A common complaint about renewable energy is that it’s intermittent – the sun isn’t always shining and the wind isn’t always blowing.  One rarely hears the opposite charge against nuclear power – it’s generating electricity even when no one wants it in the middle of the night – but renewable intermittency is a real problem.  So people have been devising all sorts of energy storage schemes for it.  Some methods are already well-established: pumped storage (pump water uphill then let it drain back E.g. the 2.7 GW Bath County Station), compressed air (fill a salt cavern to 40 atmospheres E.g. Huntdorf),   flywheels (E.g. Beacon Power), and huge battery arrays (E.g. Xcel sodium-sulfur project).

Sure, whatever.  They’re all sensible and economic and implementable, all the usual engineering virtues.   But they lack that tinge of madness that makes for interest.   E.g. anyone can build a highway bridge across a river, but the French recently built a highway bridge, the Millau Viaduct, across an entire river valley, with the road deck 900 feet in the air:

Civil engineering madness or genius? We show, you decide

For 400M euros, the French sure got a lot more tourist attractiveness than Boston did for the $16B it spent on the Big Dig.

But anyway, here are a couple of bizarre yet vaguely plausible energy storage schemes that I’ve seen recently:

• Lakes of molten metal as giant batteries – Prof Donald Sadoway of MIT wants to build really big batteries, ones that can store gigawatt-hours, not just puny kilowatt-hours.  At that scale you can’t putz around with individual containers for your electrolytes; you need one gigantic sandwich of anode, electrolyte and cathode.   He describes one possible scheme here, which consists of a pool of antimony for an anode, covered by a pool of sodium sulfate for an electrolyte, covered by a pool of magnesium.  They’re of different densities, so they’ll keep themselves separate.  As current is put into the sandwich, magnesium ions separate out on top and antimony on the bottom, and the reverse happens when current is drawn out.  A battery big enough to supply New York City’s peak demand of 13 GW would need to be 60,000 m2, or 800′ on a side.   The whole thing has to run hot – magnesium melts at 650 C.   Sadoway says “we’re now looking at failure modes”, and given how magnesium burns on contact with oxygen, and how toxic antimony is, a fire would be a truly spectacular disaster.
  

  Discharged on left, then charging and charged

  He’s a great speaker, though, and goes over his idea in a lecture here, which I had the pleasure of seeing.

• Moving big rocks up and down – A California startup, Gravity Power, wants to implement the pumped storage concept in places that don’t have convenient mountains.  Instead, they’ll dig a narrow, deep hole and put a huge concrete piston in it.  They’ll pump water underneath it to lift it up and charge it, then let it force the water out to discharge it:
  

  Its all good until the Mole Men attack

  The hole is 6m wide by 500m (!) deep. A unit that size can store 8.5 MWh with an 8000 tonne mass.  That’s only 500 Chevy Volt batteries at 16 kWh each, so this better be cheap and better be good for a lot of cycles.One key technology they don’t have is the combined pump/generator, and that’s what they’re working on. They also better have some good piston rings to keep that plug from sinking on its own.   Also, the first earthquake would hopelessly wedge the piston, so California is not the place to build this.

• Use nukes to excavate deep caves and pump in and out – OK, no one is actually working on this, but it got quite a lot of discussion on DSquared’s blog here (see the comment thread starting at 2/07/11 by Chris Williams and Ajay).  A 1-megaton bomb set off at a depth of 1000 m to avoid surface cratering would give a chamber 160 m in diameter.  That would give it a capacity of ~ 6 GWh, about half the capacity of the major UK pumped storage facility at Dinorwig, Wales.   If you’re worried about radioactive water, set off two and pump between them.  Their host, DD, responded: “And the best bit about this plan is that nothing could possibly go wrong! I do like the idea of using nuclear weapons (which we can provide at little marginal cost, since currently the best-case outcome for them is that they get chucked away unused) to build renewable energy plants. I suspect that this would go a long way to addressing the common conservative objection to renewables Alex identified a while ago (ie, having a nuclear bomb involved in the process would make the electricity produced substantially less gay).”  Ajay then goes on to calculate that with two big caverns full of water, you could set off 1 kiloton bombs in them once a day and generate enough power to run the UK.  That’s the kind of carbon-free power that even Sen James Inhofe could get behind!

Update: Oh, and I should also mention a non-maniacal but elegant system used in the Victoria and Jubilee lines of the London Underground – the stations are about 1m higher than the tunnels.  That means that the trains rise up and slow down when they come into a station and speed up when they leave.   This is a kind of regenerative braking, and it apparently  saves 5% of the Underground’s electrical usage!   Ref: Sustainable Energy – Without the Hot Air (pg 125) by David MacKay, near the bottom.