Given the data around capacity additions, growth rates, and learning rates, I’m really puzzled how you end up scoring nuclear well on your QUICK, ZERO, and FACILITATE criteria:
“Criterion QUICK: An energy strategy must lead to the highest practically feasible near-term emission reductions for the actor’s energy system, without causing significant harm elsewhere.”
“Criterion ZERO: An energy strategy pursued by an actor must, with high chance of success, reduce the total greenhouse emissions of the actor’s energy system to zero in a suitable time frame, without increasing emissions elsewhere.”
“Criterion FACILITATE: An energy strategy must facilitate, rather than impede, the path toward global zero emissions, in a suitable time frame.”
Zero-carbon energy technologies at <10 GW capacity additions per year with little to now growth over the last decade fail QUICK and ZERO. An energy technology also fails FACILITATE if it takes 5-7 years to become operational, and the same investment could have purchased more or cheaper zero-carbon resources such as VRE or efficiency that would be deployed sooner. The net result from delay nuclear entails may mean higher cumulative emissions.
It’s also weird to me that you single out nuclear as the key to technology to pair with renewables when there are several other strategies to provide that firm generation role. Your same arguments apply to other technologies such as H2 CTs, storage, or hot rock geothermal. These all get much less advocacy support, but appear to have far better unit economics, and are more or less at similar levels of development and deployment to SMRs. They also don’t entail any of the profileration, waste, and weapons concerns.
I’m not against nuclear, I just find the singular, fanatical support for it really odd given the technology landscape.
Lastly, no analysis of decarbonizing electricity is complete without acknowledging efficiency.
All generation sources have negative externalities. Efficiency and demand response reduce electric demand and have large, positive externalities. Buildings are ~75% of the load on the grid. Most buildings are horribly inefficient—it’s common to see a 3x difference in energy use between buildings of the same size and type. New buildings are getting much better. Building energy codes in developed countries are moving to net-zero as code by 2030. It’s difficult to understate how significantly zero-energy codes will change the nature of the grid in the long run. Also, building floor area will triple by 2050, with most construction in countries without building energy codes. In the U.S., energy codes over the last 50 years have counterfactually reduced grid demand by ~60%. The same could be done for developing countries.
Why don’t grid modelers include energy efficiency potential in their models? Simple—they only focus on the supply side and treat demand as exogenous, because they don’t have the expertise to do a bottom-up model of electric demand (yet—it has been done on city-scale projects). Seems to me efficiency is the closest thing to an ethical imperative if such imperatives meaningfully exist.
You are right that investing in new nuclear tends to score rather badly according to QUICK. We say the same in the paper. However, as we explain, we tend to regard out two other criteria as more important, because of lock-in and path dependency risks if we focus mostly on short-term emission reductions. (Considerations about harm from emissions being cumulative count in the other direction, though.)
You are also right about other technologies potentially being able to play a similar role. Hydropower especially often does so already. Where it is available (or where geothermal plausibly will), there is no, or at least no strong, imperative to invest in new nuclear from ZERO, though it may still be wise.
Your points on efficiency are well taken, but I don’t quite see how they are relevant. By indicating that overall demand is set to fall, at least if the right steps are taken? I’m not sure about that. In any case, the grid modellers whose works we cite tend to assume very large efficiency gains and significantly lower overall final energy demand by mid-century in developed countries. I am sceptical that these are helpful assumptions (notably, we might want to make sure that developed countries can still have energy-intensive industry and/or contribute significantly to negative emissions), but they are made.
Last year (2021) the world added 182 GW of solar PV and 90 GW of wind capacity.
Nuclear globally saw 5 GW of additions, and 8 GW of retirements, for a net −3 GW capacity. Nuclear additions will need to increase considerably to maintain net capacity as many retirements are scheduled in the next decade.
Given the data around capacity additions, growth rates, and learning rates, I’m really puzzled how you end up scoring nuclear well on your QUICK, ZERO, and FACILITATE criteria:
“Criterion QUICK: An energy strategy must lead to the highest practically feasible near-term emission reductions for the actor’s energy system, without causing significant harm elsewhere.”
“Criterion ZERO: An energy strategy pursued by an actor must, with high chance of success, reduce the total greenhouse emissions of the actor’s energy system to zero in a suitable time frame, without increasing emissions elsewhere.”
“Criterion FACILITATE: An energy strategy must facilitate, rather than impede, the path toward global zero emissions, in a suitable time frame.”
Zero-carbon energy technologies at <10 GW capacity additions per year with little to now growth over the last decade fail QUICK and ZERO. An energy technology also fails FACILITATE if it takes 5-7 years to become operational, and the same investment could have purchased more or cheaper zero-carbon resources such as VRE or efficiency that would be deployed sooner. The net result from delay nuclear entails may mean higher cumulative emissions.
It’s also weird to me that you single out nuclear as the key to technology to pair with renewables when there are several other strategies to provide that firm generation role. Your same arguments apply to other technologies such as H2 CTs, storage, or hot rock geothermal. These all get much less advocacy support, but appear to have far better unit economics, and are more or less at similar levels of development and deployment to SMRs. They also don’t entail any of the profileration, waste, and weapons concerns.
I’m not against nuclear, I just find the singular, fanatical support for it really odd given the technology landscape.
Lastly, no analysis of decarbonizing electricity is complete without acknowledging efficiency. All generation sources have negative externalities. Efficiency and demand response reduce electric demand and have large, positive externalities. Buildings are ~75% of the load on the grid. Most buildings are horribly inefficient—it’s common to see a 3x difference in energy use between buildings of the same size and type. New buildings are getting much better. Building energy codes in developed countries are moving to net-zero as code by 2030. It’s difficult to understate how significantly zero-energy codes will change the nature of the grid in the long run. Also, building floor area will triple by 2050, with most construction in countries without building energy codes. In the U.S., energy codes over the last 50 years have counterfactually reduced grid demand by ~60%. The same could be done for developing countries. Why don’t grid modelers include energy efficiency potential in their models? Simple—they only focus on the supply side and treat demand as exogenous, because they don’t have the expertise to do a bottom-up model of electric demand (yet—it has been done on city-scale projects). Seems to me efficiency is the closest thing to an ethical imperative if such imperatives meaningfully exist.
Thanks for your thoughtful response!
You are right that investing in new nuclear tends to score rather badly according to QUICK. We say the same in the paper. However, as we explain, we tend to regard out two other criteria as more important, because of lock-in and path dependency risks if we focus mostly on short-term emission reductions. (Considerations about harm from emissions being cumulative count in the other direction, though.)
You are also right about other technologies potentially being able to play a similar role. Hydropower especially often does so already. Where it is available (or where geothermal plausibly will), there is no, or at least no strong, imperative to invest in new nuclear from ZERO, though it may still be wise.
Your points on efficiency are well taken, but I don’t quite see how they are relevant. By indicating that overall demand is set to fall, at least if the right steps are taken? I’m not sure about that. In any case, the grid modellers whose works we cite tend to assume very large efficiency gains and significantly lower overall final energy demand by mid-century in developed countries. I am sceptical that these are helpful assumptions (notably, we might want to make sure that developed countries can still have energy-intensive industry and/or contribute significantly to negative emissions), but they are made.