Angelica - Thanks for pointing out the impact of the AIA rule. My radical solution for that costly rule is to eliminate it. After all, the Commission itself is on record as stating that the rule is "not necessary to provide adequate protection."
I believe there is a way to unwind the ratchet, at least to a certain degree.
Start with recognizing that some rules have been imposed with associated language stating that they are "not necessary" for adequate protection. Identify those specific rules and repeatedly ask why they exist if they are not necessary.
Writers have the ability to raise this question and perhaps gain the attention of responsible authorities among both regulators and elected officials.
I've been thinking about it and I think there are two 'killer apps' for regulation
Governments need to start creating oversight agencies empowered to invalidate decisions and processes used by other agencies based on a reasonable cost benefit consideration. Nuclear safety agencies would be great candidates as the first ones to review. Won't even need to run numbers on aircraft impact!
I think the NRC is absolutely wedded to its reputation as the gold standard and will not meaningfully change except in direct response to an oversight agency that has different incentives
The second is more of a project and controversial, being Jack Devanney's Underwriter's Certification system, if you're familiar with it. I don't see this as happening for large light water reactors but I think it's intriguing for any reactor design deemed walkaway safe, perhaps by the oversight agency
It could also be like 30th of a kind with a module factory fully implemented and favourable borrowing rates, the number you lead with until you have to give a hard number for plant one
I think it was a huge mistake for nuscale to pursue such a large reactor project as it doesnt play to the strengths of SMRs. rather they should pursue a smaller more numerous reactor design for industrial power and heat needs and remote power needs.
the goal should be paving a way for a mass-produced mass accepted reactor whose economics improve with the volume of sales.
it does not make sense to launch yourself into a market in which nuclear is already struggling to cope with cheap gas and market breaking renewables.
I would look to see if the same regulation will apply to the dow chemical x energy SMR plan - there would be some illogic in saying well the smr needs concrete shielding because terrorism but the massive chemical plant producing whatever toxic substance needs none. it would really point to the illogical nature of the regulation if that was the case.
I really don’t like all the SMR doomerism. Big reactors have a place, but there is nothing magic about 1,650GW or whatever ridiculous number EDF wants to build. I agree the problem with the NuScale design is that all the reactors fit in a single pool that ends up being a gigantic civil works project, but that problem doesn’t extend to the entire SMR reactor class.
The fact is that 500MW LWRs were the fastest and cheapest thing around back in the 1960 and early 1970s. There are a lot of reasons, but scaling from 500MW to 1GW has literally NEVER proven to be cost competitive with the smaller reactors, so the arguments about going bigger are just comparing apples to oranges.
Turbine size, reactor size, steam generator size, amount of concrete, amount of decay heat, amount of cooling, etc. All of these get worse the bigger you go and that adds cost period. That doesn’t mean we can build a 450MW reactor easy now given all the constraints everyone is facing (Solar and wind anyone?), but it can be done with 1960s tech.
300MW is going to fit a lot of grid sizes really well...
It may not be, but if you can’t build the big one then why bother comparing? The financing of 2B is doable by utilities. 15B is not. And heck, the NRC required (thanks Congress!) 100 man army might be able to be shared for the site of multiple SMRs so the scaling effect could be different. Who knows- we gotta try different approaches to find out what works best.
I think the point everyone seems to be missing is that we need to make electricity 10-20X cheaper than it is right now for AI, Synfuels, 3rd World grid access, etc. There is nothing credible that I have seen that would even come close to that using just the existing large LWRs, or renewables or fossil fuels of any stripe.
If it’s a financing issue you’re trying to solve, maybe try financial rather than
Engineer innovations. And if SMRs are just to get around the problem of funding nuclear which is otherwise too massive, what’s there to suggest it’lol ultimately produce cheaper power?
Product value should be at least as important as price. Electricity usefulness needs more parameters than just cost/kWe because not all kWe are the same. Some come at the whim of the weather, have no associated inertia, and are only available from distant generators. Others are on demand, come complete with inertia, have associated low temperature heat available at a low cost, and can be directly connected to end users.
