Dr Bunsen:we need to remember that all that power to charge the batteries will still need to come from somewhere, the existing grid is close to capacity and has massive losses between generation and distribution (up to 60% apparantly)
I've seen this figure bandied about a few times and it's incorrect in context. Transmission and distribution is c90% efficient, and that's the only figure that's worth remembering in this instance.
Generating efficiencies vary between 90% for a wind turbine to 30% for a nuke (gas 50-60%; coal 35-45%), but we're looking to displace oil as an energy source so generating efficiency doesn't matter, only delivered energy via the grid and the means by which that power is generated.
You also have load factor, which is the amount of time spent at rated capacity in a year, about 27% for wind turbines and 85% for a nuke.
Actual grid capacity is a funny one. At 1800hrs in winter it's close to capacity - with appropriate safety margins - but for much of the rest of the time it's nowhere near rated capacity. We will need urgent upgrades, but more to accomodate new generating technology and heating tech in homes as our natural gas runs out.
I'll repeat my Nazi Sharks link since it's proved a very popular paper and has some more information http://www.claverton-energy.com/download/289/ the scenario it contains requires only a 30% increase in electricty requirement and if charging was timed correctly during a 24hr period I don't believe any grid upgrades would be necessary.
Regards