Since this is about price, how is is price of solar, nuclear etc calculated and compared? It seems like naively looking at dollar per Wh for an intermittent source is like looking at the price of a particular flower that only grows at one time of year, which will be misleading for our purposes since we have a baseload demand that we can't really adjust. What corrections are done to adjust for this obvious bias?
It is complicated, but actual professionals have looked into this with more depth than a random internet commenter who has a weird feeling that they have been lied to for years by the news sources they trust and isn't quite ready to accept they've been conned and just Google the answer to their own question.
When you are ready for that next step try LCOE and LCOES (levelized cost of energy / energy and storage) as Google terms to find more info or click this link:
LCOE calculated for intermittent sources (solar, wind) isn't really comparable with LCOE calculated for dispatchable or baseload sources (hydro, gas, coal, nuclear) though.
LCOE is (lifetime costs)/(lifetime generation), but the value of that generation is very different between intermittent sources and dispatchable sources and the difference is very important at grid scale.
For example, solar has become quite cheap -- when the sun is shining. At midnight the cost per KWh from solar is infinite. The cost per KWh for a coal plant, however, doesn't really change based on the time of day. This distinction isn't captured by LCOE.
That distinction is important because grid per-KWh costs are a weighted average of generation costs. Mathematically, that cost correlates nearly perfectly when most of the generation is supplied by baseload generation. LCOE and actual grid per-KWh cost correlates well when most of the generation is arbitrarily dispatchable. The simple numerical average LCOE correlates poorly to average grid per-KWh cost when much of the generation is intermittent if that intermittent supply doesn't line up well with demand.
For example, from your link if you want a solar system which provides power overnight, you need to sum up the generating cost (busy charging the batteries during the day) with matching storage cost. From your link utility solar ($30-$41) plus wholesale four hour (100MW/400MWh) storage ($131-$232) results in a "windless night" cost of $161-$273 which compares unfavourably with coal ($65-$152) and very unfavourably with gas combined cycle ($45-$74).
Yes, there’s this sztrqnge slice of the population that really loves nuclear power because it’s hairy and expensive and dangerous, and they seem to have trouble believing that renewables can do the job, so they now grumble about „what if it’s dark?“ but aren’t interested in answers. It’s FUD, essentially.
It’s not that simple. Questions on base load are legitimate even if storage and having a large grid alleviate the issue. Then in the case of solar there are the issues of both land usage and recycling in the context of an increasing demand. Nuclear is not a panacea. It has very good space used to power produced ratio and little but complicated to deal with byproducts. That’s why people like it. Anyway it is highly unlikely that the problem of power production will be solved with only one mean of production.
In the US, land usage is just about negligible as a part of the cost of solar. The PV equipment you'd put on an acre of land might cost $100K. The land itself? You can get it for $1K/acre or less in many places, even in the East.
Combine with wind, which often blows more strongly at night (look at the curves at ercot.com for Texas, for example), and use storage (both short term efficient storage for diurnal leveling and lower capital cost storage like hydrogen for long term leveling/rare outage backup). The estimated cost of getting to 100% renewables this way is higher than the levelized cost, but still looks like it will be cheaper than nuclear.
It's a well understood way to evaluate cost. In short nuclear is at the very high end of the spectrum and not really dropping because (so far) of a lack of a learning effect.
Solar is on the very low end of the spectrum, as the article points out. And wind is somewhere in between. Both of those are still dropping dramatically because of learning effects. E.g. with offshore wind, bigger turbines mean more surface area and more load. There are some 20 MW models now coming online. These things are gigantic and obviously a bit capital intensive. But they also produce a large amount of power very reliably. And compared to nuclear which has gigantic capital cost, it's actually very good. And there are still plenty of ways left to find further cost reductions.
Baseload is a poorly defined notion often wielded in hand wavy style to argue that nuclear is needed. This overlooks a few important things.
1) nuclear is expensive for that too and cheaper technologies exist with learning effects of their own. Batteries were mentioned in the article for example. Another one is cables. Cables allow different regions to import and export electricity to each other. That provides base load and the average peaks and dips in supply and demand over large areas are far less dramatic than they are locally.
2) intermittency is not actually random but predictable and typically local. So are demand cycles. Weather forecasts allow grid operators to plan weeks/months in advance; mostly intermittency does not come as a surprise. Hot weather means everybody puts on their AC, high solar output, relatively little wind. That happens every year in roughly the same months. Of course air heats up and you get a lot of airflow to other areas, which will have lots of wind as a result. Weather predictions are mainly about figuring out where the low and high pressure areas are and how they move around.
3) base-load is a relatively poorly understood notion. How much for how long do we need? And how much extra would that be? Those would be reasonable questions to ask (and answer) if you are a grid operator. People assume a lot of answers to these questions that are probably false/irrational. The real answers are far less dramatic.
4) Some countries already have very high proportions of renewables on their grid. So far without causing issues. Better still, some countries with historically very unreliable grids (e.g. in the middle east), rely increasingly on renewables to fix that. That suggests, renewables actually improve resilience if you don't have enough base load.
So, it's not all that black and white. The answer isn't blindly putting your trust in nuclear to get some really expensive base load that you don't actually need. But sitting down and doing the math. A lot of people seem to be coming up with outcomes where renewables are cheap, feasible, and plenty reliable especially when you use cables and batteries as well. The sun always shines somewhere and wind is always blowing somewhere. If you simply connect those places with cables, there's not a whole lot of base load that you would need beyond that. If any at all.
