compressed air storage, while containing the entire shebang in a smaller array.

Supercritical CAES makes use of purpose built cryogenic liquid storage tanks. More conventional CAES, with or without heat storage, can also use natural formations such as salt caverns, depleted natural gas wells and the like. There are an abundance of such sites worldwide, which is one attraction of these methods.

It's not as cheap as King Hydro, though, and requires a good measure of compromise.

Those are our bulk, long-term storage options, and they're basic enough, but the story doesn't end there. The Grid is a complex orchestra, and it's not All About The Bass.

What about the treble?

3: Different Tunes To Play.

Grid management isn't just about keeping the lights on when a large wind farm suddenly decides that it's not interested in generating today. It's also about the small things: Managing transients, keeping the frequency harmonised and managing power quality, for many things that draw power are sensitive and need it to be well controlled.

This doesn't require the big drum beats of bulk storage, but the piccolo flute of power quality management.

Let's look at the opposite extreme: Flywheels!

4: Spin-up/ spin-down!

Offshore wind turbines grasp at the air over the froth of the Irish Sea, and some of this raw power ends up in Moneypoint, where one hundred and thirty tons of steel spins ferociously, faster and faster, to the rhythms of the distant wind.

Even simpler than pumped hydro, this one. Just spin up a big symmetrical steel flywheel on magnetic bearings in a partial vacuum and let it power a generator if you want to draw power back out of it. It brings the benefits of relatively high power density, very fast ramp times for rapid response, and the passive benefits of grid inertia.

On the other hand, the specific energy held by such a system is relatively low, though perhaps not as low as you'd think: A conventional steel flywheel system rotates at up to about 6,000 revolutions per minute, or 100 a second. So intense is this that the amount of mass in the protective enclosure is often about twice that of the flywheel, for safety reasons: You wouldn't want tens of tons of fast moving steel suddenly becoming loose in an enclosed space! 

There also exist specialist, pricey, high speed flywheel storage systems using carbon fibre composite wheels, which can rotate at 10,000-100,000 rpm, though their use is more aligned to traction and aerospace than grid power management.

In any case, flywheel systems are fast to respond, add useful inertia and have high round trip efficiency, at 90%-93%, but are let down by low energy density and a very high energy leakage, losing about 20% of stored energy per hour. 

But if your problems live in the seconds and minutes, not hours, then they're useful to have.

5: Power & Energy

Now that we've looked at two very different systems for energy storage, it's worth going back to basics and asking “what's important in grid storage anyway?”

Before we do, let's first qualify the difference between energy and power. Electrical power is current x voltage, it's measured in Watts and it's what your home appliances need to work. An LED lightbulb might need only 5 to 10 Watts, whereas a kettle might need two thousand Watts to function.

(Yes, lighting your house is really, really cheap these days. There isn't really a need to turn the lights off when you leave the room, whatever your parents used to say. -But I digress…)

Energy is power x time, usually expressed in either Joules (Watt-seconds) or KiloWatt-Hours (kilowatts x the amount of hours). This is what your utility company charges you for.

People get these two mixed up more than you'd expect: For example when “Europe's Largest Battery” was installed in England's North-East in 2022 to support the vast Dogger Bank offshore wind farm, an impressive power figure of 98MW was quoted, “enough to power 300,000 houses”... well yes, but with an energy capacity of 198MWh, it would only be able to power them for two hours, as the more honest publications admitted. As it turns out, the battery farm was less for strategic energy retention and more for power quality and smoothing the highly variable output of Dogger Bank.

In energy storage there are probably six big factors to consider: Cost per unit of energy storage, cost per unit of power output, speed of response to demand, round-trip efficiency, suitable storage duration and system lifetime. In an ideal world, you'd have a single storage system that'd be cheap, store lots of energy for a long time in a small space very efficiently, then output high power very precisely at a moment's notice, and keep doing this for decades.

No such system exists.

Sadly it's compromise all the way down. Keep that in mind as we look at our next storage option…

6: Smaller and faster: Capacitors.

The cheapest way of supplying stored power and the most expensive way to supply stored energy, a capacitor comprises layers of changed conductor separated by an insulating medium. It has the advantages of very rapid charge/ discharge, and is a compact, cheap way of storing and releasing fast-cycling high power for short periods, for current quality applications. 