IOW, electricity from nuclear plants is a higher quality, more valuable product. Producers should consider marketing it more like BMWs or Mercedes than like Chevrolets.
I would agree that nobody is going to care about a cheaper price unless it delivers good value. There is a nice distinction between the two in German: billig=cheap and günstig=good value for money.
Basic economics though- if something valuable can be had for a lower price, the consumption of it will increase which helps poorer people. Well, it helps everyone, but the poorest benefit the most!
Cost really does matter! That’s ultimately why we need lots of ideas, business plans, technologies, etc. to succeed.
I cringe when people point to decisions made in the 1960s to scale up reactors as proof that we need to make the same decisions again in the 2020s.
It's almost never mentioned, but one of the big drivers in quickly scaling nuclear plants was the fact that GE and Westinghouse wanted to dominate the nuclear power plant supply business. They were essentially a duopoly in manufacturing very large turbine generators. They sought to drive skilled manufacturers like Allis Chambers and ALCO out of the market.
The first nuclear age was also a time when utility customers made more profit when power plant costs increased. Their suppliers also liked expanding the scope of their contracts.
Giant power plants are not inherently cheaper per unit of power output, especially if going smaller allows radical simplification by eliminating systems structures and components that are not needed when decay heat production is low enough to be handled with less forced convention.
But GE and Westinghouse were not the only ones going for larger reactors, right? Nobody continues to build those smaller reactors.
I agree with you that there is the possibility for SMRs to positively impact their cost by shedding systems. But the impact has to offset the economy of scale advantage.
In the US, B&W and Combustion Engineering joined into the business, but like GE and Westinghouse, their primary experience was in producing very large components for fossil fuel fired steam plants.
Manufacturers in other countries were building lower powered reactors, often using natural uranium fuels, but they faced competition from subtlety subsidized US suppliers that had access, for example, to low-priced enriched fuel that could reduce required core sizes.
US manufacturers also had access to superior forging equipment that enabled large pressure vessels. That capability stemmed from defense factories, mainly associated with large steam powered ships (like super carriers). Again, the cost of developing that infrastructure was borne by the defense sector.
My analysis – conducted over a 30 year period – tells me that there are as many "diseconomies of scale" as there are economies. Both are very real and lead to complex optimization challenges where the right answer depends on a number of variables.
But it's worth thinking about the fact that you can produce 1000 MWe using 5 200 MWe reactors, but you cannot economically produce 200 MWe using one 1000 MWe reactor operating at 20% power.
There are two kinds of economies of scale: unit economics and production economics. For example, in the computer industry, it was thought that larger and larger mainframe would have better cost-performance for each job to be done. Then came microprocessors and personal computers, made at consumer-market volumes by mass-production techniques, and suddenly, Small Modular Computers looked at lot more attractive.
Why is a nuclear power plant like a computer? I would argue it’s more like a wind turbine—bigger is better, but at some point going too big introduces new unforeseen problems.
Computers are just an example...just about all human-made entities have scale economies when you make larger numbers of them.
Consider also how the growth in load over time is handled...if you can add capacity in smaller increments, then you will at any time be paying for less unused excess capacity.
Angelica - Thanks for pointing out the impact of the AIA rule. My radical solution for that costly rule is to eliminate it. After all, the Commission itself is on record as stating that the rule is "not necessary to provide adequate protection."
https://atomicinsights.com/unnecessary-rules-should-be-eliminated/
I love this comment Rod and I wish there is some way to unwind regulatory ratcheting!
I believe there is a way to unwind the ratchet, at least to a certain degree.
Start with recognizing that some rules have been imposed with associated language stating that they are "not necessary" for adequate protection. Identify those specific rules and repeatedly ask why they exist if they are not necessary.
Writers have the ability to raise this question and perhaps gain the attention of responsible authorities among both regulators and elected officials.
