Nuclear can be deployed a few dozen miles from urban centers, and sized to use 100% of its generation right into that urban center. Those that claim nuclear is too expensive almost always ignore that fact.
So, Solar PV rooftop with FIT arrangements = production and consumption within a mile of each other, great, but that model can't scale beyond even 10% of total grid production, need large solar plants in order to use large inverters that operate with extreme accuracy to avoid AC synchronization issues. Most anti nuclear pro solar nuts ignore this serious limitation.
The reality is solar + 2 hrs worth of local storage can greatly mitigate the AC sync issue, but even 2 hrs worth of local storage = solar solution total cost goes far more than 100% up. But it still ignores the needs of the whole grid. If 100% of households in sunny area go solar you are destroying the premise of baseload production, so you are forced to have like 4-6 hours worth of local storage, plus what are you going to do in the winter, when solar production drops hugely ?
It might be possible to have a fossil fuel grid with at least 40% baseload electricity (hydro+geothermal+nuclear+biomass).
But this study is hugely flawed, it ignores huge transmission losses.
My Brasil transmits much of its electricity demand for over a thousand miles distance, this works because that is cheap big hydro electricity.
If having a mostly solar+wind grid were practical, Hawaii would have already got rid of its very expensive oil based generators (many times more expensive even than modern peaking natural gas power plants), but the reality is you have the inverter problem threatening grid AC stability.
Germany's Energiewende is also fairly stuck, shutdown of 5 nuclear power plants offset about 50% of CO2 emission reductions, with lots of brown coal burning going up.
C'mon, look at the nuclear energy facts, instead of the environmentalists biased FUD. It's safe, it's clean, it expensive upfront, but a lot of transmission costs and losses are avoided, and people insist on comparing price/MWe generated without fully accounting for nuclear's advantage. Plus water cooler reactors is old technology that we much migrate away from, even a modern AP1000 is way more expensive to build and operate than a Russian BN800 IFR reactor. The NRC overregulation model makes GE's work on the S-PRISM walk at a snail's pace instead of at the brisk pace we need it to be.