There is no single social cost (or benefit) of carbon. It will vary across countries and locations. I'm not saying you assume otherwise but it's worth making the point.
"I wanted a density with a fat upper tail reflecting the fact that the SCC could be extremely large." Did you also allow for the possibility that the SSC is negative? In other words, a benefit rather than a cost?
but CO2 spreads more or less evenly over the planet, so from the view of the species as whole we have to base policy on the total effect on humanity. There is a single SCC for humanity as a whole.
The density that I came up with does have a 0.01 probability of net benefit. The piece is clear on this. My personal thinking is that the pluses associated with warming up Canada, etc, are unlikely to be larger than the disbenefits such as the rise in sea level, but it is possible.
But the whole point of this piece is that each person should come up with their own density. Nobody should pay any attention to my density. And when they do their own density, they may find that their optimal grid is not that different from my optimal grid. or your optimal grid. Certainly not if we allow nuclear to come close to its should-cost.
Thanks for your reply, Jack. You can certainly, in principle, find a single cost/benefit of carbon for our species but policy is made primarily within countries. I can understand a poor country desperately trying to increase its energy supply not wanting to use that global number.
I think it quite likely that increased CO2 is a net benefit. Of course, there are transition costs to *any* change from the status quo. But with the slow pace of change, those adjustments can be very gradual. For example, planting is annual or more frequent and the types of crops can shift. The world is just emerging from a time of unusually low CO2 (280 ppm) -- not far above starvation levels for plans. [I see you make that point in the piece you referenced.] We have seen made increases in agricultural productivity at least partially attributable to higher levels of the plant food, CO2. I think we are also likely to gain more land than we lose. David Friedman has done some good work looking at the overall sign of change.
Of course, we can't resolve this here. Whether positive or negative overall, we still need more nuclear at a sensible cost.
Jack, thanks, I would do that if I had the time. Maybe for a blog post later. I think it may well be a net benefit, certainly not a high cost, even if that makes it harder for nuclear. On the other side, I have supported nuclear for 45 years.
As for density, I will leave with a quote from Back to the Future: "I am your density." :-)
"I'm a supporter of nuclear." Don't you need to qualify that statement.? Are you a supporter of nuclear at $32k/kW,? At 16k/kW.? At $8k?
Nuclear has lost my "support" somewhere south of 4k. With yr SCC density, you need a nuclear that is cheaper than coal, maybe a lot cheaper , before it will be part of your grid.
Nuclear power is not a religion, or a political party. It's just a way of making electricity. A tool that's only beneficial if it's cheap enough.
Yes, I need to clarify it for some purposes. What I mean by it is that I have tried to educate people about nuclear power -- especially its safety record and risks. I also favor it because it is minimally polluting (and I don't mean CO2).
That doesn't mean I favor nuclear at any cost. Only by overcoming people's false beliefs about nuclear are we likely to reduce the excessive burden of regulation and political opposition. So, in that sense, I will say that I support nuclear or that I am pro-nuclear. I agree, of course, that this doesn't mean pushing nuclear at any cost.
You say that nuclear is not a religion or a political party. True but people do often treat as if it is. They treat it like the Devil and radiation like an evil and science-created emanation that only destroys. (But only radiation from nuclear power, not from the sun or from granite...) Much opposition does in fact come from a place that's much like blind religious or political beliefs.
A later thought: Although I doubt that CO2 is a net cost, fossil fuels -- for all the tremendous benefits they have brought us -- are polluting. There is a cost to health, so *that* is a real cost. I don't have numbers to hand (do reliable numbers exist?) but I imagine they are much lower than in earlier decades in developed countries where we use scrubbers and such. Probably more polluting in many poorer countries. So *this* cost definitely favors nuclear.
This is a great piece. When i was in grad school at CMU, after a particularly brutal class on Baysian reasoning, one of my classmates said, "I've got it! Either it happens or not. 50-50." :-)
I am the business of power analysis, but I don't really understand what you are doing. I WOULD like to understand it. Can you please the basic concept in an easy to understand paragraph for an energy expert but someone who is not an expert in Bayesian statistics. Also, when you use the term 'grid LCOE' what do you mean, exactly? Is the Total System Cost type LCOE for a specific grid?
Yes, each point on the trade-off curves represents the combination of nuke/coal/gas/wind/solar/battery/h2 storage which resulted in that grid cost and CO2 emissions. The LCOE are the model's total cost at the plant gate. No transmission costs. The points on the tradeoff curves are optimal in the sense that it is impossible to have both cheaper power and less CO2. For a much more complete description of the GKG Grid Model see
As soon as you pick a point on the tradeoff curve, you have picked a Social Cost of CO2. It's the slope of the trade off curve at that point. A Bayesian is someone who admits he is uncertain about what his SCC is, but is willing to put a probability density function on this uncertain number. It turns out a Bayesian should pick the point where the slope of the trade off curve is equal to the mean of his SCC PDF. See
There is no single social cost (or benefit) of carbon. It will vary across countries and locations. I'm not saying you assume otherwise but it's worth making the point.
"I wanted a density with a fat upper tail reflecting the fact that the SCC could be extremely large." Did you also allow for the possibility that the SSC is negative? In other words, a benefit rather than a cost?