Frequently used in electrical substations, you won’t see capacitors or supercapacitors used for bulk energy storage of any form.

7: Run it hot: Thermal energy storage!

In Morocco the Noor concentrated solar plant spreads like an inland sea, riven by the frozen waves of hundreds of concave mirrors, superheating oil in fragile tubes. At its apex is Noor Phase 3: A tall spire threading the sky at the centre of concentric rings of flat reflectors that aim the sun's fury straight at its peak, turning the desert daylight into focused fury. The spire's peak burns, a mote in the eye of God. 

Storing the sun.

A dramatic and bulky, though comparatively cheap form of bulk storage is thermal, which pairs well with generation processes that function on the basis of heat cycles, which is most of them. In essence, a thermal substance is either heated or cooled to store energy by imposing a temperature differential that drives a heat engine and subsequently a generator. 

Thermal storage systems can be separated into hot & cold, with the direction of travel being driven by heat exchanging in both circumstances, but in opposite directions: Hot systems involve storing energy in a heated medium, which can be molten salt, water, crushed rock… all kinds of things. Molten salt is a popular pairing with concentrated solar or some proposed nuclear power generators, as the temperature ceiling is very high and it's comparatively easy to handle, so long as it remains at elevated temperature. The classic disadvantage of molten salt systems is corrosion-proofing the working materials at elevated temperatures.

Molten salt is an example of a ‘sensible’ thermal storage system, which stores thermal energy without a phase change. The opposite is ‘latent’ storage, which involves a phase change to enhance energy storage density: Solid to liquid or liquid to gas, for example.

Cryogenic storage is another flavour, which we briefly covered earlier when talking about compressed air storage, and has the advantage of high energy density, but at increased capital cost and harsh round-trip efficiency. 

As a family of storage options, thermal energy storage is cheap to deploy per unit of stored energy, has good medium to long term storage potential with low daily self-discharge, but is primarily let down by poor round-trip efficiency caused by dependence on multiple heat exchange cycles: This is typically 30%-60% on thermal systems as a whole, but only 40%-50% on cryogenics.

That said, if it's used to capture heat from a thermal power plant before electrical generation, then this will reduce the overall efficiency impact: The Natrium nuclear reactor concept with thermal battery works on this principle, as does concentrated solar thermal power.

Compromise, compromise. Everything, everywhere is about compromise.

Will we ever find a one-size-fits-all solution?

8: Everyman’s Energy Storage: Lithium-Ion Batteries!

Powering the very device you're using to read this, the Li-ion battery is the true everyman of energy storage: A little bit expensive, though rapidly getting cheaper, it has a pretty high energy density, can charge moderately quickly depending on chemistry, can run at high voltages and has a high round trip efficiency of 85%+. It also boasts a reasonably low self-discharge rate of a few percent a month, though this can be much higher.

So why isn't it everywhere?

Well, it is the fastest growing grid storage option worldwide for many of the reasons just listed, though it’s still damnably expensive compared with the likes of pumped hydro, whose low, low costs per Joule are extremely hard to match.

As a point of comparison, a 2020 study which will be listed at the end of this article showed pumped hydro at between $5 and $100 capital costs per stored kWh, versus $600-$3800 for Li-Ion.

But while Li-ion may be expensive (it is!), especially compared to pumped hydro, what you get back is high flexibility and high energy density: A pumped hydro station will store between 0.5 and 2 Watt-hours per litre, whereas a Lithium ion battery manages 200 to 500, meaning storage capacity can be added where space is at a premium and geography doesn't allow for the entry of King Hydro. The batteries can be plonked down just about anywhere and can be good for either medium term storage or current management events measured in milliseconds. With appropriate control gear you can even use them to create grid inertia like the vast Moneypoint flywheel that we started the article with.

Truly, what you get for your money is an awful lot of flexibility. It's like rail travel versus owning a Range Rover.

But not all of us can afford Range Rovers, and sometimes a journey is so long that it's nicer to take the train.

Lithium ion covers short and medium term storage, and does it well, but are there batteries that can do really long distances, while we sit back and relax in our train travel analogy, knowing that it's all being taken care of?

