[Schroedingersat]

You're excessively fixated on nameplate capacity. Compare net watts to net watts, or more accurately joules of radiative forcing avoided per dollar. You're also triple counting those inefficiencies.

The reality is real net power from real utility PV plants is being sold without subsidy, sweetheart loans, or unlimited publicly funded insurance for as little as $15/MWh in some places. A tenth of the cost something like Vogtle or Hinkley C needs to break even. Even at mid latitudes (north of 90% of the world's population) it's a third of the price. Include all the failed projects, or the public holding the bag for underfunded decommissioning and it's even worse.

Nuclear isn't even free of the need for storage and backup. It is capacity limited, so you can't even provide for variable demand without storage or paying double again for your already absurdly overpriced power. And it's not all that reliable -- stations in france are approaching similar capacity factors to new offshore wind -- AND the failures tend to be correlated which is a huge issue.

Also why would you need 24/7 power to serve EVs? EVs ARE storage. Weeks of it for many people.

[robomartin]

> You're excessively fixated on nameplate capacity.

Sorry, my perspective is precisely the opposite. Nameplate capacity is a farce --I have said this much-- because it is only valid under ideal laboratory test conditions. Solar zealots are the ones who use nameplate ratings, or worse, solar radiation per square meter, to justify solar fantasies. Building and actually looking at the data from my system (something most solar panel owners don't do) delivered an education I probably could not have gotten any other way. If anything, it made me think and eventually decide I needed to to understand it the way I do any other engineering project I approach.

> Nuclear isn't even free of the need for storage and backup.

These are not problems.

> Also why would you need 24/7 power to serve EVs?

You'd have to model this in order to understand it. Beyond a certain threshold or concentration of EV's in an area, you eventually get to a situation where you have a massive number of vehicles plugged into the grid 24/7. That's the simplest way I can put it.

Here's a simple attempt to show the mechanism at play:

https://i.imgur.com/arNVRea.png

Again, super simplified. The idea is you have 25 vehicles, all charging for 12 hours. The charge start time is staggered by 1 hour. The bottom line shows how many cars are charging simultaneously.

What this shows is that you eventually get to a peak simultaneous charge requirement that will remain pretty much constant 24/7.

What if you rapid charge in 15 minutes, or 1 hour?

Well, sure, the number of simultaneously charging vehicles will be reduced, however, the power and energy requirements will not. In fact, due to efficiencies and losses one might very well require more power under such scenarios. Power is the killer (not energy) because it has to be delivered instantaneously.

The other thing this oversimplified illustration does not show is a distribution of vehicles of different types (from motorcycles to semi trucks) requiring more or less power, different charge durations, usage patterns (delivery van vs. working from home and barely driving) and varied power and energy demands from the grid. Including that requires writing a reasonably detailed simulation with hundreds of variables, which is what I did years ago in order to try and understand the relationship between EV's, power and energy.

In short, we need to create a complete doubling of our entire power generation and distribution system. In some cases, more than double. We need at least 1200 GW of power; which is equivalent to 1200 nuclear power plants (this should give anyone pause and a real sense of proportion). My model predicted a range between 900 GW to 1400 GW. I believe this range represents a confidence of 95%. In other words, the real answer is in there.

It is amazing to me that this isn't front-and-center in the national discourse. EV's at scale will not happen without the equivalent of about 1200 new nuclear power plants being built and the power distribution system adapting to delivery that power.

[nostrademons]

Your model is simplistic. No sane electrical engineer would build a system where EVs charge on a uniform schedule evenly distributed across all 24 hours.

Rather, you can build a system where EV chargers are basically the inverse of natural gas peaker power plants. They turn on when grid supply is high relative to demand (and hence get off-peak rates), and then turn off when grid demand is high. Most peoples' EVs would charge while they're plugged in at work; the few stragglers would charge late at night when everyone's lights are off. Unless you are about to make a cross-country trip overnight, there is no reason to charge an EV during peak hours between 5-9 PM.

That's what GP is alluding to when they say EVs are storage.

Technology for this is already being worked on:

https://www.sciencedirect.com/science/article/abs/pii/S03787...

https://www.energy.gov/eere/evgrid-assist-accelerating-trans...

[robomartin]

>Your model is simplistic. No sane electrical engineer would build a system where EVs charge on a uniform schedule evenly distributed across all 24 hours.

Unbelievable.

Your comments are pointless, insulting. Now you are calling me insane.

My findings were confirmed by none other than Elon Musk --who I had the pleasure of working with for two years, BTW--. Here he is, confirming what I have stated when asked the question during a conference:

https://www.youtube.com/watch?v=TcI6FaaDp8g&t=3510s

Regarding my simplistic model:

I went out of my way to point out that this model was "super simplified" and an "oversimplified illustration". We use oversimplified examples to illustrate points, not to provide precise models of reality.

I further explained that creating a proper model "requires writing a reasonably detailed simulation with hundreds of variables"

And then I said "which is what I did years ago in order to try and understand the relationship between EV's, power and energy."

