In his post on the potential of our current grid to support electric cars, John McGrath mentioned V2G in passing.

Electric cars (either hybrids or full EVs) have the potential to be a real-life silver bullet. Anyone who advocates for increased use of renewables is inevitably confronted with the problem of intermittency. With wind, the rule of thumb is that if grid energy supplied by wind grows to more than 25-30%, utilities need to spend prohibitive amounts on “spinning reserve” to even out supply.

Well, a nation driving plug-in hybrids makes for a spinning reserve of amazing proportions according to one estimate (PDF), the U.S. fleet would power the U.S. electrical grid seven times over.

What these estimates neglect is the capital costs of the batteries themselves. The assumption seems to be that since car owners have the batteries anyway, the economics can be calculated based on operating costs — electricity and inconvenience.

But battery cell lifespan is measured in cycles — charges and discharges. If utilities start drawing on car batteries, they will shorten the lifespan of the battery cells. At current battery prices and cycles, this adds 25 to 80 cents per kWh — before electricity prices are considered!

Surprisingly, this does not make V2G uneconomic; it just limits V2G value to spinning reserve only, while ruling it out as a practical means of peaking or evening out day-to-day variability by renewable sources.

The key is that both peaking and day-to-day shaping draw on storage capacity frequently. Where the lowest cost extended batteries suitable for electric cars (NiMH) cost about $300 per kWh of capacity and last 500 cycles, for $350 per kWh you can buy large scale Vanadium flow batteries that will last 10,000 cycles. Over those 10,000 cycles you would have to buy 20 NiMH batteries (or 10 LiON at $1,000 each, or 35 lead acid cells).

However, pure spinning reserve is not taken advantage of frequently. It is drawn on in case of failure. So renting capacity infrequently, even at a very high price, makes more sense than owning capacity you almost never use.

In other words, use flow batteries (and pumped storage) for day-to-day shaping of variable renewable electricity, and to replace fossil fuel peaking plants. Use V2G to compensate for equipment failures, seasonal peaking, and unexpected shortages of renewable power beyond what such dedicated storage can compensate for.

You still won’t want to use it to compensate for extended shortages. The study cited suggests a fleet of PHEV could provide about eight hours of backup for an entire grid. But part of the fleet is on the road at any one time; you can’t draw too deeply on any car, because you compromise the ability of the owner to travel. And even before you reach that point, if you draw down an owner’s power so deeply that she starts recharging again in a few hours, you have not gained much — if the shortage is ongoing.

So V2G would only be useful as true spinning reserve — to keep the power going in certain types of emergency until a backup, off-line operating reserve can be brought on-line.

This is definitely useful. But it is not the key to making renewables fully dispatchable.

There is a fascinating side point in the V2G article that may well be that key; it proves me wrong about something I’m really happy to be proven wrong about. I will write about it in my next post.