7 ms·
I disagree with this take. The main claim here is that naval reactors are very expensive, and getting even more expensive as the civilian nuclear industry dwind
by credit_guy 1y ago
I disagree with this take. The main claim here is that naval reactors are very expensive, and getting even more expensive as the civilian nuclear industry dwindles.
If only we could compare the cost of a submarine with and without the nuclear reactor. Fortunately we can: it is the French Suffren-class which is nuclear and has a conventional version, the Orka-class ([1], [2]). The cost of one Suffren-class is listen on wikipedia as €1.94 in 2022 money (or €2.1 in 2024 money) , and of one Orka-class as €1.4 in 2024 money. So the nuclear powered version is 50% more expensive, it would seem. Except that it has a much larger displacement, at 4690 tons vs 3248 tons for the conventional one (i.e 44% larger). So, all in all, it does not appear the reactor adds that much to the overall cost. Why? Military stuff is extraordinarily expensive already. One single torpedo fired by one of these submarines costs a whopping $5.39m in 2022 dollars. Firing such a torpedo is the equivalent of destroying about 20 Ferraris.
[1]https://en.wikipedia.org/wiki/Suffren-class_submarine https://en.wikipedia.org/wiki/Suffren-class_submarine
[2] https://en.wikipedia.org/wiki/Orka-class_submarine https://en.wikipedia.org/wiki/Orka-class_submarine
[3] https://en.wikipedia.org/wiki/Mark_48_torpedo https://en.wikipedia.org/wiki/Mark_48_torpedo
- alephnerd 1y agoThe issue is economies of scale. On average, only 1 Suffren-class submarine has been laid every 3 years since 2007, and no new orders have been placed since 2020. That is not enough dealflow to support an SoE like TechnicAtome, let alone a private sector nuclear energy player like a Westinghouse. On top of that, the reactor used (the K15) has only been used in just 12 naval applications since 1980. Entire decades can pass without new military deals for a K15-type reactor, which can lead to skill and experience attrition for what is a fairly complex piece of technology. Meanwhile, the know-how to build a diesel-electric transmission for naval applications is constantly in demand as the same technology is used in commercial naval vessels as well as locomotives. By trying to justify a dual-use case for SMRs for energy as well as commercial propulsion (which is what both China and India are investing in), it makes it easier to help maintain the knowledge needed to develop nuclear propulsion systems as well as ensure deal flow exists so a Westinghouse-style bankruptcy does not occur.