My argument has nothing to do with ‘greenness’. What I am talking about already has a name: marginal productivity or marginal efficiency.

My point is if you have to dig and burn 2.5 rocks to get 1 rocks worth of energy you are really concerned with that ratio for a whole host of reasons and affects everything you do (it is one reason why coal plants are huge for instance).

Now say you can get a rocks worth of energy with no rocks to dig and burn. Of course you still want to get the most amount of energy and there is an upper limit to how much you can collect but you no longer think about it in terms of the number of rocks you have to dig out and burn. That is a huge difference.

That's fine, and that's a metric that's worth considering, but you're not talking about the same kind of efficiency arcticbull was talking about. When someone says that "a solar panel is 20% efficient" what they mean is that, of the energy contained in the light that hits it, 20% of it is converted into electricity and the rest is converted into waste heat stored in the panel. Yes, if the panel hadn't been there 100% of that energy would have gone unused, but that's a different discussion. It doesn't make what arcticbull said false. Optimistically a new solar panel is 20% efficient, and that efficiency becomes worse with age.

Articbull is saying nuclear is better than solar for the energy system. He bases this on a claim that solar panels are inefficient along a number of dimensions. Look at his actual comments, they aren’t narrowly related to lab or theoretical efficiency (he is making points about land use, death, etc). There is no real concept of physical inefficiency there is only relative efficiency. And relative efficiency of Rankine cycle prime movers vs semiconductors is not a strictly easy comparison. 20% of something free that would have been entirely wasted is different than 40% of something you have to find, dig out, transport, crush, then filter. I also didn’t say what he said about theoretical panel efficiency is wrong (except all his actual numbers were wrong until he edited them). I am broadly saying this point doesn’t support his conclusion.

Edit: you are also wrong about what happens to the energy that hits the panel. Some light is reflected, a lot of light passes right through the panel. Further, panels are designed for particular wavelengths which is why multi gap panels have theoretical efficiencies higher than any thermal plant. And some is turned into heat. All of this would have happened had the panel not been there too except no usable electricity would be created.

> Look at his actual comments, they aren’t narrowly related to lab or theoretical efficiency (he is making points about land use, death, etc).

I don't know about deaths, but the point about land use is related to theoretical efficiency. If an existing solar plant that is 20% was 100% efficient it could generate five times as much power in the same space, or generate the same power in a fifth of the land it currently uses. The land used by solar power plants is very relevant when considering its practical utility.

> 20% of something free that would have been entirely wasted is different than 40% of something you have to find, dig out, transport, crush, then filter.

It's not free, though, because those rooftop panels will lose efficiency over time. Eventually they will need to be replaced or they will generate so little power that they may as well not be there. Since they need to be replaced over time then harvesting that 20% of sunlight is no longer free, but carries an ongoing resource cost. Is that more or less than the cost of getting the same total energy from a nuclear power plant?

> I am broadly saying this point doesn’t support his conclusion.

The conclusion is that solar and wind can at best only complement nuclear, and this is supported by various points about the inefficiency of solar panels. Setting wind aside for the moment, why do you think the fact that solar panels are inefficient and degrade over time doesn't affect their viability as an alternative to nuclear power?

Response to edit: I don't know whether the 20% figure counts in the theoretical 100% light that is reflected or unabsorbed. It seems to me that it shouldn't be counted, since if you're going to count light that couldn't have interacted with the cells anyway, then you may as well count light that hits the ground next to the panel, which would mean the panel is 0% efficient.

If nuclear plants were twice as efficient that would be great too. They aren’t.

Land use isn’t relevant for rooftop solar. There is basically no opportunity cost.

Everything becomes less efficient over time. Sunpower warranties their panels for a 20% decline in performance over 25 years. That means a 20% efficient panel is warrantied to be 16% efficient after 25 years in the field. To put that in perspective, panels made 15 years ago were around 16% efficient. There isn’t a cliff and they are still useful.

Suppose you abandon the system for whatever reason. You go back to the steady state of not producing power on that site. There is again no cost.

I am not saying nuclear is bad. I am outright rejecting the claim that solar panels are inefficient on any dimension that counts. I think this argument about degradation is beyond silly. I also think you are an astroturfing troll.

If you're going to assume bad faith then there's no point to the discussion.

I didn’t assume. I think you demonstrated. I agree there is no point.

>Eventually they will need to be replaced or they will generate so little power that they may as well not be there. Since they need to be replaced over time then harvesting that 20% of sunlight is no longer free, but carries an ongoing resource cost. Is that more or less than the cost of getting the same total energy from a nuclear power plant?

