What is it about energy storage that excites people?
There are many answers to this question, but for opponents of nuclear energy it is predominantly seen as the most promising way to overcome the intermittency and unreliability of those most photogenic of renewable energy technologies, photovoltaics and wind turbines. In fanciful majority solar/wind future scenarios, without storage to back up low production times – like windless nights or even dark, very stormy days – a magnitude of overbuild is needed which leaves much spare, idle capacity in optimal production times, else other technologies like hydropower, biomass combustion or (predominantly) natural gas turbines are unavoidable.
The first major irony is that backed-up or overbuilt renewable models try to overlap a bunch of technologies to meet baseload demand – conventionally met by just one or two main dispatchable technologies (like coal or nuclear) and erroneously dismissed as a myth. The second is that in a nation-scale storage-reliant grid, solar and wind capacity would require a comparable magnitude of overbuild anyway – to both meet demand and charge the batteries enough* on good days.
It’s not that I oppose storage – I love batteries, I’m pro-EVs, I’m excited by technological advancements and I support hydropower (as well as solar and wind, all where appropriate) in our future mix – or other measures like increased efficiency and sensible demand management. I just refuse to delude myself that storage, with its intrinsic limitations, can fully enable any clean technologies that cannot manage the decarbonisation job themselves, at the required scale and in the dictated time frame (if at all).
Tom Murphy dealt with pumped storage – still the most economical form – years ago and anyone interested in energy needs to have read this. Why is it the most economical form? All you need is suitable land, accessible water, and you build the structure once to last a century or more. As Murphy demonstrates, this is supremely, physically unachievable in a country like the US.
But let’s say a pumped storage site were commissioned to backup some considerable solar and wind capacity. The storage operator will derive income from flowing water through turbines and selling power to retailers, and the operating costs will mainly be buying power to pump that water back up to the reservoir. They will want to sell at times of high demand to maximise return on investment, and to store during low demand (cheap power), and this is exactly how such capacity is currently used. But high demand generally coincides with sunny afternoons, when PV capacity is being used directly and the price is high. Wind will blow, or not, largely regardless of demand, or lack there-of. The storage operator will be expected to pay a premium to store power during the time when it makes most sense to generate, and to hope for windy, overproducing nights. Who will compensate the operator for this forgone income? Who will invest in constructing this facility on this basis?
The most ubiquitous chemical storage technology in the world today is based on the oxidation states of lead. Lead is inexpensive, common, and the vast bulk used in batteries is economically recycled. Emerging technology like liquid metal batteries are exciting and deserve substantial support to rapidly penetrate the market and contribute as best they can, but the power is still limited by the magnitude of change in oxidation states of metals, and this can’t be increased.
That net energy release is only part of the story, but it is sufficient to illustrate the quantities we must accept. The oxidation of lead provides the equivalent of 34.35 kJ/mole, where kJ are the same metric kilojoules we consider for food, and a mole is a chemist’s way of levelising between different elements and compounds. Specifically, when coal is burned, the energy is overwhelmingly provided by the oxidation of carbon-carbon bonds, and yields 153 kJ/mole. Again, don’t directly compare these values – one is a reversible electrochemical half-reaction, the other is the energy liberated from a bond by combustion, and they are subject to vagaries and efficiencies of technology design – but instead accept that they fundamentally frame the scale at which chemical energy is stored. Tom Murphy also dealt with what happens when we want to scale this up. The point is, the fundamental coulombic interactions being manipulated through combustion or electrochemical storage are immutable and, despite likely improvements in the future, there will be and cannot be any Moore’s Law of battery storage.
And guess what – the same economic problems facing pumped hydro would also apply to operators of grid-scale battery storage, but without the extended infrastructure lifetime. Remember, pumped hydro is still the most economical form of storage.
The Best Storage of All
The rough comparison above alludes to treating coal like a form of storage – popularly, stored sunlight from millions of years ago. To the extent that this describes fossil fuels, it also describes nuclear fuels. The steady decay of 18.89 moles (4.8 kg) of plutonium-238 oxide in Curiosity’s RTG provides over 118 kJ every minute (1968W). The fissioning of uranium-235 liberates 20 000 000 000 kJ/mole, stored by the strong nuclear force in the hearts of primordial supernovae that seeded our little region of the galaxy with plentiful heavy elements billions of years ago. I can see how that image may be a bit daunting. But given nuclear power’s proven carbon-mitigating capacity, dismissing it in favour of an inferior energy storage fantasy is provably irresponsible. We can now build reactors that release the comparable energy from all U-238, and even from the plutonium from used nuclear fuel as well as dismantled weapons.
The irony of ironies is that the ultra-low carbon, dispatchable power provided by nuclear fits the economic operation of pumped hydro far better. The addition of even modest conventional storage to the fast reactors described above results in formidable clean electricity generation indeed.
I think the reality is that some significant storage will be integrated into grids (where technically feasible) of developed countries (who can afford it). It will certainly play a small-scale role in domestic systems, and odds are you know someone (who knows someone) who already operates batteries connected to rooftop PV. But that * up at the top of this article is crucial – how much is enough? If rejection of dispatchable electricity capacity – from fossil fuels as well as ultra-low emission nuclear – is the goal… rather than rapid and effective decarbonisation of reliable supply, how can anyone know how much impossible storage would be enough?