Translation: coal, as nuclear plants are providing the base load at night already, the wind dies down at night, and natural gas and hydroelectric power can be cycled more easily than coal fired boilers. So let's calculate how much coal is burned to fuel a Tesla.
To get 60kW-H, we divide that amount by the energy content of coal (6.7 kW-H/kg) and the average efficiency of coal fired power plants of 31% to find that charging one's Tesla sedan fully requires the burning of about 29kG of coal, which will release about 90kG of carbon dioxide, or about a pound of carbon dioxide per mile driven. The Nissan Leaf and Tesla Roadster, also using about .3-.4 kW-H per mile, will have similar numbers.
For comparison's sake, this is about the same amount of carbon dioxide as would be emitted by most half ton pickups these days. And my "favorite" vehicle, the Chevy
Would it improve much if power generation moved more to natural gas? Well, with 55MJ/kg vs. coals 35MJ/kg, and a different hydrogen/carbon mix, you could get to about 9 ounces of carbon dioxide per mile for the Leaf or Tesla (a bit more than the Jetta TDI today), and about 15 ounces of carbon dioxide per mile for the Volt, about the same as my minivan or the Model T Ford, which got about 18mpg.
In short, the best an electric car can do is what compact cars have been doing since the 1940s--the Volt matching only the Model T--as (AHEM) any decent engineer familiar with the Carnot cycle could have told you.
Instead of subsidizing these boondoggles, maybe we should assess a road damage tax of $2000 or so to cover the gas taxes they're not going to be paying, as well as an environmental damage tax to mitigate the damage from all the coal they're burning.