Also, don't make the assumption that every electric vehicle will require the same energy per mile as a Tesla does today. It is easy to see that something like an electric F150, Suburban, Jeep, Toyota Sienna (minivan), commercial cargo van, commercial truck or big rig will require far more energy on average and far more power to charge if charging is to happen within a reasonable period of time (30 minutes to 8 hours, depending on utilization profile).
While I do not claim my simple model to be accurate, even with a large error we still are looking at having to build a massive number of nuclear power plants in order to support this electrified ground transportation future. We can't even build one new power plant. How are we going to build 50, 100 or more?
My fear is that the hand-waving that is going on out there will result in a reality where the well-to-do will be able to drive electric cars and everyone else is going to get stuck having to drive conventional vehicles.
If we don't get serious about adding a serious amount of power generation capacity we will ultimately face an asymptotic relationship between cars and electric power. This will limit the viability of electric cars, which will limit sales, which means prices will remain high and the cost of electricity will skyrocket. On top of that, oil prices are likely to go up at the same time, which means the poor, lower-middle and even middle class will get royally shafted at every conceivable level.
And, BTW, as I said in my original post, even if I am off by a factor of ten, we need 300 nuclear power plants. I'll take your 14x suggestion and take up a notch to being off by a factor of twenty. That's 150 nuclear power plants. Someone needs to explain how that is even remotely possible.
EDIT:
This is important. Most people make the mistake of thinking about energy rather than power. This idea of filling-up one a week vs. every day is great...when what you are doing is pumping a liquid. If what you are pumping is electricity, well, there is no free lunch. The energy you did not pump into the batteries every day of the week will need to be pumped into them on Sunday. Rough numbers means you need seven times more power in order to achieve this when compared to charging a little bit every day.
As I have been saying, most people have hand-wavy conversations about this, which means they develop the wrong model for the of electric charging. You have to sit down and make an attempt at a half-way honest model of what it would be like to charge hundreds of millions of electric vehicles of different kinds under reasonable usage regimes. Some people will charge every day for eight hours to top off. Others will need to top off in half an hour. And, yes, some will charge on Sunday (or whatever). In the end, physics is physics, and the mathematical results do not get better if you charge once a week, they get worse, much worse.
This is what I come here for, and let me add in a few more confounding factors I haven't been able to chase down.
Power transmission is neither free, trivial to implement, or perfectly efficient. The more power you are moving somewhere, the more dangerous and costly the "pipe" you use to get it there. If you look at residential situations, it isn't hard to find households with >3 cars. Is everyone plugged in at once? Who gets the fast charge? Can the house handle it? Can the neighborhood handle everyone doing it at once?
What happens to residential electric? Do we start doubling up on transformers? One to keep the Electrician's codes from having to change, and one to feed auto chargers? What does that mean for linemen/women/people? How does the increase in higher power transmission lines affect the biosphere?
What happens the first time a feed or arterial link goes down? You can ship gas, electricity, not so much, especially if swappable batteries are not a thing?
How about equipment? Cars aren't the only things that use gas, Gotta count in charge maintenance and battery attrition there too as well as consumption.
What happens when the almighty Centralized Generation(TM) goes down? How decentralized is our generation going to get? How do we ensure the grid can be maintained safely without linefolks also having to go house to house to isolate them from the upstream grid they need to work on?
Electricity, despite the near magical experience of the layperson, is frigging complicated, and it's successful deployment has been built on a lot of interconnected assumptions, standards, and what have you. If the transition is slow enough, I figure we can spread answering all of these questions over the span of a generation or so; but I'm really uncomfortable with the optimism of people going "we JUST need to go electric".
Hell, all of that has to be considered on top of increased need for power for climate control, the possibility of higher temps leading to actively having to mitigate water shortages or any of the other myriad of other knock on effects nobody has wracked their noodle enough to realize may be a problem yet
I like to say that reality is a complex multivariate problem. Reducing it to a single variable --which is what you see out there nearly 100% of the time-- is, at best, a complete misrepresentation of what solving real problems in the real world looks like.