Please. Pretty please. With sugar on top. Stop insulting people with zero-effort, zero-knowledge comments. Instead, try asking questions, engaging in constructive dialog and, perhaps, learning a thing or two.

You are arguing against conclusions that were confirmed by Elon Musk. Not sure what else to say other than, you might want to reflect on just how little of this topic might actually understand.

This is why we go nowhere with this and other issues.

On top of that you post nonsense links. What do you think you are going to accomplish, other than confirming you truly have no understanding of the problem?

We have a 1200 GW deficit in power generation and distribution. No amount of intelligent management is going to make up for this. Not one bit. Not enough. Not even close. This isn't a rounding error, this is a serious hole in future power capacity; a precondition for EV deployment

That's what you need to grok before continuing to post comments anyone with a reasonable command of the subject would easily classify as noise.

[Schroedingersat]

> Your comments are pointless, insulting. Now you are calling me insane.

> My findings were confirmed by none other than Elon Musk

> We have a 1200 GW deficit in power generation and distribution. No amount of intelligent management is going to make up for this. Not one bit. Not enough. Not even close. This isn't a rounding error, this is a serious hole in future power capacity; a precondition for EV deployment

There are a few possibilities as to how you could get to 1200GW

You could have gotten there by looking at the energy content of every gram of oil that enters the US and assuming it would all need to be replaced with electricity. Which would indicate that you at least failed to understand what a combustion engine is.

You could have gotten there by assuming that every EV is a tesla and needs to be fully charged every day. In which case you have at least failed to understand what an EV is.

You could have gotten there by assuming around 10kWh/day per car, but assuming that peak demand would be 10x average. In which case by proposing enough idle capacity to meet said demand will be provided by nuclear power you are assuming people will be happy to pay $1.50/kWh wholesale for their electricity and $150 to fill their teslas.

Or finally you could be intentionally spreading disinformation about renewables.

[Schroedingersat]

> Sorry, my perspective is precisely the opposite. Nameplate capacity is a farce --I have said this much-- because it is only valid under ideal laboratory test conditions. Solar zealots are the ones who use nameplate ratings, or worse, solar radiation per square meter, to justify solar fantasies. Building and actually looking at the data from my system (something most solar panel owners don't do) delivered an education I probably could not have gotten any other way. If anything, it made me think and eventually decide I needed to to understand it the way I do any other engineering project I approach.

This sounds like pretty big fixation to me. I'm sorry you had unrealistic expectations for your solar install, but that doesn't change the inviability of nuclear. Also did you think to get a system with bypass diodes or are you also suffering from reduced efficiency of the whole system during partial shade?

> These are not problems.

Then why are they suddenly problems when renewables are involved?

> You'd have to model this in order to understand it. Beyond a certain threshold or concentration of EV's in an area, you eventually get to a situation where you have a massive number of vehicles plugged into the grid 24/7. That's the simplest way I can put it.

Charging things in a stupid way far more than they need it is stupid. News at 11.

Simply charge whichever EVs are stationary and not full, wherever they happen to be, whenever there is surplus power (4 extension cords, 4 transformers and 4 metal boxes per person is hardly a big investment compared to $40-200k of nuclear reactors to meet peak demand so they can all charge at once at 5pm). This is one of the few problems that is actually very simple to solve with markets (put a price on charging outside of the hours with approximately free solar power).

Mean driving distance is about 30 miles. With a reasonably efficient EV this is about 7kWh/d or 350GW if it happens only when solar electrickty is cheap.

Why would you spend $12 trillion on this problem when $1 trillion of solar, wind and storage would solve it (and this will halve or better before your first nuclear plant comes online)?

Additionally you can solve it from the other end. Forcing people to drive monster trucks 30 miles a day is an intentional policy decision. If you stop forcing the issue it will correct itself. If you put some of those $12 trillion into decent infrastructure, driving will halve or better. Even throwing LEVs into the mix for any family's second+ vehicle reduces that 7kWh/day to around 2.

Here's another simple model to play with (only uniform demand unfortunately). https://energy.model

Main caveat for somewhere like the US is it will aggregate weather over the entire country without considering problems like interfacing with texas. Maybe consider a smaller country with similar weather to get a more realistic estimate. Compare US nuclear capital costs ($10-12 per nameplate watt) to the 80c/nameplate watt non tilting or $1.3/nameplate watt tilting of recent US projects, or about half that for projects still in the permiting phase.

[robomartin]

> 4 extension cords, 4 transformers and 4 metal boxes per person

Oh, please.

[Schroedingersat]

You don't need fast charge for every joule. A regular 20A cord (or 2A 1kV to reduce copper investment) can keep a car topped up just fine. Then the only issue is producing enough net watts per week (which dovetails excellently with variable production)

If we correct our insane legal framework around the use of streets then a vehicle which consumes 100W average when in use, charges itself with a single 300W panel and does 20mph covers 95% of uses.

Either option is vastly more viable than a multi trillion dollar handout for infrastructure to help sell luxury cars.