[coldpie]

I admit this is a poor and hand-wavey response, but I'll try anyway. If we agree that solar+storage is off the table, then the question is what should we build instead? And I would guess that the people who make these decisions do consider hydro and importing power, but still decide that nuclear is the right answer. Given they're the experts and have all the info, and I don't, I'd defer to them in deciding what the best option is given all the inputs. As an example of downsides for Canadian hydro power, I would be thinking about our current geopolitical nonsense, and also transmission losses. Perhaps nuclear is the winner when you account for those? But like I said, I don't know.

[PaulHoule]

Solar + storage has to buffer three kinds of variation:

(1) Diurnal. You need to store maybe 12 hours of production to get through the night. It's believable that this could be affordable with batteries.

(2) Seasonal. In a place like Minnesota you either need to overbuild solar panels by a factor of 3 or so, or you need a lot of storage, probably not batteries, but maybe some kind of chemical or thermal storage. Casey Handmer would point out that you could use excess energy in the summer for industrial activities but that could be easier said than done because the capital cost of a factory that runs 1/3 of the time is 3x that of one that runs all the time.

(3) Dunkelflaut. Sometimes you have a rough patch of cloudy weather and little wind, so the requirements are worse than (1).

It's rare to see credible analysis of the grid-scale cost of a solar + storage system because of (3) -- you can quote a reasonable price for batteries that will supply power "almost" all the time, but costs rise explosively as you increase "almost". With different requirements for reliability the cost of a storage-based system could be "a bit less" than "nuclear power plants built without bungling" or it could be much more. It also has to vary with your location though people talking about the subject don't seem to talk about that which contributes to people talking past each other. (In upstate NY I could care less about Arizona)

[bryanlarsen]

Credible analysis doesn't go to 100%.

https://ember-energy.org/latest-insights/solar-electricity-e...

The cheapest grid is 90-97% renewable (depending on location) in 2025. As battery prices go down, that number gets higher.

[jfengel]

If we agree that solar+storage is off the table, then the question is what should we build instead?

The answer is actually "nothing". We keep gas generators around for the winter months in extreme northern climates.

We don't have to drive fossil fuels down to zero. If we need to run fossil fuel plants 10% of the time, then we've cut 90% of our power-generation CO2. Cutting the remaining 10% is far less important than other greenhouse gas sources (transportation, concrete & steel manufacture, agriculture, etc.)

We already have all of the gas plants we need to do that job. Replacing the with nuclear is unnecessary.

If it turns out that we can build nuclear fast and cheap enough to supplement the existing zero-emission transition, so much the better. But there's no need to prioritize the last dregs of fossil fuels. Just the opposite: whatever gets rid of most of the problem, fastest, is optimal for reducing the harm from climate change.

[pedrocr]

I'm confused. Who has agreed that solar+wind is off the table? Approximately no one has effectively decided nuclear is the right answer for a long time. If the proof is in what the market is actually building, solar and wind are the winners by huge margins.

What's commonly done in these arguments, and you did some of that, is declare that from first principles nuclear is the solution and we aren't only doing it for other reasons. Yet while there are plenty of simulations of doing full grids with only solar, wind and batteries there's never one where a full nuclear roll-out actually makes sense economically.

[coldpie]

> I'm confused. Who has agreed that solar+wind is off the table?

Ah okay! That's our disconnect. Do go run the numbers on how much natural gas we're burning up here. It's a lot, like seriously a lot. How many batteries will we need to ensure that amount of energy is available for (say) 2 weeks of continuous cloud cover at -10 ~ -40 degrees F? Keep in mind that if it fails, people will die. I don't feel confident enough in my own analysis to share it, but do try it out yourself for an exercise. It's pretty eye-opening.

> Yet while there are plenty of simulations of doing full grids with only solar, wind and batteries

I would love to see this! Can you share some? Do they account for converting Minnesota's heating needs from natural gas?

