If building up the grid to support 500 million electric cars only costs trillions, it sounds like an order of magnitude easier than the batteries.
Batteries are not a $30 trillion and 30 year problem, not even remotely close. Batteries are an order of a magnitude smaller of a problem.
We can build the Tesla S already. Musk is going to spend $5 billion building a plant that can produce 500,000+ car battery systems. We can already make effective battery systems for mainstream electric cars, the challenge now is to make them better and scale up manufacturing (which has already begun).
It costs $5 billion in the US just to build one major nuclear power plant.
I rarely see anybody discussing where the trillions of dollars in new spending is going to come from to boost energy supply and grid to handle 100 million electric cars in the US. It'll cost trillions just to maintain what we already have now over that time.
As for where the money comes from, doesn't that just fall out of supply and demand? As demand for electricity increases, prices will go up, funding the infrastructure improvements needed to accommodate the increased demand.
The solution may not even lie in the grid. Batteries can solve the grid problem as well as the battery problem. Give every electric car customer enough solar to cover his charging needs and a fixed storage installation to keep the power for when the car comes home, and the grid won't even know the cars are there.
I'm sure this will all be wonderfully challenging for those involved, but it doesn't look like something to worry about from the outside.
Looking forward to oil-shareholders claiming how expensive electric cars for the elites are making electricity unaffordable for the poor.