Potentially, yes…

9: Hell’s Battery: Sodium-sulphur.

What use has sun parched, energy-rich Abu Dhabi for energy storage, you might wonder? In a land where easy oil flows like industrial lifeblood from the sand and sea, could you not just turn up the furnaces and generate whatever you like, whenever you like?

The Emirate answered this with the installation, in 2019, of the world's highest capacity battery plant, capable of storing 658MWh from a fleet of very unusual batteries.

For reference, that's enough to keep a small city powered through the coldest hours of the night. Why?

And how?

The chemistry of these batteries, and the fact that they were installed with Japanese expertise, gives a hint to their unique provenance. Abu Dhabi's huge storage scheme is made of a fleet of sodium-sulphur batteries, strange niche devices that must maintain an internal temperature of 300 celsius or more just to keep running.

These monstrosities are not new: They were invented in the 60s, but their unusual properties made them rare, possibly until now, for they possess unique properties making them ideal for grid scale storage.

For one, they're big. For another, their self-discharge is nearly zero, so they can hold energy for a long time if needed.

And they're cheap. A fraction the cost per kWh of pricey Lithium-ion across their lifetime, depending on how they're used. In part this is due to the commonly available materials used in sodium-sulphur batteries. 

But it's not all sunshine and roses. Sodium-sulphur batteries need to keep their respective electrodes in a molten state, which means, no matter how good the insulation, an element of maintenance expense is required. And their constituent electrodes don't play well with air or water: Molten sodium reacts very energetically with water, so safeguarding against such damage becomes part of the use case.

As ever, there is no free lunch.

There is a lasting hope in some quarters that a free lunch can be found if we look hard enough. That the mercurial temperament of weather-based electricity can be managed by just the right storage technology. That if we try hard enough we can live in a Sylvanian paradise, where our technological needs are powered by the wind & sun.

Sounds nice, but as we've just shown, right now we can't afford the storage, or at least can only do so in certain places.

Is there a perfect storage solution? A golden goose that can make all our wind-driven fantasies come true?

Some say there might be, and that the answer may lie in the lightest element of all…

10: The Infinity Fuel.

Hydrogen.

The maker of the cosmos. Star fuel. An unwise choice for airships. Hydrogen is many things but also two more: Rare on planet Earth, and a potential energy storage medium.

It has some appealing features, for certain: Though it can’t hold energy in as compact a manner as kerosene, compressed or cryogenic hydrogen still has an energy density many, many times greater than any battery in existence. It can be produced anywhere there is electricity for electrolysis, stored wherever you like and used either in fuel cells to create electricity, or else fed directly into a gas turbine as a source of motive power. 

This a least explains the bubble of excitement around hydrogen, which has risen as fast as the super-lightweight gas itself and created a smorgasbord of odd creations; experimental train & bus routes running on hydrogen, a couple of doomed hydrogen fuel cell car designs and an incipient boom in new regional transport aircraft with compact fuel cells and high power density electric motors. 

It has also caused politicians like Ed Milliband and Ireland’s own Eamon Ryan, equally infuriating in their own special way, to repeat the phrase “hydrogen economy” so frequently you’d swear they were trying to will it into being with words alone.

Which is pretty much exactly what they’re trying to do.

But the hydrogen economy has its costs as well, and the price of entry is the shirt on your back. 

The total expense per kWh of hydrogen storage is atrocious, and that's not because of the capital costs: Electrolyzers aren't cheap but they're not all that brutal across their lifetime. Instead, the main expense with hydrogen is simply down to its  atrocious round-trip efficiency.

While the creation of hydrogen from electricity is about 80% efficient or thereabouts, you still need to store it and then turn it back into usable juice again, and just as with thermal storage this is where you lose out. A single cycle open gas turbine can do this at about 35%-40% thermal efficiency, a fuel cell will manage 50% and a big combined cycle gas powerplant, if optimised at some expense for hydrogen, might manage mid-60s, but that's an awful lot of commitment right there. Lots of capital and lots of efficiency loss throughout the process.

So with hydrogen storage of electricity you end up with a pitiful 40%-50% round trip, meaning you'd be as well off squeezing air instead.

