Thread by @Jordan_W_Taylor

In 2022, short on electrical grid inertia and long on renewable power, Ireland installed the world's largest flywheel, 130 spinning tons.

Why did we do something so preposterous?

And are there other, better storage technologies? Let's find out.

It's the grid storage thread!

In this thread we'll cover all the major storage techniques and what they're good for. Be warned: There is NO perfect method.

Before we get started, the difference between power & energy:

Power (MW): How much oomph/ what can you power with this.
Energy (MWh): Power x time.

The classical use case is load shifting: Storing electricity in low demand periods and supplying it back in high demand periods: Hours, days or weeks later. These require high energy capacity, crucial for renewable-dominated grids.

Shown: Turlough Hill pumped hydro, Ireland.

Another use case is grid stabilisation: Maintaining grid stability in the face of faults, sudden load or draw changes and supplying inertia. These emphasise power delivery over energy.

An example is the synchronous condensor flywheel mentioned earlier.

For the impatient, here's how all the different storage solutions stack up against each other, in several acronym-heavy graphs.

But you want more detail than this don't you? On with the show...

Pumped Hydro .

The most popular solution globally with 150GW power & 9000 GWh energy capacity, this pumps water to an elevated reservoir when electricity is cheap and sends it back through turbines later. After frictional losses, it has a 76%-85% round trip efficiency.

It has a lot to commend it: It's affordable, can do long term storage, is fairly efficient and has a 50 year+ lifespan.

But it's volumetric energy density is very low, so to get meaningful amounts of storage you need massive installations, which is geography dependent.

Compressed Air Energy Storage (CAES)

A quirky and old fashioned storage method used for power smoothing for decades and to power mining vehicles before that, it has the advantages of pumped hydro but with a higher energy density.

But a big problem keeps it off-grid: Heating.

When you compress a gas you heat it, and the loss of energy keeps CAES systems at just 40% efficiency.

Advanced Adiabatic CAES: Compressed air is cooled by heat exchangers, storing thermal energy (e.g in crushed rock) for re-injection during expansion, for 70% efficiency.

Another solution, supercritical CAES: Air is compressed & cooled to a liquid state for cryogenic storage, and heat stored elsewhere.

Both systems allow higher efficiency & energy density, trading off complexity & lessened long term energy retention. Pilot plants are underway.

CAES and pumped hydro represent our two "bulk energy storage" solutions, adapted for large scale, long period storage: There is a 3rd but we'll get to it later.

Now let's look to another extreme: High power density, short term storage: Flywheels!

Flywheel Energy Storage (FES)

A symmetric steel rotor on magnetic bearings rotates in a partial vacuum. With similar specific power but lower specific energy than batteries, it excels in low cost of power, long life, efficiency & reliability. Good for grid stabilisation.

Capacitors & supercapacitors.

Pitiful specific energy but high specific power, capacitors have long lifespans, high efficiency, but cannot store long term.

Used sometimes in substations, these work well for power control applications but are useless for load shifting.

Superconducting Magnetic Energy Storage (SMES)

Spookily storing energy in magnetic fields, you might see these in particle accelerators or fusion reactors but never as grid storage. A thread on them is linked below.

So what about the jack of all trades, the Lithium-Ion battery?

A great technological leap, the rechargeable Li-ion battery can be modified for high or low specific power or energy depending on chemistry. Suitable for grid stabilisation and short-mid term load shifting.

It's efficient, at 85%-95%, flexible, can be built anywhere and turn it's hand to most things. It's the fastest growing grid storage globally, though not challenging King Hydro yet.

But charge degradation means it's unsuitable for seasonal storage, and it remains very expensive

Sodium-sulphur batteries (Na-S)

With electrodes of molten sodium & molten sulphur, these high temperature batteries are cheap-ish, pretty efficient (75%-90%) and can do long term storage, but are let down by a poor operating lifespan: Just a few thousand cycles.

Thermal Energy Storage (TES)

Divided into sensible heat storage (no phase change) and latent heat storage (uses a phase change), TES can be done with a variety of materials: Advanced concentrated solar plants use molten salt TES to supply electricity at night and when cloudy.

Hydrogen fuel cells (proton exchange membranes).

The great white hope of green economics, hydrogen energy storage is scalable, high energy density, high power, easily transportable and suitable for long term storage.

It has a big problem though...

... Utterly terrible round-trip efficiency, of 25%-40%. Losses not just in electrical generation, but also hydrogen production, by conventional electrolysis or using solar or nuclear process heat (thermochemical water splitting).

This inefficiency makes it niche storage only.

So how does everything stack up? In terms of long or short term, here's the breakdown. The bulk storage solutions, long period and with the heft to manage entire wind farms going down, are pumped hydro & CAES, though sodium-sulphur batteries could do it if the price is right.

At the low end the usual suspects: Capacitors, flywheels and many battery types, though flywheels can chip into the lower reaches of load shifting applications and lithium ion remains jack of all trades, constrained mainly by price.

Speaking of price...

The full levelised cost of storage holds some surprises on the bulk long term end: Pumped hydro, a mature technology, won't get cheaper but keep an eye on advanced compressed air & sodium-sulphur battery systems!

Hydrogen remains hamstrung by inefficiency.

At the short-term, current quality end, li-ion will continue it's march downwards in price and across in capability, driven by a now-colossal consumer industrial base. Flywheels, already stealthily popping up everywhere, have more room to run, but will fight with Li-ion.

The surprising reality is that not only is storage getting cheaper, and fast, but that's it's mostly mechanical, not battery driven.

There are many storage niches, see graph shown, and great profit potential, but no single technology ticks all the boxes.

Like it or not, the vast expansion of renewable power will drive a many-fold explosion in grid storage capacity worldwide, and it will be a smorgasbord of different technologies, including some genuine surprises!

It is, at least, getting cheaper.

I only included the most mature technologies here, so I'm sorry if I missed your favourite one! You can read about 47 (!) different methods in the paper shown.

But it's a long paper: Charge your batteries...

I hope you enjoyed this!