I think we'll see a significant ramp up of EVs long before we see widespread CNG/LNG adoption.
China has sub-$15k equivalent EVs today. Cheap EVs are certainly physically possible even with today's scales, and partly it's just the catch-22 perception hump of "EVs are only for early adopters and luxury cars" (for both consumers and manufacturers) keeping EVs from hitting some scale benchmarks to push them cheaper.
The used market still isn't showing a lot of EVs, but that seems to be that statistically most EV owners keep their cars for longer. That's an interesting statistic, that maybe points to reliability on one side, and a possibly very different replacement rate on the other side. Which is to say that the average EV first owner is currently keeping their vehicle for 7-10 years rather than the 4-5 year replacement rate that has been of typical of ICE vehicle sales. (If EV owners were only keeping their cars for 4-5 years, as you suggest of some hypothetical "HN bubble" that would be a lot more of them on the used car market, and in turn fewer complaints of the high cost of EVs. The long first life of an EV is also keeping the costs up and affecting depreciation rates.)
I'd love to get an EV one day, when the pricing is less nutty than it is today. But it has to be economically viable for me. Right now it's not.
And for electricity, the utilities actually need the extra demand because of energy efficiency reducing lighting costs and distributed solar and subsidized wind providing occasional gluts. Since EVs can fairly easily be set to charge (most the time) on a schedule when electricity is most abundant, this is good for everyone.
EDIT : Whoops, my bad, somehow forgot that we were talking about transportation, which is "only" a quarter of power use...
So you need to go from 24 to 24+(34-1)/4 = ~32 Gwh/year, a 33% increase, and you expect to be able to do that in 2-3 years ?!
(Napkin calculations, not taking into account various (in)efficiencies and intermittency issues, but the order of magnitude should be roughly this one...)
US average electricity is ~475 Gigawatts. But there's a huge difference between average electricity and nameplate capacity. Natural gas peaker plants have about a 200GW nameplate capacity in the US but only produce about 25GW average. If we ran just the peaker plants all-out, we'd generate an additional 175Gigawatts.
US gasoline consumption is about 400 million gallons per day. I drive a Volt, and typically I get about the same range in 10 kWh that I would in a single gallon (Model 3s do even better, however). So roughly speaking, We'd need about 4 TWh per day, or about 167 Gigawatts average extra. Technically, averaged over the year, then, our peaker plants would have enough extra capacity to produce enough electricity. But natural gas combined cycle plants would also generate about another 100GW if ran all-out. (None of this is terribly realistic as there will be local transmission constraints and weather-related demand spikes which the peaker plants are needed for, but it is instructive.)
However, it'd take a good 10-15 years to mostly turnover the whole gas car market in the US even with a revived "cash for clunkers" program, so the capacity expansion would have plenty of time to sort these things out. We can expand solar and wind over this time even while phasing out gas.
The main thing, however, is the battery production.
It's 4% of electricity coming from oil, sorry for the confusion. (Hmm, I might have made another mistake here - eh, it's smaller than the uncertainty anyway...)
167/475 = 35% Nice to see that the math seems to check out with your data too. The issue of intermittence would warrant whole books of studies, hard to do it properly in comments here...