Let's be honest with ourselves: In the electricity market, the hydrogen economy is a nonsense. It may have some potential, perhaps, in creating clean industrial hydrogen feedstock for chemicals or fertilizers, but as a way to store electricity it's hopeless.

With some teasing, it might be useful to decarbonize certain forms of difficult-to-electrify transport, but that's another article.

But we shouldn't end with a downer. Grid storage is a big, and growing, deal, so what's the wrap-up?

11: Conclusions and comparisons.

I drink my coffee with bleary eyes today, for the overnight gale that is still powering Ireland at the time of writing kept my kids awake through the night, and me as well.

And no doubt, as the gusts shook the house, somewhere giant flywheels spun and water was pumped uphill. The Eirgrid smart dashboard shows that today wind has powered almost 80% of Ireland, which is a feat. The remaining 20% is mostly gas power, though on many days this ratio is reversed. As the topographies of supply and demand run raggedly up & down, driven by mercurial weather, the need for grid storage becomes apparent: To make best use of what is there, smooth the mountain ridges and fill the valleys. It's becoming increasingly obvious that if we are to continue our grand civilizational gamble of power-by-nature, we will need a lot more of it.

And it needs to be cheap.

Here we can end on a forward looking note, since all things in our world are guided, molded and tainted by money: What does everything cost, and how is this changing? We can look at the next diagram as a rough guide to the future of storage, because while storage may be saintly, cash is King.

Firstly we shouldn't expect much change from pumped hydro. It's a mature,  scaled technology that won't magically get cheaper, but in the long term storage market there are likely to be some surprising new entrants.

Compressed air (CAES) storage is, for all its history, still pretty experimental. Isothermal and liquid air systems in particular are cutting edge and full of risk, though the more conventional Adiabatic CAES with thermal recuperation has potential for scale up and cost-reduction that could bring it into direct cost competition with hydro.

Probably the biggest surprise could be Sodium-Sulphur batteries, telegraphed by Abu Dhabi's recent big installation. The technology is well known but not yet fully optimized for long life and industrial scale-up, and is already pretty cheap as bulk storage goes, so keep an eye on Sodium-sulphur, for it could upend the entire energy market given time.

Lithium ion batteries will continue their onward march through the world energy economy, spurred on by the increasingly vast economies of scale provided by the electric car and consumer products industries. The entry of electric power into automotive will on its own be a catalyst for chemistry improvements and cost reduction in this fast growing area. Expect more developments like the Dogger Bank battery farm, and even at a local level in peoples’ homes, as off-grid solar becomes an increasing trend.

-And that could really be a thing. I know many professional engineers at what can be politely described as “that age” who, kids now functioning adults, have decided against the mid-life Porsche and bought solar & battery installations for their house instead. 

It could be a trend.

In other areas of the grid storage market, flywheels could be surprising winners as the up & down variation of renewable power becomes a bigger & bigger factor in keeping the lights on, and grid inertia needs providing somehow. Gas storage like hydrogen is, alas, not likely to get cheap enough fast enough, and I have my doubts whether the sector will survive absent serious subsidisation. It's noteworthy that Europe's hydrogen electrolysis plans have consistently outstripped reality in recent years as private investors balk at the fat risks and slim rewards.

Whatever. You win some, you lose some.

At the end of the day, unlike every other creature on the planet we have the ability to trade time. It's our unique superpower. From favours of food and shelter between tribal affiliates to promises made & kept, debts owed and paid, we have a remarkable ability to plan and it's what sets us aside from other animals. 

We can borrow from the future, to pay for the present.

Grid storage is just another way we're exercising this species-wide mania, to manage the passage of the one thing we can't control: Time and nature. 

Will we win? Who knows, but we'll give it a damned good try. It's in our nature.

So: Build more batteries!

Credit for images & diagrams in this article go to the excellent paper A Comprehensive Review of Energy Storage Systems: Types, Comparison, Current Scenario, Applications, Barriers and Potential Solutions, Policies, and Future Prospects by Eklas Hossain et al, 2020.

P.S: For those system and battery engineers out there, I know I missed a lot out by selfishly looking at only two battery types and not two hundred. Yours is a vast and challenging field and I salute your efforts, but this needed to be readable for us laymen, so I kept it to Lithium-Ion and Sodium-Sulphur alone.

For now.