I've been thinking about it and I think there are two 'killer apps' for regulation
Governments need to start creating oversight agencies empowered to invalidate decisions and processes used by other agencies based on a reasonable cost benefit consideration. Nuclear safety agencies would be great candidates as the first ones to review. Won't even need to run numbers on aircraft impact!
I think the NRC is absolutely wedded to its reputation as the gold standard and will not meaningfully change except in direct response to an oversight agency that has different incentives
The second is more of a project and controversial, being Jack Devanney's Underwriter's Certification system, if you're familiar with it. I don't see this as happening for large light water reactors but I think it's intriguing for any reactor design deemed walkaway safe, perhaps by the oversight agency
2 billion/300MWe for BWRX is quite competitive anyway.
The 1 billion estimation was too good to be true and it's better to be treated as a marketing gimmick.
Unfortunately a lot of people took it serious and nuclear companies crater their already thin credibility when they play stupid games
Well, in their defense they had to compete against windmill companies that were lying through their teeth about their costs... 😉
It could also be like 30th of a kind with a module factory fully implemented and favourable borrowing rates, the number you lead with until you have to give a hard number for plant one
I think it was a huge mistake for nuscale to pursue such a large reactor project as it doesnt play to the strengths of SMRs. rather they should pursue a smaller more numerous reactor design for industrial power and heat needs and remote power needs.
the goal should be paving a way for a mass-produced mass accepted reactor whose economics improve with the volume of sales.
it does not make sense to launch yourself into a market in which nuclear is already struggling to cope with cheap gas and market breaking renewables.
I think it came out of the aircraft impact rule. If you needed this expensive shell, it’s only economical to stick a bunch of modules in it.
I would look to see if the same regulation will apply to the dow chemical x energy SMR plan - there would be some illogic in saying well the smr needs concrete shielding because terrorism but the massive chemical plant producing whatever toxic substance needs none. it would really point to the illogical nature of the regulation if that was the case.
I really don’t like all the SMR doomerism. Big reactors have a place, but there is nothing magic about 1,650GW or whatever ridiculous number EDF wants to build. I agree the problem with the NuScale design is that all the reactors fit in a single pool that ends up being a gigantic civil works project, but that problem doesn’t extend to the entire SMR reactor class.
The fact is that 500MW LWRs were the fastest and cheapest thing around back in the 1960 and early 1970s. There are a lot of reasons, but scaling from 500MW to 1GW has literally NEVER proven to be cost competitive with the smaller reactors, so the arguments about going bigger are just comparing apples to oranges.
Turbine size, reactor size, steam generator size, amount of concrete, amount of decay heat, amount of cooling, etc. All of these get worse the bigger you go and that adds cost period. That doesn’t mean we can build a 450MW reactor easy now given all the constraints everyone is facing (Solar and wind anyone?), but it can be done with 1960s tech.
300MW is going to fit a lot of grid sizes really well...
So why did they keep building them bigger?
I just have a hard time imagining building say 3 BWRX-300 can be cheaper than building 1 full size ABWR.
It may not be, but if you can’t build the big one then why bother comparing? The financing of 2B is doable by utilities. 15B is not. And heck, the NRC required (thanks Congress!) 100 man army might be able to be shared for the site of multiple SMRs so the scaling effect could be different. Who knows- we gotta try different approaches to find out what works best.
I think the point everyone seems to be missing is that we need to make electricity 10-20X cheaper than it is right now for AI, Synfuels, 3rd World grid access, etc. There is nothing credible that I have seen that would even come close to that using just the existing large LWRs, or renewables or fossil fuels of any stripe.
If it’s a financing issue you’re trying to solve, maybe try financial rather than
Engineer innovations. And if SMRs are just to get around the problem of funding nuclear which is otherwise too massive, what’s there to suggest it’lol ultimately produce cheaper power?
Why should the ultimate goal be "cheaper" power?
Product value should be at least as important as price. Electricity usefulness needs more parameters than just cost/kWe because not all kWe are the same. Some come at the whim of the weather, have no associated inertia, and are only available from distant generators. Others are on demand, come complete with inertia, have associated low temperature heat available at a low cost, and can be directly connected to end users.