Yeah, is it more or less? I don't know either. But finally we're getting to the core of the question instead of debating about the number of angels that can dance on the head of a pin.

So at the end of its lifetime, be that 50 or 100 years in, a nuclear plant will have to be replaced, too. Basically a sudden 100% degradation. What is the resource cost for that, why are we not talking about that in this context? Because it's not "ongoing"? What's the resource cost for cooling with increasingly sparse water bodies that due to climate change will be less and less suited for it over the lifespan of the plant, BTW degrading its efficiency? What's the resource cost of plant dismantling and disposal, which goes on for decades after a plant has ceased to produce power? Would you like to include nuclear waste disposal as an ongoing resource cost in the consideration? Where do you draw the line? Are there any reliable estimates about these externalities that don't range wildly depending on which faction informed them? Do we want go down that rabbit hole?

If not, then what's the point of fixating on this single aspect with regards to solar and solar only? Just because this one is a figure that's easy to quote out of context and sounds kinda bad? Can I make you realize that in this light it's hard to not think of this as deliberately cherry-picking an argument for argument's sake?

These are all retorical questions, of course. FWIW, I don't claim to have any definite answers on the topic, and actually I do think that nuclear and renewables will have to complement each other to varying degrees around the globe, depending on many factors. I'm just trying to get across why the specific point your're clinging to seems so extremely moot to me.

>Setting wind aside for the moment, why do you think the fact that solar panels are inefficient and degrade over time doesn't affect their viability as an alternative to nuclear power?

Because it's already correctly priced into the running costs, just as - I suppose its proponents will claim - the corresponding facts above are for nuclear. Also, as has been pointed out, the ever repeated claim of solar inefficiency is baseless in the most literal sense, due to the lack of a comparable reference point for nuclear. 20% is an entirely meaningless number as long as the system boundaries are drawn so arbitrarily that it's impossible to benchmark. To put it bluntly, what would be sufficiently efficient for your taste? 30%? 60? 90? Can you explain why? It's just apples and pears on this level, the only reasonable comparison could come from a much broader view that you're explicitly not interested in.

Anyway this discussion is going in circles. I'm not assuming bad faith on your side, but I don't think we'll come to an agreement either.

>What is the resource cost for that, why are we not talking about that in this context?

Presumably because nobody knows it. Personally, I'm not aware of any nuclear power plant that has been replaced. I also don't know why one would need to replace the entire plant and not just the core and possibly surrounding structures.

>Would you like to include nuclear waste disposal as an ongoing resource cost in the consideration? Where do you draw the line?

I don't think a line should be drawn anywhere. If we're talking about sustainability and total impact, we should look at all the facts, otherwise we're drawing conclusions off incomplete data. It should be possible to reduce the entire situation down to a single "impact per Joule" value that's comparable between any two power production systems, and every impact that's caused in order to produce that Joule (mining, land use, pollution, etc.).

>If not, then what's the point of fixating on this single aspect with regards to solar and solar only? Just because this one is a figure that's easy to quote out of context and sounds kinda bad? Can I make you realize that in this light it's hard to not think of this as deliberately cherry-picking an argument for argument's sake?

The point is that treating rooftop solar as "free" because it's supposedly a one-time cost is at best naive, at worst disingenuous. I have no interest in making nuclear look good or solar look bad, but let's call a spade a spade. Neither of them has zero impact on the environment. Pretending that they do means we're not making rational decisions.

>Also, as has been pointed out, the ever repeated claim of solar inefficiency is baseless in the most literal sense, due to the lack of a comparable reference point for nuclear.

I wouldn't say it's baseless. The figures on efficiency by themselves are measurable and true. It is true, though, that nothing can be said to be (in)efficient in an absolute sense, for the reasons you say. That's why I never said "solar panels are inefficient". All I said on the matter was that efficiency in this sense is exactly the same kind of efficiency that an IC engine has, because, as I said, dpierce9 was trying to reframe the discussion for no reason. Well, I don't think it was for no reason.

>Because it's already correctly priced into the running costs, just as - I suppose its proponents will claim - the corresponding facts above are for nuclear.

I'm not sure either are. Is the pollution caused by semiconductor production, or as you mentioned the warming of natural waters used as coolant, priced into the corresponding products?

>the only reasonable comparison could come from a much broader view that you're explicitly not interested in.

Yes, I agree, that's the only reasonable comparison. I would be interested in that if anyone came forward with an actual analysis from which we could derive a figure like the one I mentioned above.