3 ms·
> A few years ago, France had 1/3 of its reactors offline at the same time for maintenance. And the wind stopped blowing several times today, the sun was only
by ComradePhil 5y ago
> A few years ago, France had 1/3 of its reactors offline at the same time for maintenance.
And the wind stopped blowing several times today, the sun was only around for a few hours.
There's a clear difference. I invite you to think about it.
> Indeed we would have to build enormous overcapacity with wind and solar to achieve 100% uptime
Except, that does not guarantee 100% uptime, not by a long shot, regardless of how much over-capacity you build. One area not having wind usually means that there is no wind in a much larger area close to that. If you have wind in another part of the globe, that is useless to you because it can't be transmitted to your place; too much transmission loss for it to be viable.
Similarly, if you don't have sun, it is very likely that your entire hemisphere doesn't. So, it is crucial to have backups that are more reliable than that... and there is nothing that is better than nuclear for this... at least as of now.
- akamaka 5y agoYes you can in theory power the entire world with renewables. The sun is always shining and the wind is blowing somewhere. Of course, doing that would be stupid because of the diminishing returns. But the same is true of nuclear power. Not even France tries to provide 100% of power with nuclear, because it would be an absurd over-investment to handle a few peaks or unexpected maintenance periods. It’s pointless to argue that renewables can’t provide 100% of power, because in theory any single source of energy could, and in practice no single source of ever should.
- q1w2 5y agoThis is a big false equivalency. Nuclear uptime is orders of magnitude better than solar or wind.
- phicoh 5y agoCapacity factor for wind at sea (North Sea at least) is typically above 40%. So nuclear can be at most a factor of 2.5 better. Given how cheap wind is, you can easily overbuild a factor of 2.5 and still be cheaper than new nuclear.
- q1w2 5y agoYou cannot make those multiplications. When there's no wind, having 2.5x the wind turbines doesn't help at all. In very general terms, adding a 9 to the reliability stats on engineering projects, usually adds an order of magnitude in the cost. But more importantly, nuclear plants are up WAY over 90% of the time. If they need to do maintenance, they do one reactor at a time in the 4-5 reactor plant. This is similar to how they do maintenance on one turbine at a time in a hydro plant.
- phicoh 5y agoNuclear, wind, solar, they all need storage. Wind and solar because sometimes there is no wind or sun. Nuclear because building out for peak capacity is unaffordable. But I responded to 'Nuclear uptime is orders of magnitude better than solar or wind'. That is simply not true. Uptime cannot be 10 times better. It could be rephrased as the downtime of nuclear is one order of magnitude smaller than wind. But even that is not clear. And certainly not 'orders of magnitude'.
- ComradePhil 5y agoThe "backup" for nuclear is another nuclear plant. It is unlikely that both will go down at once. This won't work with solar or wind becuase no sun in one location usually means no sun in a much larger area around it. Same for wind. The sun and wind much further away are of no use because it is not practical to tranmit that electricity to the location it needs to be. The existing nuclear plants have ceased production once every few years at the worst. In contrast, wind mills and solar panels stop producing electricity multiple times everyday.
- phicoh 5y agoNuclear cannot do peak power in any affordable way. So it doesn't matter how reliable nuclear is, if there is not enough power in late afternoon/early evening then the grid is broken. Given the current costs for nuclear power, you can do a lot of crazy stuff with wind and solar (like transporting it over long distances) and still be cheaper than nuclear.