TMI Lessons: what was learned and what should have been learned

Third in a three post series on the Three Mile Island meltdown

Figure 1. The TMI Operators, all Navy vets: from left, Schiemann, Zewe, Frederick, Faust.

Lessons that were Learned

The Three Mile Island meltdown generated 300 shelf-feet of paper including two ``independent" reports:¹

1) The President's Commission Report, often called the Kemeny Report, which was funded by governmental sources outside the NRC.\cite{kemeny-1979}

2) The Rogovin Report was funded by the NRC but with ground rules designed to try and make this an outsiders take.\cite{rogovin-1980}

Both are products of the establishment. Neither qualify as totally independent, and both tread very lightly when it comes to accountability. But both are factual and apparently honest efforts.

The most important lesson that these reports took from Three Mile Island is nuclear power is unacceptably unsafe. Casualties such as TMI, which produced no detectable radiation harm to the public, are both intolerable and preventable. The Rogovin Report takes this as axiomatic.

Barring basic changes in the industry and the way it is regulated, it could happen again. The fact that it happened here at Three Mile Island, in the glare of media spotlights, may be the best insurance that it will not recur.\cite{rogovin-1980}[p 5]

Here's how the President's Commission puts it.

To prevent nuclear accidents as serious as Three Mile Island, fundamental changes will be necessary in the organization and practices --- and above all --- in the attitudes of the Nuclear Regulatory Commission and, to the extent that the institutions we investigated are typical, of the nuclear industry.\cite{kemeny-1979}[p 7]

The entire sentence is set off and underlined to indicate its importance. But the President's Commission adds these fundamental changes may not be enough to make nuclear power safe. ``We do not claim that are proposed recommendations are sufficient to assure the safety of nuclear power." The word fundamental is repeated over and over again.

unless portions of the industry and its regulatory agency undergo fundamental changes, they will overtime totally destroy public confidence and hence they will be responsible for the elimination of nuclear power as a viable source of energy.\cite{kemeny-1979}[p 25] [Emphasis in the original.]

But the fundamental changes these reports recommend don't turn out to be very fundamental. Changes to the NRC organization chart, replace commission with a single administrator, move most NRC employees into a single building or campus, have a degreed nuclear engineer on watch.² Moving organization charts around does not change the basic bureaucratic incentives. It does not align NRC goals with society's need for cheap, dispatchable, low pollution, low CO2 electricity at a tolerable level of risk.

According to both the Kemeny and Rogovin report, the NRC is a failure:

With its present organization, staff, and attitudes, the NRC is unable to fulfill its responsibility for providing an acceptable level of safety for nuclear power plants.\cite{kemeny-1979}[p 56]

and the regulations are counter-productive

The existence of a vast body of regulations by NRC tends to focus industry attention on the meeting of regulations rather than on a systematic concern for safety.\cite{kemeny-1979}[p 20]

But the solution is more regulators and more procedures, and most importantly more paperwork. The safety-critical QA program failed, so we need a more comprehensive QA program. The single failure based Probabilistic Risk Assessment(PRA) failed, so we need more comprehensive PRA. This illustrates the First Rule of Bureaucratic Expansion: screw up and you get bigger. In a truly competitive market, screw up and you disappear.

More of the same was exactly what we got. There were no fundamental changes. At the end of the day, nothing much was learned, at least by the nuclear establishment.

The Lessons that should have been learned

The Two Lies are Lies

This should have been the big take away. Three Mile Island brazenly and unmistakably exposed the nuclear establishment's two big lies.

1. The Negligible Probability Lie. The probability of a significant release is not negligible. The probability of the next release is 1.00. It is inevitable. It is only a matter of when.

The usual way that the nuclear establishment tells the Negligible Probability Lie is they throw out a bogusly small frequency number, say 1 in 17000 reactor years, expecting the public to interpret this as a ``no need to worry about it" number. But in a fully decarbonized, all nuclear world we will need at least 25,000 large reactors. So even if the 1 in 17000 number were correct, we could expect a TMI-or-larger release about once a year. The actual performance to date is about 1 TMI+ release every 4000 reactor years, in which case we are talking about roughly 6 TMI+ releases a year.

2. The Intolerable Harm Lie. The radiation harm associated with a release is not intolerable. In this case, if there was any harm, it was far too small to be detected. If the TMI release had been more than 1000 times larger as it was at Windscale, there would have been no detectable harm. If the TMI release had been 300,000 times larger as it was at Fukushima, there would still have been no detectable public harm due to radiation.

At one point, the Rogovin Report appears to realize this:

Just as the regulators must change their attitudes to appreciate that this [the public's] perception of risk cannot be dealt with by trying to convince the public that it ``can't happen", so renewed efforts must be made to educate that the risks and benefits of nuclear power must be weighed against the very real health and environmental risks associated with other forms of power generation.\cite{rogovin-1980}[p 91]

But there is no follow up. Nor does this insight show up in any of their recommendations. The problem is, if the two lies are false, then there is no need for an NRC. And neither the Rogovin nor the Kemeny report is going to go there.

But MetEd in its own way did. The utility filed a four billion dollar suit against the NRC alleging the NRC's failure to tell MetEd about what happened at Davis Besse caused the loss. The way this works is MetEd first had to file a complaint with the NRC. The NRC Commission rejected the complaint on the grounds that it is not responsible for what happens at a nuclear power plant.

The commission does not, thereby, certify to the industry that the industry's designs and procedures are adequate to protect its equipment or operations,\cite{upi-1981}

This rejection allowed MetEd to go to the Courts. The court came up with a different out. It turns out the Davis Besse loss of feedwater was listed in a routine monthly Licensee Event Report that the NRC sends out to all the plants. According to the court, that's all it had to do.

