Large scale energy storage options

Cheap energy storage is the holy grail which will enable roll-out of wind and solar energy - so we can store the excess electric power when we have too much and use it later. Recently there has been news of some promising developments in alternative energy storage. Here is a comparison of some existing and potential technologies.


We already have some energy storage for balancing the grid, using pumped storage hydroelectric dams. However it takes a surprising amount of water to store a useful amount of energy, so we have to keep coal power stations idling on standby as well.

The average household uses about 10kWh of electricity/day (that's just the electricity). If your loft is 6m above the ground level, you would need to store 600 tonnes of water there to provide 1 days worth of energy (1). 50kg of Lion battery would be more practical but still very expensive.

The largest pumped storage facility we have at the moment is Dinorwig in Wales. It can generate 288 MW for 7 hours (2) so the overall storage capacity is about 2 GWh. For comparison, the UK currently uses about 1000 GWh/day (of electricity alone - overall energy use is 5 times that). Unfortunately we don't have suitable sites for another few hundred Dinorwigs.

Type	Energy density kWh/m3	Notes
Petrol	9100	For comparison only - we have no sensible way (yet) of manufacturing this from electricity.
Hydrogen at 200 atm	500	Containing hydrogen at 200 atmospheres is doable - but it isn't cheap. It is easy to make hydrogen from electricity but by the time you have turned it back to electricity you will have lost about 60% of the energy. Alternatively we could distribute hydrogen through the existing gas mains but if this dilutes the methane significantly all our gas boilers and cookers would have to be adjusted to allow for the difference in energy density.
Lion batteries	300	Based on 0.18 kWh/kg from (3) and densities from a car battery catalogue (4). Scaling this up will be expensive - unless we all have electric cars, keep them plugged in when not in use and allow the power companies to use the batteries for short term storage. 10 million cars each with a 20kWh battery could store 200 GWh.
Compressed air	30	This is a new one being developed by SustainX (www.sustainx.com). The 30 kWh/m3 is a theoretical maximum (from wikipedia but I checked it) based on ideal isothermal compression/expansion from 1 to 200 atmospheres. SustainX claim they have about 90% efficiency. That is, if 1.0 units of electricity is used in compression they can get 0.9 units back later. There is some fancy engineering involved and, as with the hydrogen, storing the compressed gas won't be cheap. Adiabatic compression would give you a better energy density but it generates a lot of heat which you also have to store. The practical efficiency would be much less - perhaps 70%.
Gravel hot/cold sinks	30	Another new one, this time from Isentropic. The energy is stored by heating up 1 vat of gravel to 500C and simultaneously cooling another one to -160C. They claim better than 80% efficiency. Two tanks 7m diameter and 7m high could store 16 MWh.
Methane gas at mains pressure, just over 1 atmosphere.	10	Not a great density but gas holders are cheap to build. Something 25m high and 35m diameter could store 24,000 m3 or 200 MWh. Also there are depleted gas fields in the North Sea which could be useful. Methane can be generated from electricity, though the efficiency is only about 60%. First you have to split water into hydrogen and oxygen, for example by electrolysis. Then you have to chemically reform your hydrogen with carbon dioxide into methane. This takes high pressure and temperature. A demonstrator is operating in Stuttgart (6). Methane can be distributed and used as is, without converting it back to electricity. However, if you do need electricity the overall efficiency will be only 30%.
Hydrogen at mains pressure.	2.5	Storage at 1 atm is easier than at 200 atm though not quite so easy as with methane, because hydrogen reacts with metals and the molecules being small and very light they leak more easily.
Flywheels	0.6	This is based on a product from Vycon for regenerative braking on trains (7). It stores 1.7kWh and weighs 3 tonnes. I believe most of that is armour just in case the flywheel explodes. It runs at up to 9000 rpm.
Hydroelectric	0.2	This is the theoretical maximum based on a 70m head, as at Diorwig.

There is no clear winner and we will doubtless end up needing a range of solutions. My instinct favours the gravel storage over the compressed air system, but it all depends on the relative costs. Both are more efficient and have higher energy density than methane but will doubtless be more expensive. I can't help thinking the methane gas option has a lot going for it, since methane is so easy to store and distribute.

(1) Potential energy is mass.g.height. 10 kWh is 10x3600 kJ. So 10x3600/9.81/6 = 612 tonnes
(2) First Hydro Company
(3) Battery University
(4) Euro Car Parts
(5) Guardian News
(6) Got Powered
(7) Vycon Energy