Max,
Yes, some areas will benefit from global warming. as this old piece makes clear:
https://gordianknotbook.com/download/how-much-time-do-we-have-an-engineer-looks-at-global-warming-part-1/
but CO2 spreads more or less evenly over the planet, so from the view of the species as whole we have to base policy on the total effect on humanity. There is a single SCC for humanity as a whole.
The density that I came up with does have a 0.01 probability of net benefit. The piece is clear on this. My personal thinking is that the pluses associated with warming up Canada, etc, are unlikely to be larger than the disbenefits such as the rise in sea level, but it is possible.
But the whole point of this piece is that each person should come up with their own density. Nobody should pay any attention to my density. And when they do their own density, they may find that their optimal grid is not that different from my optimal grid. or your optimal grid. Certainly not if we allow nuclear to come close to its should-cost.
Thanks for your reply, Jack. You can certainly, in principle, find a single cost/benefit of carbon for our species but policy is made primarily within countries. I can understand a poor country desperately trying to increase its energy supply not wanting to use that global number.
I think it quite likely that increased CO2 is a net benefit. Of course, there are transition costs to *any* change from the status quo. But with the slow pace of change, those adjustments can be very gradual. For example, planting is annual or more frequent and the types of crops can shift. The world is just emerging from a time of unusually low CO2 (280 ppm) -- not far above starvation levels for plans. [I see you make that point in the piece you referenced.] We have seen made increases in agricultural productivity at least partially attributable to higher levels of the plant food, CO2. I think we are also likely to gain more land than we lose. David Friedman has done some good work looking at the overall sign of change.
Of course, we can't resolve this here. Whether positive or negative overall, we still need more nuclear at a sensible cost.
Max,
If you can come up with your density, I will run it.
If more CO2 is a net benefit, you are going to need a very low nuke cost before you switch from coal. The nuclear establishment hates guys like you.
Jack, thanks, I would do that if I had the time. Maybe for a blog post later. I think it may well be a net benefit, certainly not a high cost, even if that makes it harder for nuclear. On the other side, I have supported nuclear for 45 years.
As for density, I will leave with a quote from Back to the Future: "I am your density." :-)
Max,
"I'm a supporter of nuclear." Don't you need to qualify that statement.? Are you a supporter of nuclear at $32k/kW,? At 16k/kW.? At $8k?
Nuclear has lost my "support" somewhere south of 4k. With yr SCC density, you need a nuclear that is cheaper than coal, maybe a lot cheaper , before it will be part of your grid.
Nuclear power is not a religion, or a political party. It's just a way of making electricity. A tool that's only beneficial if it's cheap enough.
Yes, I need to clarify it for some purposes. What I mean by it is that I have tried to educate people about nuclear power -- especially its safety record and risks. I also favor it because it is minimally polluting (and I don't mean CO2).
That doesn't mean I favor nuclear at any cost. Only by overcoming people's false beliefs about nuclear are we likely to reduce the excessive burden of regulation and political opposition. So, in that sense, I will say that I support nuclear or that I am pro-nuclear. I agree, of course, that this doesn't mean pushing nuclear at any cost.
You say that nuclear is not a religion or a political party. True but people do often treat as if it is. They treat it like the Devil and radiation like an evil and science-created emanation that only destroys. (But only radiation from nuclear power, not from the sun or from granite...) Much opposition does in fact come from a place that's much like blind religious or political beliefs.
A later thought: Although I doubt that CO2 is a net cost, fossil fuels -- for all the tremendous benefits they have brought us -- are polluting. There is a cost to health, so *that* is a real cost. I don't have numbers to hand (do reliable numbers exist?) but I imagine they are much lower than in earlier decades in developed countries where we use scrubbers and such. Probably more polluting in many poorer countries. So *this* cost definitely favors nuclear.
This is a great piece. When i was in grad school at CMU, after a particularly brutal class on Baysian reasoning, one of my classmates said, "I've got it! Either it happens or not. 50-50." :-)
I am the business of power analysis, but I don't really understand what you are doing. I WOULD like to understand it. Can you please the basic concept in an easy to understand paragraph for an energy expert but someone who is not an expert in Bayesian statistics. Also, when you use the term 'grid LCOE' what do you mean, exactly? Is the Total System Cost type LCOE for a specific grid?
Steven,
I apologize. The piece really is not self-standing. Please check out
https://jackdevanney.substack.com/p/nuclear-and-windsolar for some background.
Yes, each point on the trade-off curves represents the combination of nuke/coal/gas/wind/solar/battery/h2 storage which resulted in that grid cost and CO2 emissions. The LCOE are the model's total cost at the plant gate. No transmission costs. The points on the tradeoff curves are optimal in the sense that it is impossible to have both cheaper power and less CO2. For a much more complete description of the GKG Grid Model see
https://gordianknotbook.com/download/the-gkg-grid-model
As soon as you pick a point on the tradeoff curve, you have picked a Social Cost of CO2. It's the slope of the trade off curve at that point. A Bayesian is someone who admits he is uncertain about what his SCC is, but is willing to put a probability density function on this uncertain number. It turns out a Bayesian should pick the point where the slope of the trade off curve is equal to the mean of his SCC PDF. See
https://gordianknotbook.com/download/thinking-quantitatively-about-co2-uncertainty
for a proof.
Great, thanks, I will study for sure.