[pedrocr]

You're again talking about simulating only Minnesota I suspect. If you want a realistic simulation there are others in the thread and RethinkX has had a whole-US simulation for a long time. What I've never seen is a nuclear roll-out simulation that argues that's a good option. Do you have one of those?

[coldpie]

> What I've never seen is a nuclear roll-out simulation that argues that's a good option. Do you have one of those?

I don't know what a "nuclear roll-out simulation" is, exactly. As stated earlier, my position is that we should be building both nuclear and renewables. We should build whatever makes sense for the area in question. If renewable+storage can solve all of an area's needs, then that's fantastic and we should absolutely do that.

If I understand right, you are arguing we should not be building any nuclear, even in Minnesota. I'm unconvinced that renewables+storage alone can solve the Minnesota winter problem. I'm asking if you can provide a link to an analysis showing that we can feasibly and cost-effectively solve the Minnesota winter problem without any nuclear power. Can you please link to one?

[pedrocr]

> I don't know what a "nuclear roll-out simulation" is, exactly.

Any simulation where building nuclear power plants makes economic sense would do.

> I'm unconvinced that renewables+storage alone can solve the Minnesota winter problem.

You're again asking for simulations about Minnesota specifically which doesn't make sense. Unless you're thinking of seceding from the union and closing the borders to energy trade, as long as the US as a whole can do it Minnesota in particular can be a net energy importer in winter if that's what's needed. Here's the RethinkX simulation of that:

https://www.tonyseba.com/wp-content/uploads/2020/11/Rethinki...

"Our analysis makes severely constraining assumptions, and by extrapolating our results from California, Texas, and New England to the entire country we find that the continental United States as a whole could achieve 100% clean electricity from solar PV, onshore wind power, and lithium-ion batteries by 2030 for a capital investment of less than $2 trillion, with an average system electricity cost nationwide of under 3 cents per kilowatt-hour if 50% or more of the system’s super power is utilized."

This is almost 5 years old at this point. Others have linked other such analysis. At this point asking people to show them simulations for renewables while trying to argue for nuclear is disingenuous. Renewables are the ones being built out at scale all over the world while nuclear struggles to deliver new projects and doesn't seem to have a viable path to being cheap.

[coldpie]

> You're again asking for simulations about Minnesota specifically which doesn't make sense.

No I'm not, I have no idea how you are getting that idea. I'm asking for an analysis showing that Minnesota's winter needs can be met without building nuclear plants. That's it. You can solve that problem in any way you like, including importing power from other states and nations.

> Here's the RethinkX simulation of that

Thanks for the link. I focused on the New England scenario, as it's the most similar to Minnesota of the 3 scenarios. It doesn't seem to account for heating. This is the problem I keep coming to in these analyses. See page 25:

> Our model takes as inputs each region’s historical hourly electricity demand ... For the New England region, our analysis applies to the ISO New England (ISO-NE) service area which provides 100% of grid-scale electricity generation for the states of Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont.

Our heating is not supplied by electricity. I definitely believe that our current electricity demand may be met by renewables in a feasible timescale, but that leaves out the massive hole of heating our buildings.

The only reference I could find to New England's heating is this little note at the bottom of page 46:

> If New England chose to invest in an additional 20% in its 100% SWB system, for example, then the super power output could be used to replace most fossil fuel use in the residential and road transportation sectors combined (assuming electrification of vehicles and heating).

But I don't see any actual numerical analysis backing this up. Given their analysis earlier only spoke about electricity usage, I'm not super convinced by this one sentence.

Additionally, the New England scenario suggests they need 1,232 GWh of storage to supply only 89 hours of electricity for the area. Even if we agree that's a sufficient amount of time, the currently largest energy storage facility on the planet is only 3 GWh[1]. We would need 410 such facilities for New England alone. Can we really scale battery tech up that much, especially given resource constraints like Lithium and copper? Maybe! Hopefully! But it's a big question. Meanwhile, nuclear is here now, and it works. I don't think we should be betting our future on unproven tech.

[1] https://electrek.co/2023/08/03/worlds-largest-battery-storag...