@Lats @ajsadauskas @australianpolitics
Well, right now there is much more derailing of nuclear in the hope of solving storage than derailing solar+wind in the hope of re-enacting a nuclear buildup (like in France, Japan, Germany (1970s-80s), Ontario, China, India…) going on.
Get both on the road, they do not much compete for resources. It will be faster than only one.
@Lats @ajsadauskas @australianpolitics The problem as I see it is that solar+wind+storage alone will not get you there ever. It will go up to 40% solar+wind, then maybe 10—30% with storage+solar+wind (depending on your technooptimism). And then you start replacing everything built every 20 to 30 years. Buys time, but not sustainable.
What you say is true: you need to build up the entire nuclear industry. International cooperation for bootstrapping will be important. Better get started.
@planet @clojure That link seems broken, even if it has a real date. But this one seems to work: https://xtdb.com/blog/dev-diary-feb-24
@ajsadauskas @australianpolitics
What would »grid scale solar & storage« cost, and how long would it take?
This is the competition:
No one in the whole world has ever built (2). There is no mature industry, and no technology even matching the only grid scale storage we have so far (pumped hydro).
For (1), there are several international players with established designs.
I wouldn’t stop either one.
@kuna I like the irony of it being a webp.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
I think you do not realize how much of our population only exists because of Haber and Bosch.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
Sorry, but the term »degrowth« is a red flag for me.
Sure, we are getting more efficient over time. That’s why even Germany’s emissions fell over the last two decades.
But cutting power that is actually needed means poverty, and that will immediately end support for long-term thinking as well as severely limit our technical options.
There are too many people for romantic visions of rural self-sufficiency.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
Yes, but I’d like to add that we need to think about lifetimes.
Let’s imagine having built all we need in 30 years, through sometimes extreme efforts.
Current solar panels, wind turbines, and batteries have a lifetime of (a bit generously) 30 years. So we’d have to immediately start again with the entire effort just to keep it up. I’m worrying that this might not be … sustainable.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
https://en.wikipedia.org/wiki/Load-following_power_plant#Nuclear_power_plants
For a grid of 100 GW peak demand, you either need
- 100 GW nuclear plants, or
- 100 GW storage output, plus (100 GW × storage loss factor) storage input (volatiles or whatever), plus additional transmission capabilities, or
- a combination of 60% nuclear plus, say 10% hydro, plus 30% volatiles
I’d say some variation on the last looks most plausible to me.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
You seem to argue that our /current/ fossil grid would also need more storage, but it works just fine as is. Nuclear is better at load following than fossils, so what gives?
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
Again, £50 per MWh is at current penetration levels of volatiles. This doesn’t scale linearly.
See that you get to more-of-the-same-kind nuclear reactors. This does.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
Nuclear is faster at load following than everything but pumped hydro and (very dirty) gas peakers. It was even a design requirement for the german Konvoi type in the 70s and 80s.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
At least Germany never had subsidies for commercial nuclear power.
On the other hand, »renewables« are still subsidized heavily, and there is much moaning right now because the build-out is slowing down, as the best places are taken.
And France has no /real/ problem with its riverside plants. Last year (much bemoaned) had 0.05% (one twentieth of a percent) curtailing for river temperatures.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
If you don’t have power output from storage equal to *PEAK* demand, it’s the same argument for any storage. And storage doesn’t /produce/ energy, it /consumes/ it (because of conversion losses, which are significant).
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
Ah, but historically, France is not an outlier. Here are the largest 10-year deployments of clean energy sources. The green ones are nuclear.
Nuclear doesn’t take long.
Here is an overview of historic build times.
The task is not fearing we might get a bad case, but creating an environment in which we get a good one.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
And again, nuclear can load follow /just fine/.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
Anyway, I don’t want anyone to stop building renewables, but I don’t want anyone to stop building nuclear either. We need every option.
(Even if nuclear is a safer bet.)
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
You seem to assume that only one reactor will be built at a time, and nothing learned. But that’s not how you do it, and not how France already did it, obviously.
I have a little problem understanding how one can acknowledge the success of the Messmer plan and at the same time claim it unrepeatable.
@matthewtoad43 @MattMastodon @BrianSmith950 @Pampa @AlexisFR @Wirrvogel @Sodis
There are already single events of more than a few hours where sunshine and wind are lacking. But that is only the immediate perspective; you need to integrate over at least several years to see the longer-term shortages that need to be handled as well. And that is quite obviously much more than a few hours. Therefore, I have some problems regarding such studies as credible.
@planet @clojure I am a bit miffed that #CommonLisp is not mentioned at all. It would fit into the article on the measure of both market share and support for functional programming.