There’s a lot of very simplistic analysis that makes things seem vastly worse but you don’t need crazy assumptions for things to look reasonable. Basically the more sources of energy you can add to the mix the cheaper things become as you only include them when you save money. Thus if you ignore Wind in an estimate adding wind can only lower costs, ditto for nuclear, geothermal, etc.

Hydro is reliable and better than continuous power because it’s flexible enough you can fill in gaps based on what you forecast production will be across the next week. US only gets 6.2% of its power from hydro so the rest of the world actually looks better than these numbers. So, let’s assume a US grid with only Solar and Hydro, but that Hydro is flexible enough to reduce the need for batteries by 5%.

How much Solar to build is part of a complex optimization problem, but let’s simplify it 2c/kWh of capacity but 1/2 of all solar is wasted over a year, we’re not even close to accurate enough for reasonable increases or decreases to matter. Further, let’s assume 1/2 of all solar is used directly with the rest available to charge daily batteries again pessimistic with that kind of oversupply and HVDC lines etc but whatever. Now we need 50% - 5% (hydro) = 45% of total demand from batteries. (Notice how even a little wind could significantly change these numbers.)

Projecting LCOE to batteries sees constant reductions moving forward. Battery degradation is closely correlated with use so let’s use 120$/MWh for our long term estimate just as a reasonably high but not crazy ballpark. By the time production really ramps up in 15+ years inflation adjusted prices will likely to be much lower, but again high estimates are fine.

Now our worst case is something like 4c/kWh (solar/hydro) + 12c/kWh (batteries) * 45% = 9.4c/kWh for 24/7/365 power with zero fossil fuels and zero nuclear. On the surface that might seem terrible or awesome depending on what your local rates look like, but this was a very pessimistic estimate. Wasting half of all solar production seems unlikely to be the most efficient option, batteries just keep getting cheaper, wind is 3.2c/kWh and significantly reduces need for batteries etc etc.

But our pro nuclear investor needs to be pessimistic in the other direction. Remember demand is shifted to cheaper nighttime rates, if daytime electricity is cheaper then nighttime use drops further. Baseline numbers of say 4c/kWh production + 30% batteries at 8c/kWh = 6.4c/kWh 24/7/365. If you’re considering building nuclear and projecting 60+ years out those kind of numbers represent a real threat.

Most of the estimates I've seen for storage requirements for current electricity usage are not optimistic[1]. When you add in the huge amount of energy used for heating, which is currently not electric, it gets even worse.

I'd like to see an analysis of how much generation and storage is required to handle Minnesota's natural gas usage[2], with a reasonable guarantee that power will not be lost for more than an hour or two during the entire yearly six month cold weather span. Are those generation & storage estimates reasonable? How much will they cost, how reliable are they, how much room will they take, how many natural resources will they consume? Keep in mind Minnesota will not be able to bogart the entire battery manufacturing capacity of the globe, and remember that this analysis is in addition to the amount required to meet current electricity usage needs. Then, extrapolate that analysis to all other cold weather areas, such as other northern US states and most of Canada. Is renewable+storage really feasible to meet that need?

I've never seen this analysis done, and what I've seen from other analyses makes me think it's not feasible without depending on unproven future tech. But I'd love to be wrong!

[1] For example, https://news.ycombinator.com/item?id=36949165

[2] Approximately 500 billion cubic feet of natural gas per year. Note also that the usage is not evenly distributed through the year, we use a lot more in February than we do in July. https://www.eia.gov/dnav/ng/NG_CONS_SUM_DCU_SMN_A.htm

There is so much peer reviewed work saying the opposite of what he claims that you should default to skeptical about him. It's the same way one should treat a creationist.