I avoided adding layers of complexity to my simple model because all I was after was a sense of the scale of the problem rather than an accurate number (not sure that's even possible).
Once you start to add power system transmission, conversion and delivery efficiencies the problem quickly grows. On average, the energy loss from power generation to delivery is in the order of 6%. This might not seem like much, yet all of that power is converted into heat. Which means that if we need to deliver 500 GW to homes, we actually need to generate 532 GW, of which 32 GW will be converted to heat.
I have no clue what this means because I don't yet have a sense of proportion on such matters. That is is a very large amount of energy being wasted goes without saying. I can't really voice an opinion beyond that other than to highlight that this is yet one of the other variables that needs to be considered.
Power conversion systems and charging circuits (in general, switch-mode power supplies) of course have losses. SMPS efficiencies (I used to design them) are in the 85 to 95% range. It's a complex combination voltage and current that produces switching losses. At higher voltages you can be more efficient, however, because we are drawing such high currents the losses quickly mount. In the end, you are going to convert another 5% to 15% of all of incoming power into heat.
So, if cars need 500 GW and chargers are, say, 90% efficient, the total power delivered to the building connections will have to account for the 6% transmission and delivery loss and 10% SMPS loss. This starts to add up fast. Now we need 591 GW and will convert 91 GW of that into heat (the aggregate efficiency being 84.6%.
I could go on. This is not a simple problem. And those who just hand-wave and say stuff like "solar and wind can handle it" can only do so because they have not devoted a single minute to doing some pretty basic math before forming --or believing-- these flawed ideas.
https://news.ycombinator.com/item?id=27978080
Also, don't make the assumption that every electric vehicle will require the same energy per mile as a Tesla does today. It is easy to see that something like an electric F150, Suburban, Jeep, Toyota Sienna (minivan), commercial cargo van, commercial truck or big rig will require far more energy on average and far more power to charge if charging is to happen within a reasonable period of time (30 minutes to 8 hours, depending on utilization profile).
While I do not claim my simple model to be accurate, even with a large error we still are looking at having to build a massive number of nuclear power plants in order to support this electrified ground transportation future. We can't even build one new power plant. How are we going to build 50, 100 or more?
My fear is that the hand-waving that is going on out there will result in a reality where the well-to-do will be able to drive electric cars and everyone else is going to get stuck having to drive conventional vehicles.
If we don't get serious about adding a serious amount of power generation capacity we will ultimately face an asymptotic relationship between cars and electric power. This will limit the viability of electric cars, which will limit sales, which means prices will remain high and the cost of electricity will skyrocket. On top of that, oil prices are likely to go up at the same time, which means the poor, lower-middle and even middle class will get royally shafted at every conceivable level.
And, BTW, as I said in my original post, even if I am off by a factor of ten, we need 300 nuclear power plants. I'll take your 14x suggestion and take up a notch to being off by a factor of twenty. That's 150 nuclear power plants. Someone needs to explain how that is even remotely possible.
EDIT:
This is important. Most people make the mistake of thinking about energy rather than power. This idea of filling-up one a week vs. every day is great...when what you are doing is pumping a liquid. If what you are pumping is electricity, well, there is no free lunch. The energy you did not pump into the batteries every day of the week will need to be pumped into them on Sunday. Rough numbers means you need seven times more power in order to achieve this when compared to charging a little bit every day.
As I have been saying, most people have hand-wavy conversations about this, which means they develop the wrong model for the of electric charging. You have to sit down and make an attempt at a half-way honest model of what it would be like to charge hundreds of millions of electric vehicles of different kinds under reasonable usage regimes. Some people will charge every day for eight hours to top off. Others will need to top off in half an hour. And, yes, some will charge on Sunday (or whatever). In the end, physics is physics, and the mathematical results do not get better if you charge once a week, they get worse, much worse.