IOW, electricity from nuclear plants is a higher quality, more valuable product. Producers should consider marketing it more like BMWs or Mercedes than like Chevrolets.
I would agree that nobody is going to care about a cheaper price unless it delivers good value. There is a nice distinction between the two in German: billig=cheap and günstig=good value for money.
Basic economics though- if something valuable can be had for a lower price, the consumption of it will increase which helps poorer people. Well, it helps everyone, but the poorest benefit the most!
Cost really does matter! That’s ultimately why we need lots of ideas, business plans, technologies, etc. to succeed.
I cringe when people point to decisions made in the 1960s to scale up reactors as proof that we need to make the same decisions again in the 2020s.
It's almost never mentioned, but one of the big drivers in quickly scaling nuclear plants was the fact that GE and Westinghouse wanted to dominate the nuclear power plant supply business. They were essentially a duopoly in manufacturing very large turbine generators. They sought to drive skilled manufacturers like Allis Chambers and ALCO out of the market.
The first nuclear age was also a time when utility customers made more profit when power plant costs increased. Their suppliers also liked expanding the scope of their contracts.
Giant power plants are not inherently cheaper per unit of power output, especially if going smaller allows radical simplification by eliminating systems structures and components that are not needed when decay heat production is low enough to be handled with less forced convention.
But GE and Westinghouse were not the only ones going for larger reactors, right? Nobody continues to build those smaller reactors.
I agree with you that there is the possibility for SMRs to positively impact their cost by shedding systems. But the impact has to offset the economy of scale advantage.
In the US, B&W and Combustion Engineering joined into the business, but like GE and Westinghouse, their primary experience was in producing very large components for fossil fuel fired steam plants.
Manufacturers in other countries were building lower powered reactors, often using natural uranium fuels, but they faced competition from subtlety subsidized US suppliers that had access, for example, to low-priced enriched fuel that could reduce required core sizes.
US manufacturers also had access to superior forging equipment that enabled large pressure vessels. That capability stemmed from defense factories, mainly associated with large steam powered ships (like super carriers). Again, the cost of developing that infrastructure was borne by the defense sector.
My analysis – conducted over a 30 year period – tells me that there are as many "diseconomies of scale" as there are economies. Both are very real and lead to complex optimization challenges where the right answer depends on a number of variables.
But it's worth thinking about the fact that you can produce 1000 MWe using 5 200 MWe reactors, but you cannot economically produce 200 MWe using one 1000 MWe reactor operating at 20% power.
There are two kinds of economies of scale: unit economics and production economics. For example, in the computer industry, it was thought that larger and larger mainframe would have better cost-performance for each job to be done. Then came microprocessors and personal computers, made at consumer-market volumes by mass-production techniques, and suddenly, Small Modular Computers looked at lot more attractive.
Why is a nuclear power plant like a computer? I would argue it’s more like a wind turbine—bigger is better, but at some point going too big introduces new unforeseen problems.
Computers are just an example...just about all human-made entities have scale economies when you make larger numbers of them.
Consider also how the growth in load over time is handled...if you can add capacity in smaller increments, then you will at any time be paying for less unused excess capacity.
Not convinced. We can make serial manufacturing of larger reactors too. But of course yes ok smaller grids should take smaller reactors.
Last Energy, Thorcon, Rolls Royce. All are focused on engineering for scale rather than focusing on cool reactor tech. That should be rule number one.
Do you have a link to the original tweet by Jigar Shah about the cost of the BWRX-300?
Someone needs to call him on this NOAK cost of $6667 / kW. It comes off being fairly negative and not helpful. Their own report on Pathways to Commercial Liftoff, https://liftoff.energy.gov/wp-content/uploads/2023/05/20230428-Advanced-Nuclear-Pathways-to-Commercial-Liftoff-Webinar-vF_web.pdf , gives a generic FOAK cost of $6200 / kW and NOAK of $3600 / kW.
Don’t expect much help resolving this with our current Energy Secretary.