[pfdietz]

By 1/4th, do you mean 1/4th of nuclear? Sure, I'm totally happy with that. Once EROI is high enough, further improvement has only marginal effect. If EROI of PV is 10 (say) and nuclear is 40 (say), that just means nuclear is 1/10 - 1/40 or .075 better than PV by this metric. The contribution of this to the cost difference is minor, and is swamped by the other contributions to the cost of nuclear.

[ya_throw]

I don't think you understand how the maths works.

The EROEI of nuclear is 106, by the link I previously provided. The very best solar installations have an EROEI of 7. That means nuclear gives 106/7 = 15x more energy output than solar, for a given amount of manufacturing energy input.

It's a simple ratio calculation, I'm not sure why you decided to subtract the reciprocals of the two numbers.

> Once EROI is high enough

The EROI of solar is 7, which means you get 7 units of energy output for every unit of energy expended in manufacturing. That is a terrible return. In no way is that "high enough". In fact, we shouldn't be bothering with it at all.

[pfdietz]

> The very best solar installations have an EROEI of 7.

That sounds like a regurgitation of Ferroni and Hopkirk's analysis, which has been well debunked,

https://www.nrel.gov/docs/fy17osti/67901.pdf

The estimate for EROI of PV IN EUROPE is somewhere around 8, and of course Europe is a terrible place for solar -- the EROI for PV in a sunnier place, like Chile, Namibia, or the middle east, would be nearly twice this. If energy costs were really important, one would not put the PV factory in a place where energy were expensive.

That the EROI of solar is adequate should be obvious because the economic return on investment is good. If solar in Dubai can come in at less than $0.02/kWh then the energy cost (which will always be just a small fraction of the total manufacturing cost) will be reasonable.

Attempts to show PV has bad EROI very often run into methodological problems, extending the system boundaries beyond what analyses of the competing systems use (if you extend the boundary far enough, to the whole society, in steady state the EROI always converges to 1, since all energy produced is consumed somewhere. This is not a meaningful result.)