When an agency determines the amount of information necessary to fulfill its regulatory mission, it is exercising the essence of its discretionary function.\cite{yorkdailyrecord-1984}

Let me get this straight. The NRC admits it is not responsible for nuclear plant safety. And the courts say the NRC has the discretionary power not to tell the plants that the training that the NRC has approved and required is both wrong and dangerously misleading. What does it have to do?

To survive, the NRC must spread fear and then sell a bogus solution to that bogus fear. To survive, the NRC must promulgate the Two Lies, even if events like Three Mile Island prove both are false. That's what it has to do and continued to do.

People who screw up must pay

Accountability shows up almost no where in the Kemeny and Rogovin reports. There is one exception. The TMI Reactor Operators lost their licenses and had their careers ruined, for doing what they had been trained to do.³ And when that failed, they did a pretty good job of coping with the resulting mess. They were not culprits; they were scapegoats.

The real culprits went unscathed. Joseph Fouchard, the NRC PR man who apparently came up with the radiation shine lie, and the three Commissioners Gilensky, Bradford, and Kennedy who signed off on it should have been fired with cause, no pensions. Everybody in the Office of Inspection and Enforcement (OIE) who sat on Creswell's complaints, should have been fired for cause. Everybody at NRC who sat on the Michelson report should have been fired for cause. The authors of the specious OIE report that blamed the operators, all fired for cause. Roger Mattson, the author of the hydrogen explosion panic, should have been fired for gross incompetence.⁴

B&W bears immense responsibility. They were in the best position to understand the problem and the solution. And parts of B&W did just that. At least Karrasch, Hallman, and Walters should have been out on the street immediately. Their bosses as well. At the corporate level B&W should have borne a large portion of the costs of Three Mile Island. The utility did sue B&W; but the suit was settled out of court for a measly 47 million dollars. I do not understand how that could have happened.⁵

B&W had a moral responsibility to inform their customers of a problem they had clear knowledge of. But they did not, just as Westinghouse did not earlier. If B&W had gone bankrupt over their failure to do the right thing, it would be the last time that a nuclear vendor pulled this stunt.

Lastly the MetEd shareholders should have borne the portion of the cost of Three Mile Island that B&W did not.⁶ They had bet on and voted in a management that lost a multi-billion dollar plant. All that management should have been out on the street. No pensions. No deferred compensation.⁷

The group that should have borne none of the costs is the MetEd rate payers, who had nothing to do with all the screw ups. In fact, they shouldered the bulk of the costs in increased electricity prices.

If people/corporations/bureaucracies who do not live up to their responsibilities, do not pay for their mistakes, then those mistakes will be repeated.

The NRC is a Farce and must be done away with.

This brings us to the NRC. Both the Kemeny and Rogovin Reports document the ludicrous incompetence of the NRC, which on one occasion descended into an outright lie, a lie that was preposterously stupid. These guys should not be in charge of anything. It is not so much that the individuals were stupid, although there is considerably evidence that some of them were unfit for their jobs; but that bureaucratic incentives led just about all of them to do stupid, and sometimes evil, things.

In its response to the MetEd suit, the NRC claimed it is not responsible for what happens at nuclear plants. By its own admission, it has no function. The lesson that should have been learned is that the NRC should be euthanized and replaced with an entirely different system. That would be a fundamental change.

Once you renounce the Two Lies, you realize that nuclear power can be regulated like other hazardous, beneficial activities, such as ocean transportation or high pressure steam. GKN calls this system Underwriter Certification. Even forgetting about radiation, a nuclear power plant loss is an exceedingly expensive industrial casualty as TMI proved. If the cost of such a casualty is borne by the people responsible for that loss, then those people will work very hard to insure that those casualties are very rare. And they are the best people to figure out how to do just that.

1. I'm going to skip over the obvious technical lessons: the need to change the manuals, the training, and the simulators to reflect pressurizer leak reality, and the need to improve control room design.

The release also showed that many of the assumptions underlying the NRC's models of what happens in a meltdown were unrealistically conservative. At TMI only a tiny amount of iodine and essentially no cesium were discharged. Instead they condensed out on internal surfaces in the plant. According to the models, a sizable portion of the plant's inventory of both should have been emitted.

2. Davis Besse, where essentially the same scenario happened, there were five operators present. Two of them were degreed engineers, one with a masters. The operator who figured out what was happening was a high school grad.

3. This was based on an NRC Office of Inspection and Enforcement Report which pinned the blame squarely on the operators.\cite{nrc-oie} The problem was their mindset.

There is considerable evidence of a ``mindset", that overfilling the reactor cooling system (making the system solid) was to be avoided at almost any cost. Undue attention by the TMI operators to avoiding a solid system led them to ignore other procedural instructions and indications that the core was not being properly cooled.\cite{nrc-oie}[p 2]

Nowhere in this 800 page exercise in unapologetic deflection is there any admission that this strange mindset was a product of NRC approved and required training, nor that the NRC had ample, multiple warnings that the training was dangerously wrong.

4. In a bit of exquisite irony, Mattson was put in charge of the TMI-lesson-learned team, completing a rare quadfecta: judge, jury, executioner, and perpetrator.

5. The situation was complicated by the fact that B&W was ``just" the supplier of the nuclear steam supply system. The plant was actually designed by Burns and Roe, an engineering firm who also managed the construction. Under NRC rules the utility was the licensee, and the licensee is solely responsible for whatever happened at the plant. A better system would be turnkey construction of the plant and joint responsibility between the plant builder and the plant operator, much like commercial air transportation.

6. Burns and Roe shares some culpability. Among other things, they were responsible for a poorly laid out control room, which went bananas as soon as the system started to break down.

7. I need to be fair. The MetEd President, Walter Creitz, was demoted to highly paid Advisor.