Nuclear power reactors: a study in technological lock-in

The interesting Robin Cowan has a well-cited paper with that title:

Recent theory has predicted that if competing technologies operate under dynamic increasing returns, one, possibly inferior, technology will dominate the market. The history of nuclear power technology is used to illustrate these results. Light water is considered inferior to other technologies, yet it dominates the market for power reactors. This is largely due to the early adoption and heavy development by the U.S. Navy of light water for submarine propulsion. When a market for civilian power emerged, light water had a large head start, and by the time other technologies were ready to enter the market, light water was entrenched.

Here are some recent articles on light water reactors in Japan.


A link for knicker-twisters.

lol... interesting point of view, dearieme :) is that your blog? I mean, is that your opinion? what can I say... I'm far away from the reactors, but I'm a little worried about the guys that are in the nearby.. they're human, just like us

If it turns out that climate change is the sort of disaster that some fear -- mass extinctions, crop failures, sea risings, 1 billion deaths, etc. -- this earthquake will go down as the thing that cemented its coming. Nuclear power is the only thing that we, in theory, could have built fast enough to slash our carbon output and with the scare tactics the anti-nuke nuts can use now, we're not going to be seeing many new nuclear plants going up any place where ordinary people have a say in these things. China, yes. USA, no, unless we have a lot more political will that I think.

That's not to say the failure of these plants that were theoretically designed for Japan's earthquakes isn't worrying, but what's the over/under on years of human life lost to any radiation leak that takes place this week? We probably don't have the data to calculate yet, but it's certainly four or five orders of magnitude smaller than the expected loss of life years from climate change, even if you only give the worst-case scenario a 1 in 20 shot of happening.

so many people are pushing that blog

he is a supply chain MBA, i am still very worried.

The critical line of research is how climate change caused the earthquake.

Or can we rule out Al Qaeda?

What the anti-nuclear crazies are going to forget is that the most fundamental part of the risk management strategy--the containment structure--worked as designed. Yes, a meltdown occurred when the colling system failed, and the reactors that were flooded with water as a last resort are ruined, but the containment structures did exactly what they were supposed to do: contain the mess and thus prevent contamination outside the site.

Even in the Three Mile Island incident, in which the humans made wrong decisions at every turn, there was no disaster because the most basic safeguards in the design--again, a robust containment structure--kept the worst of the mess inside the reactor building.

Nuclear plants built to Western standards are safe, even when they fail. It's the ramshackle Soviet-style designs that present a hazard. (Chernobyl had no containment structure.)

They should be built fail-to-safe. The link explains why this is non-trivial, but neither is it trivial getting people to deliver a miracle in the middle of a disaster, which we found out with the Gulf oil leak. The problem is that the relative moratorium on nuclear development is partly to blame.

Even taking the Japanese problems into account, nuclear is still the safest source of power bar none.

It's certainly a hell of a lot safer than coal.

The ironic thing is that there are reactor designs where meltdown is a physical impossibility, but new reactor designs are unacceptable with environmental lunies. We can only use 50 year old technology that has been grandfathered in.

Hmm, if the lock-in is the result of USN funding, why is it the people who build and sell reactors are Japanese and French?

The evidence that the light water reactor is inferior to the heavy water and graphite reactor is pretty weak. In fact I would contend that probably most of the evidence points in the other direction (but is still inconclusive).

There has been signifigant experience with heavy water reactors in Canada and while they are better in some areas they are worse in others and the general consensus from nuclear folks I knew was that they were inferior.

The helium/graphite reactors have had some experience as well (in England and Germany) and have always run into mechanical issues. That is a technology that from a high level looks great so science and engineering geeks love it, but when it comes to the nuts and bolts of making it work it has not proven sucesseful.

I would view with intense skepticism any paper that claimed there was a clear enough advantage between the 3 to illustrate any economic principle.

What exactly is the economic principle here? Are they saying there's a systematic reason why inferior technologies are being selected?

I've heard that most if not all reactors are based on designs for producing weaponized material, and that research into much safer reactors is possible.

A pebble-bed reactor uses micro-cores that are each encased in their own carbide shields. Multiple micro-cores are placed within a larger vessel and form a single reactor core. Sub-critical reactors use an external neutron source to initiate and sustain the chain reaction, usually a particle accelerator of some sort. Most of the radioactivity of the spent fuel is restrained to 7 years of pulling the material. With sub-critical, pebble-bed reactors, you just pull some of the spent micro-cores and add some new ones. The old ones keep producing heat, and thus power, while being exposed to the neutrons fissiles the radioactive waste down into less radioactive isotopes. If you keep them old micro-cores in there long enough, eventually they all just turn to iron (the most stable element) or lead. When the micro-cores are finally exhausted, they tend to be of much lower radioactivity than normal spent nuclear fuel and they have contributed much more of their mass to energy, so you use less fuel to begin with. That being said, there isn’t much practical knowledge on pebble-bed fuel cycles. There have been some experiments but they were all critical reactors.

Interesting if true!

Slate has an excellent article on nuclear power and public opinion in the wake of Japan's experience

Also, how much expected benefit will thorium reactors need to clear the hurdle of light water entrenchment

Or maybe it has to do with Bill Clinton scrapping the Integral Fast Reactor program in 1994. The IFR can use nuclear waste and fissile material from weapons for fuel. It produces a tiny fraction of the waste of other reactors, and uses up almost all of the fuel.

The pebble bed reactor is also a good design, but it has much higher reprocessing costs.

The BWRs in Japan are not "unsafe" reactors. While they were supposedly build to withstand earthquakes and probably got a D on that test, understand that the Earth's awesome power of moving continents is a bit difficult to overcome.

