Variable renewable prices are excellently low, with solar PV falling from $350/MWh to like $43 in 10 years. The variability and massive land/material use is where I suspect they will eventually run into modest troubles at scale (around 50% of total energy). Already, in San Bernardino county (of all places!), large new solar installations in the desert were banned, largely (from what I can tell) because of NIMBYism. Thus, I personally believe that at scale variable renewables will eventually struggle, as almost all other energy systems have in the past. The exponential curve in the sub-10% range gives way to a plateau in many cases. I think there's sufficient reason to consider this, if not likely, at least imaginable. Many studies, including some from NREL, show costs of solar going absolutely vertical at certain levels of penetration, for intuitive reasons that when you're building that last solar panel that's only needed on December 15th at 1pm for 1 hour, it's pretty darn expensive. Energy storage helps, but has externalities and cost. As usual, there is no free lunch.
So, you want some ideas to make nuclear cheap? Ok! Barring what's been dubbed the Mobile Chernobyl, building large nuclear plants in huge shipyards and floating them to their location where they're either embedded on land and operated terrestrially or (even better) operated 10 km offshore is the kind of idea that can cut costs dramatically while actually improving safety. Sounds crazy, but hear me out!
Shipyards are like assembly lines. It's Henry Ford meets decarbonization. You can pump out massive components and assemble them in a controlled environment with local labor without any temporary structures or shops or roads or anything that makes heavy engineering projects expensive. You get economies of mass production and economies of scale.
When operating at sea, you are intimately coupled to an infinite heat sink that can cool you in all situations. If you're in deep water, tsunamis are long wavelength and non-threatening. Earthquakes don't exist. Weather can hurt, but the Prelude was designed to operate in category 5 cyclones.
For maintenance, you just get relieved by a spare powership and go to the shipyard. All fuel handling equipment is shared across the fleet, not duplicated in each plant. The uptime at the customer site is 100% because of the spare ship.
In case of military attack, you design it to sink safely and coolably, and have a salvage operation built into the design.
Absolutely epic. This could decarbonize the planet extremely rapidly.
Anyone working on this? You bet! This is Thorcon's entire pitch, and they think they can hit $1000/kW capital cost. MIT studied it too. [1].
[1] http://news.mit.edu/2015/new-look-floating-nuclear-power-062...