[DennisP]

The cheapest option is probably to build nuclear to the level that meets minimum demand at night, and solar for everything on top of that, with just enough storage to even out the remaining discrepancies.

Storage is quite a bit more expensive than nuclear, and while the cost is dropping it has a long way to go. At the same time, new nuclear technologies like molten salt reactors could well drop the price of nuclear. For that to be a factor, we'd likely have to get more aggressive with licensing new nuclear technologies; i.e. we'd have to treat climate change with the urgency it deserves.

Here's a Lazard report on levelized cost of storage: https://www.lazard.com/perspective/lcoe2019

[cogman10]

Besides the cost problem, deployment is also a huge issue with nuclear. It takes 10->20 years before a new nuclear plant produces it's first watt hour.

I can get behind the idea of pushing for nuclear in the case of baseload and even in the case of high population density areas where wind/solar are simply not practical. But, if there is anywhere to dump a bunch of money, it is wind and solar. We can have those producing electricity within a year, easily. Even if you want to talk about manufacturing costs, those are payed back within 1->5 years. Still shorter than the timeframe to getting a new nuclear plant online.

I'm not saying this to be anti-nuclear. I think it was a great solution. I just think that solar and wind have become the better solutions (at least for the shorter term).

[DennisP]

I agree nuclear is more of a long-term solution. Solar is the way to go right now, with nuclear coming online as solar gets to such high market penetration that storage is a serious issue. Until then, we can back solar with the remaining fossil on the grid.

The deployment time is a solvable problem though. We've built nuclear faster in the past, and some places still do today. France converted to 80% nuclear in 20 years, South Korea has recently built modern reactors in about five years [1], and some new designs can be mass-produced in factories. Thorcon is working on building molten salt reactors in shipyards, at massive scale [2].

[1] https://www.forbes.com/sites/michaelshellenberger/2018/06/21...

[2] http://thorconpower.com/production/

[cogman10]

Guess I don't disagree :)

I do thing there is some peak amount of producible renewables that we aren't really anywhere near. New nuclear would work well to decrease the amount of new renewable build-out required year over year.

Particularly, I think nuclear is a great option for islands (Hawaii, for example) where land is at a premium anyways.

[g_sch]

This is a great point - solar/wind and nuclear are, as the grid currently works, not drop-in replacements to each other, but complements. Nuclear provides base load (because it's difficult and slow to ramp a nuclear plant up or down), Wind/solar provides as much daytime load as practical, and meeting any remaining demand (especially peak evening demand) will require a peaking source that can dynamically be ramped up or down to meet demand.

The optimist in me believes that we'll eventually get good enough at solving the unit commitment problem in the energy grid that we'll reduce the need for carbon-intensive peaking sources and eventually even eliminate nuclear power. But we're not there today.

[08-15]

Nuclear ramps just fine. French nuclear plants reportedly ramp routinely between about 50-100% capacity, so this isn't just theoretical. Going from full power to less than about 40% tends to run you into a Xenon pit, which causes a two day shutdown, but fortunately nighttime load is roughly half of daytime load, so this point is moot.

[DennisP]

Here's a source to back up that claim: https://www.neimagazine.com/features/featureload-following-c...

[cogman10]

They all suffer from the same problem, none of them are ramping sources. You need a peaker plant for all of them (though, you'll need more for solar and wind).

Storage is a problem that simply needs to be solved for any green solution. Nuclear and wind included.

A big issue is that solar, wind, and nuclear simply don't play well together. Nuclear wants to be a base load, solar and wind push down the maximum base load feasible.

Without storage to smooth out the demand curve, you can't efficiently operate either.

[DennisP]

The question isn't whether you'll need storage at all. It's whether it's cheaper to build enough storage for windless nights, or baseload nuclear and enough storage for remaining demand discrepancies.

I think the latter is more feasible without too much wind, since solar predictably goes to zero when demand is lowest. I don't think it's at all clear that a grid with high amounts of wind instead of nuclear would be cheaper.

[cogman10]

Yeah, hard to tell given the dramatic drop in storage costs.

Today, I agree that if we could make the switch overnight that solar + storage would probably be a lot more expensive than nuclear + storage.

However, who knows in 10 years. I could see liquid metal batteries becoming extremely economical.