If a meteorite slammed into a reactor and there was a containment breach, would that be a reason to avoid nuclear energy?

I think we're going to continue building cities on the ocean near the Ring of Fire.

BTW, we all know this disaster has somethong to do with George Bush's policies.

John Kerry had much more to do with killing the IFR than Clinton.

> What exactly is the economic principle here? Are they saying there’s a systematic reason why inferior technologies
> are being selected?

I think they're saying that "if competing technologies operate under dynamic increasing returns", say because there is a natural monopoly or large returns to scale, there is a strong tendency to get stuck on the first idea that works even if it's not optimal. Space launch may have worked this way, for example.

It isn't saying that first ideas are necessarily suboptimal, although one could think of reasons why that might generally be true. A corollary might be that design spaces are generally under-searched. But it doesn't prove that alternatives are better - that would be a stronger statement.

The cited article is a case study of this effect on reactors. The article specifically says this effect applied in the case of reactors. The author says that the evidence, although not incontrovertible, supports light water reactors are "not the superior technology."

charlie> nuclear reactors required building entire new industries from scratch. Zirconium, for example, was produced in gram quantities in 1950 but reactors need it in tons. Using any design radically different from PWR/BWR would have meant paying enormous development costs thst the US navy had already paid for you, for a design that might not work.

yes, but the USN uses sodium reactors now.

No, the US briefly experimented with a sodium reactor in the 50s, but quickly rejected it in favor of PWR.

Again these reactors have actually been built and operated.

There are 29 CANDU heavy water reactors around the world.

"When first being offered, CANDUs offered much better "running" time statistics, the capacity factor, than light-water reactors of a similar generation. At the time, light-water (LWR) designs spent, on average, about half of their time in maintenance or refueling outages. However, since the 1980s dramatic improvements in LWR outage management have narrowed the gap between LWR and CANDU, with several LWR units achieving capacity factors in the 90% and higher range, with an overall fleet performance of 89.5% in 2005.[13] The latest-generation CANDU 6 reactors have demonstrated an 88-90% capacity factor, but overall fleet performance is dominated by the older Canadian units which generally report capacity factors on the order of 80%.[14]"

In addition there were a number of graphite reactors built in the UK. These too have not proven to be better than the LWR. In fact the evidence is that they are worse.

"Like the Magnox, CANDU and RBMK reactors, and in contrast to the light water reactors, AGRs are designed to be refuelled without being shut down first. This on-load refuelling was an important part of the economic case for choosing the AGR over other reactor types, and in 1965 allowed the CEGB and the government to claim that the AGR would produce electricity cheaper than the best coal fired power stations. However fuel assembly vibration problems arose during on-load refuelling at full power, so in 1988 full power refuelling was suspended until the mid-1990s, when further trials led to a fuel rod becoming stuck in a reactor core. Only refuelling at part load or when shut down is now undertaken at AGRs. [3]"

The nuclear industry is filled with wide eyed boosters promoting this technology or that as the next greatest thing if only the govt was willing to give it a boost to get started.

Just like the qwerty keyboard. Less efficient but ubiquitous.

Not entirely clear that is the case:

After months of retraining (note: months of retraining), some typists have shown some increase in efficiency.

Not every lock-in is inferior.

It is just not true that alternate designs were not evaluated. F.e. there was a production-size 'pebble-bed' reactor in Germany (at least one) however it was shut down after 5 years because of frequent problems (cracks in the pebbles, problems with pushing control rods between pebbles, high temperature difference between the middle and outer areas, difficult to extract/exchange pebbles)
And that is just the one I know a little bit more about.

While I kind of agree with the negative comments on alternate designs here, I'm not an expert on nuclear reactors.

It is fair to say that a lot more money has been spent on light water reactor designs than on alternates. Alternate v1.0 is competing with light water reactor v12.3. That's not a fair assessment of the merits of the alternate. If it's even close, that's suggestive of a problem. And the economics tells you that it's easy for this sort of problem to develop. Now, actually solving this problem, say by funding alternates heavily, might be more trouble than it's worth...

@Tyler Cowen
No, you are wrong this is not the case here. Its more to due with the impossibility of getting regulatory permits for non-light water reactors.

@Vertical Driver
You are absolutely correct.



@Doc Merlin You do know that there are reasons why its more difficult to get permits, right? (of course I don't know the reasons that the administrations have, but I could think of a few)

"is that your blog? I mean, is that your opinion?" Not my blog; my opinion tends that way. But I should caution that the only reactor whose design I have studied in detail was the one at Chernobyl: I was pressed into service (in a distinctly humble role) for Her Majesty's Government when that one went awry.

P.S. Othe whole I think governments tend to be lamentably incompetent, but the speed at which a huge pile of bumf about that reactor was put on my desk impressed me no end.

Every single thing that the authorities have announced is practically impossible to happen ... has happened. So. by now, I figure the smart play is to spend the next ten years at the bottom of Carlsbad Cavern.

It's not a question of technological lock-in when there has been no regulatory approval for a new plant in the US in decades. I'm not saying the regulators have necessarily been wrong, but that it's hard to see how technology can advance when no production is allowed.

actually is rewarding, or should be critical? Milton friedman in describing American win "core" when, still put secret U.S. domestic tremendous consumer market as an important weapon. But now people see it is this tremendous consumer market, without restraint, become overwhelmed overspend U.S. economy the last straw.

September 2010, the European Parliament and Council reached a political agreement to adopt three Commission proposals to effect a substantial change in the way financial supervision is to be exercised at the European level: to set

Comments for this post are closed