Why we need it – and how hard it is
I mentioned in my recent IET talk that the supply of electricity to the grid is likely to become increasingly variable as wind generation grows, and because as net energy importers we are increasingly dependent on foreign energy outside of our control. We don’t worry about this problem too much with gas because we can store it relatively easily in giant gasometers. If we could store electricity too that would be a huge help in de-stressing our electricity grid (and its engineers!).
As we all learnt at school, the 2nd Law of Thermodynamics says that all forms of energy eventually end-up as heat. “High grade” energy such as electricity tends to end-up as “low grade” energy such as heat, and it’s hard to go the other way. Entropy and all that. This is one of the reasons why electricity is hard to store – unless you have access to a lot of liquid helium and some very large superconducting coils, you generally can’t store electricity as electricity. You have to convert it to some other form of energy that can be stored, and then convert it back to electricity when you want it – and energy generally gets “lost” (i.e. converted to heat) in the conversion process, there and back.
The other reason why electricity is hard to store is simply energy-density. The aforementioned superconducting coils are rather big and rather expensive. The familiar energy store which has much higher energy-density is liquid hydrocarbon fuel such as petrol (gasoline) … but we don’t have any easy ways to turn electricity into such fuels just yet.
So for now we’re stuck with the humble battery, which stores electrical energy as chemical change. At Splashpower I learned rather more about battery technology than anyone should have to, because it’s not a particularly pretty picture. Their energy-density is poor (all those batteries in your Prius don’t move it very far). Their “kWh/£” is poor too (half the cost of an EV is batteries). They wear out if you actually try to use all their capacity (your laptop battery loses much of its capacity after just a few hundred deep-discharge/charge cycles). And because batteries are already a mature technology, these numbers are only improving at about 10% per year at best. On the plus side, the electrochemical reactions within batteries can be surprisingly efficient – you can lose less than 10% in the round-trip (plus maybe the same again in the power electronics). [Factoid – did you know that parts of the charge cycle of your phone’s lithium-ion battery are actually endothermic – the battery is absorbing heat rather than producing it!].
Nevertheless, despite all of these issues, electricity storage is by no means a dead duck – simply because we need it so much.
Trouble in da ’hood
Consider your local neighbourhood. A long piece of wire runs from your nearest substation down the street, with connections off to each house (actually there are multiple phases, but that’s not relevant here). Your local DNO (Distribution Network Operator) is required to keep the voltage at the point that the wire enters your house within the limits set by EU law.
But this wire isn’t perfect – it has resistance. So when the houses are all consuming lots of electricity, those at the far end will experience a significant “droop”, i.e. their input voltage falls close to the lower limit (trust me, I know – our house is the last house in the village and our supply voltage goes all over the place). And if the homes near the substation have solar panels and it’s a sunny day with everyone out at work, then the supply voltage at that end can get dangerously close to the upper limit.
To make matters worse, big new loads are appearing in homes – an Electric Vehicle (EV) or a Heat Pump (HP) has a load equivalent to several of today’s typical houses. These can significantly drag the supply down further.
And then we have the “cluster effect”. People tend to like keeping up with the Joneses, so if you get an EV or an HP then your neighbour is more likely to as well, bringing your end of the street closer and closer to the lower limit. And the solar panels cluster too.
So what is the DNO to do? Smart Meters can act as the eyes and hands of the Smart Grid, ameliorating the problem by controlling large loads like EVs, but a) it may be eight or more years before your street gets Smart Meters, and b) that doesn’t help the situation with lots of solar panels. So they’re faced with upgrading the substation (not cheap) and digging up the road to install thicker cables (not cheap). How much could this cost? Maybe £500k-£1m for a suburban road (your mileage may vary). Ouch.
Could a little bit of electricity storage help smooth out the problems? DNO’s and others are starting to wonder. Perhaps if they just put a little local storage (a bank of batteries) in or near a couple of the houses, perhaps just at the far end, they could ride-out the worst of the peaks & troughs, and provide enough smoothing to avoid having to dig up the streets, at perhaps 1/10th of the cost or less. And of course a “smoothed” neighbourhood is a grid-friendly neighbourhood too.
It’s early days, but certainly something to watch.
The above approach is limited in scale simply by the cost of the batteries. If we want to store a lot more electricity, it’s just too expensive. Is there anything alternative?
A recent spinout from Cambridge called Isentropic thinks there is.They are still some way from proving it at scale but it looks promising to me. They have developed a heat pump which is highly reversible. Off-peak electricity is used to pump heat between two gravel pits, making one very hot (500C) and the other very cold (-160C). Then when you want electricity at peak times, you throw the whole shebang into reverse, applying this large temperature differential across the pump, driving it to generate electricity.
Despite everything I said about entropy, Isentropic claim this can be done really quite efficiently – with round-trip efficiency of the order of 80%, which is comparable to the Dynorwig pumped-hydro storage facility in Wales. The advantage being that of course you don’t need a mountain nearby – particularly useful here in fenland Cambridge! If 80% doesn’t sound too wonderful, then consider that even domestically off-peak electricity only costs about one quarter what on-peak electricity does – and the spot pricing of electricity on the grid goes through even bigger swings.
Isentropic’s cunning plan centres around how to achieve this efficiency, which like any heat engine is determined by heat difference. When you’re generating, the larger the heat difference between the gravel pits, the more efficient. So the main trick to keep the efficiency high is to ensure that a clean “front” of heat moves through each gravel pit.
Consider the situation where you only get to charge the pits by 1/10th before you want to get the electricity back out again. As you put energy in, rather than heating all the gravel up just by 1/10th, you heat the first 1/10th of one pit to maximum temperature, and cool the first 1/10th of the other pit to minimum temperature. Then when you reverse the system you get a really large temperature difference which drives your heat pump efficiently.
What kind of scale will it work on? Isentropic believe their technology is probably best suited to substation-scale applications. The big advantage of storing electricity locally in this way, rather than up a single big mountain in Wales, is that it stresses the grid much less. Hot stuff.