One plant should not be judged on its intermittency alone. Its just not how the system works at a connected-grid scale.
So we can store our storage centers closer to population centers, but we can't generate it close by. We now have lots of transmission losses and opportunities for failures.
The problem with all this is that it is exceedingly complicated and most people are trying to simplify the problem. But here even including second order factors doesn't give you good approximations of the solution.
While costs of transmission infrastructure required for country scale (larger distances for lower correlation) energy dispatch are recognized, its more abstract challenges are less well acknowledged. Dispatch at this level is not just about developments in grid integration or hardware like solid state "transformers", it also has a complex routing coordination aspect requiring research in control and even game theory [1].
A lack in wind and solar can sometimes occur simultaneously. Analysis of German wind turbines data observed that experiencing a stretch of almost a week with generation as low as 10% installed capacity was likely within a given year [2]. Surprising/extreme weather events like Europe's recent "wind drought" are rare but there remains a large amount of uncertainty in how changes in climate will affect the tail of this distribution. Tools such as coordinating distributed generation and improvements in storage tech will surely help smooth generation, nuclear is another powerful tool in that toolbox.
[1] https://www.nrel.gov/docs/fy15osti/63037.pdf
[2] https://iopscience.iop.org/article/10.1088/1748-9326/ab91e9/...
https://www.sciencedirect.com/science/article/pii/S096014812...
The problem is that Nuclear may not be economical if it is only used when both solar and wind run out. Nuclear has large fixed costs. And almost zero marginal costs. So the average costs -- what needs to be charged in order to avoid bankruptcy, increase as you use it less.
That means every solar panel you add makes the nuclear power a bit more expensive. And that incentivizes adding more solar. Up until you drive the nuclear out of business, and then suddenly you don't have reliable power anymore.
Then you are faced with a situation of
a) only having nuclear power which can provide for all of your needs, in which case adding solar is an unnecessary expense
b) only having solar+wind and an unreliable grid, which means you need to add batteries to cover solar+wind. And the price of those batteries may be more than the price of the nuclear plant.
c) having nuclear and solar both, with enough subsidies given to the nuclear plant to keep it in business so that the total solution is more costly than just going with nuclear.
So yeah, there really is a tension between nuclear and solar.
This is not the situation, however, with solar and coal. Because coal plants are damn cheap, and they have higher marginal costs. Thus solar can coexist with coal or with gas much better than with nuclear.
Therefore the economics is such that as people promote solar the result is a decrease in nuclear and an increase in coal and gas.
Current electricity demand is heavily biased to daytime use even with cheap nighttime prices causing people to shift demand to use that. Start to ramp up solar to the point where daytime demand is higher and a great deal of nighttime demand drops off.
Grid storage isn’t cheap enough to store energy at current nighttime rates, but it’s cheap enough to have a balanced grid backed by hydro, wind, and solar even with zero fossil fuels. The tipping point to cheap daytime rates and expensive nighttime rates isn’t inherently better or worse, it just reflecting the future economic reality.
You are correct that my model is simplified. I ignore the issue that demand isn't really stable, and you need some peaker plants.
The problem is that Solar isn't really a good solution for peaker plants, because those need to be reliable. So I don't think this simplification undermines the tradeoffs I was describing, although I agree that in the space of peaker plants, there can be some combination of solar and gas to handle peaks when it is sunny and also when it is not sunny. Just be prepared that you need enough gas and coal to cover all the generating capacity you are getting from solar and wind, which is again very expensive.
> grid backed by hydro, wind, and solar even with zero fossil fuels
This requires a lot of hydro, more than most nations have together with really punitive electric rates when there is an absence of wind or solar. I mean, massively punitive rates, because demand for electricity is highly price inelastic. So be prepared for rates to go up 10x or 20x or even 100x when there is a stretch of windless days with weak sun. I think there is a reason why no nation has gone this route except oddballs like Iceland with their reliable geothermal.
At that point you are still going to get multiple day stretches where wind and solar only cover ~1/2 of daily demand but hydro can make up the difference on such occasions even if it’s only supplying 6.6% of annual US demand. Basically you get 1-2% hydro on most days and on 5% of days you a lot of energy stored.
As to high costs, because of the excess solar you’re generally filling batteries with nearly free electricity. Average nighttime wholesale prices therefore end up at ~10c/kWh or whatever the battery storage costs settle on, but daytime rates when most demand actually takes place are going to tank. That’s a net reduction in average prices. Trying to make a grid from Nuclear + batteries on the other hand means your paying Nuclear prices at night, but nuclear + battery prices in the daytime which is the opposite of what you want. Nuclear + fossil fuels on the other hand simply doesn’t go far enough.
Now in a mostly solar world a very low percentage of electricity may end up generated by fossil fuels, but a 99.X% solution is success by any reasonable metric.
PS: As a sanity check you can look at what people are paying when their off grid and then realize that’s very much a worst case.
Also, be careful when looking at wind and solar minimal percentages. It’s the difference between median output and minimum output that matters not maximum output. Long term it’s likely something like 30 to 50% of all solar generation is going to be wasted simply because it’s just that cheap.
In terms of fish, large dams are needed for flood control and water. But rivers and streams often have huge numbers of small dams that are equally problematic and far less useful.
Likewise we could make power stations that could withstand hurricanes, or earthquakes, or tsunamis, and so on. We don't usually though, it's too expensive (not to defend the decision not to weather proof more in this case, it wasn't really that far outside of the expected operating conditions as I understand the situation).
I’m not an energy grid expert but I guess it’s hard to have nuclear, gas wind and solar playing together because if it weren’t, people would just get on with it?