Is there a third way?
Figure 1. The Star Chamber. This 16th Century tribunal was more balanced than the NRC.
There is a fundamental difference between engineers and lawyers in the way they think. Engineers are taught from day one everything is a trade off. Every design is a balance of cost, function, safety, etc. Your job is to come up with a good balance.
Lawyers on the other hand are taught from day one that everything's a conflict. The way you resolve these conflicts is to have the two sides argue it out and see who wins. This requires a third party arbiter: a jury, or judge, to make that decision.
The original Atomic Energy Commission was based on the engineering paradigm. But thanks to fear of radiation stoked by anti-nuclear weapons groups led by the Rockefeller Foundation, we decided to abandon that model. We set up a conflict between the promotional side of the AEC, which became the Department of Energy, and the nuclear safety side, which became the Nuclear Regulatory Commission.
We went to the legal model. But we left out one essential feature: the third party arbiter. We gave the safety side the final decision. It's a court in which the prosecution decides on the verdict. To make it even worse, the prosecution writes the laws and can change those laws whenever it feels like it. And if there's a need to interpret those laws, the prosecution is also the judge. We set up an all-powerful tribunal whose only rule is we set the rules and told them to prevent a release.
The resulting tragically unbalanced, lopsided "conflict" quickly led to the mess we are in. Any solution must have an inherent balancing mechanism. I can think of only two possibilities:
1) An entity which is responsible for both providing reliable, affordable electricity, and doing this with an acceptable level of risk.
2) A reactive system in which the provider of electricity is forced to pay for any harm he causes.
My question is: can anybody come up with a third alternative? Remember: it has to have a built in balancing mechanism. The decisionmaker must incur a real penalty for both over-weighting safety and under-weighting safety. Good intentions and platitudes do not count.
Some related and collected thoughts on this, and my 2 cents.
Regulatory reform is one of, if not the most, important requirement for a truly global nuclear buildout. We need at least 100 GWe/year of new nuclear construction, forever. No way this can be achieved under the current glacial system. The preferred system these days of the omnipotent national regulator who has no benefit from clean power but owns all the problems of a casualty, and (in the US case) profits from delays by milking applicants, is truly an affront to common sense, good regulation, and the principles of trias politica. Following game theory the regulator is incentivized to drag out te licensing process as long as possible so as to both not own problems (paper reactors are the safest reactors) and to maximize revenues ($300/h for our time with no limit) this creates the predictably disappointing situation of lots of applicants and no running reactors. The incumbents that have survived this bureaucratic onslaught are not any better, as they take their hard-won moat for granted, again following standard game theory rules. But even the incumbents aren't building anything anymore. Everyone is simply operating according to the rules of the game and that game leads to stasis.
Not entirely sure where Jack's newfound love for the EPA comes from. They are very much anti-nuclear and use their influence as part of the licensing process (via the environmental impact statement) to stall progress. They literally make you write reports as to why this nuke has to be built, why couldn't it be a wind turbine or a solar farm, or a pretty rainbow. It is like applying for a building permit to build a shed in your garden and the municipality makes you hire an expensive engineering consultant firm to explain in great detail why it can't be a a haystack, a skyscraper, or a road-side restaurant. EPA is clearly not functioning properly.
As Jack has noted the heart of the matter is to find the balance between benefits and risks, so as to maximize societal gain. The present day system of the omnipotent regulator is exactly wrong here, completely lop-sided on the risk side of things, and to make matters worse, to employ endless red tape to achieve it. Ironically this not only pushes cost out to the point of nuclear build extinction as was already proven 40+ years ago, it also weirdly leads to lots of safety problems and lack of progress and innovation. Safety problems are created both because real safety and paperwork are not always in sync, and because the drowning in red tape disparages anyone, even utilities, from fixing real problems as the license amendment would be so draconian as to preclude it.
The hardest part is to have the powers that be acknowledge that this problem exists and that the rules of game theory don't allow progress, thus necessitating rather radical changes. If it is realized that the gold standard goes nowhere then we are there most of the way already.
What is left is what to replace it with. We seem to have options.
1. Underwriter certification as Jack prefers. Seems to work very well in ships.
2. Nationalist style regulation a la French. This worked well, the French kept cost at bay even as late as the 90s, only recently have costs escalated due to abandoning this process (privatizing and breaking up the power structure). This has proven to work. A key legal strength is people aren't opposed to each other, the same government that licenses the plants is the one that has to run them so they better make it work. It helped that the French did not allow litigation during construciton. But this approach runs into political problems. It could be palatable in some left leaning countries like Spain. In more right-bent areas like the US under Trump and now also here in Holland under the nationalists, it would be quite problematic.
3. Aviation style regulation. Better than NRC but as Jack noted there are problems.
4. Regulate it like we regulate coal plants. In most places that means BAT style regulation.
For me, the fourth way would be the most obvious way - to drop the massive double standard around nuclear that pretends it is uniquely dangerous, and treat it like any other dangerous industry. LNG terminals and oil refineries are objectively more dangerous than modern LWR designs, so in our case we should drop the national nuclear law and bring it under the fold of the EU Sevilla process. It is a bit of a palaver but the process works as all stakeholders are involved in putting the BREFs together. This is where the balance comes from - the regulator doesn't make the BREFs, it is a co-operative effort between government, industry, suppliers, engineering firms, etc. Importantly it is not possible to make changes during construction - that'd require a new round of negotiations.
Licensing would be the same as a coal plant - it'd fall under the regional government jurisdiction, or in our case recently their detachments, the industrial regulator wings (Omgevingsdiensten). I worked for one for 4 years as an industrial permit officer. They'd need to hire some nuclear engineers and safety analysts but it's not a big deal. A nuke is actually a lot simpler than a modern coal plant.
We would need to write a BREF for nuclear power. It could probably be just one document, with various subsections for the various types - water cooled, metal cooled, gas cooled, salt cooled. It would be of high importance to keep all requirements functional and precise, no focus on crippling paperwork or ALARA. So e.g. there could be a requirement for a pressure-tight containment with no more than x % leakage per day, and that it be pressure tested, and then it's up to the permit seeker to show how they will comply with the requirement. We could always deviate from the BREFs, if there was a good reason, such as a better new technology with lower emissions but that isn't in the BREFs yet. I thought it was a good system, and our industry is very safe. I would love to volunteer to help draft the BREF but the hard part is to abolish the national absolutist regulator.
https://eippcb.jrc.ec.europa.eu/reference
Disolve the NRC.
I was an original badge 100 at Comanche Peak.
The NRC was nothing but an ipediment.
They contributed NOTHING to quality and therefore safety.
The federal government declares energy a national security issue and builds the plants on military sites.
We are out of time. It has to be done now and this is the only way to move forward rapidly.