Learning from Crisis: An Interview with Three Mile Island Historian J. Samuel Walker

In this interview, Jowanza talks to J. Samuel Walker to uncover the story behind the headlines of the Three Mile Island accident—from the control room confusion to the high-stakes decisions that defined the response. Walker provides a rare, insider's account of the human factors, communication breakdowns, and enduring myths that continue to shape our understanding of this pivotal moment in industrial history.

Background

On March 28, 1979, a stuck valve at Pennsylvania’s Three Mile Island triggered the most consequential accident in U.S. commercial nuclear power. What began as a routine mechanical malfunction spiraled into the nation's worst commercial nuclear accident, bringing a reactor core within hours of a catastrophic meltdown and exposing fundamental flaws in both technology and human preparedness.

The partial meltdown at Unit 2, located on the Susquehanna River near Harrisburg, Pennsylvania, lasted five harrowing days and released radioactive gases into the environment. While no one died and radiation exposure remained minimal, the psychological and political fallout was seismic. The accident shattered public confidence in nuclear energy, triggered the cancellation of dozens of planned reactors, and effectively ended the nuclear industry's golden age of expansion.

Through Governor Thornburgh’s excruciating evacuation calculus and Harold Denton’s steady on‑site leadership, the conversation reveals why information gaps—not secrecy—drove uncertainty, and how those lessons rewired operator training, control‑room design, and crisis communication across the nuclear industry.

When America's Nuclear Dream Nearly Became a Nightmare: Revisiting Three Mile Island Through the Eyes of Its Definitive Historian

Nearly five decades later, the lessons of Three Mile Island remain as relevant as ever.

This conversation offers rare insights into the human factors, regulatory failures, and communication breakdowns that transformed a mechanical malfunction into a national crisis. More importantly, it reveals how the industry's response to Three Mile Island—from enhanced training protocols to improved control room design—fundamentally reshaped nuclear safety culture and continues to influence how we manage complex, high-stakes industrial systems today.

Jowanza: Before we dive into Three Mile Island, I want our readers to understand the unique perspective you bring to this conversation. Could you share about your background and how you came to write the definitive book on this crisis?

J. Samuel Walker: My background is as a professional historian. I earned a PhD in American diplomatic history, which has nothing to do with Three Mile Island. But I wound up working for the Nuclear Regulatory Commission because I couldn't find an academic job—there was an academic job crisis just when I finished my PhD, which still continues.

I started my job with the NRC just a few weeks after the Three Mile Island accident. At the time, I didn't have any training or background in nuclear technology or nuclear power. So I had a steep learning curve, and as I read lots of books and articles coming out about Three Mile Island, I thought, "How will I ever write a book about this?" Well, 25 years later, my book came out, and by that time I had learned a whole lot as the historian of the Nuclear Regulatory Commission.

Jowanza: That's quite a journey. Now, let's talk about March 28th, 1979. Could you walk us through what actually happened at Three Mile Island?

J.S. Walker: There was a whole series of malfunctions, none of which were especially serious. But the one that really caused the accident was that a valve stuck open—this was not unprecedented, but it was not recognized by the operators. As a result of that valve sticking open, thousands of gallons of cooling water ran out of the core.
The core has to be kept cool with water. Once that valve stuck open, the cooling was severely compromised, and eventually enough water ran out that the core was exposed—the last thing you want to have happen. That raised the genuine possibility of a fuel meltdown and, in worst case, a breach of the containment building.

Jowanza: What was surprising about this crisis to the technicians and engineers working on it?

J.S. Walker: The problem they had was that all kinds of alarms were sounding, all kinds of lights were flashing in the control room, but there was no single indication that showed what was happening. They knew something was happening and that it was getting increasingly serious, but they really didn't have the training to figure out exactly what had happened.

The operators were concerned about something different than what was actually happening. They were worried about the pressurizer going solid—having too much water in it. So they took steps that were appropriate if the problem had been the pressurizer, but were exactly the wrong thing for a loss of coolant accident. They shut off the emergency core cooling system and the reactor coolant pumps, which reduced the flow of coolant water to a trickle.

We found out later that within a fairly short amount of time, about half the core melted within the pressure vessel.

The Human Factor Challenge

Jowanza: What was the regulatory culture at the NRC during this time, and how did it contribute to this crisis?

J.S. Walker: The NRC was fairly new—it had only been created in 1975—but both the NRC and its predecessor agency, the Atomic Energy Commission, enforced tight regulations. The problem we learned about at Three Mile Island was that they didn't pay enough attention to what became called "human factors."

This included training of operators and design of control rooms so that you could get useful information if you had a major mishap. The problem wasn't a lack of strict regulations, but a lack of recognition of the possible consequences of human error and equipment that wasn't useful for what it was intended for.
There was no single piece of equipment in the control room that showed that the core was losing coolant. There wasn't anything like a gas gauge in a car that showed you the level of coolant in the core.

Jowanza: What kind of training was required to be an operator at that time?

J.S. Walker: Most operators had their first training in the nuclear Navy and received rigorous training there. The operators at TMI had passed all their qualification exams from the NRC and done better than average. So it's not as though these people were incapable—they just didn't have the equipment they needed to figure out what was going on. They got increasingly confused because there were conflicting signs and increasingly frustrated because something obviously was happening.

Communication in Crisis

Jowanza: How did the communication process work as this disaster unfolded? How quickly did information reach the public?

J.S. Walker: It was never a secret, but it was not well planned for. The utility, Metropolitan Edison, was the first to go public and kind of blew it off. They said, "Oh, no big deal. We've lost a couple of fuel rods that seem to have overheated, but the plant will be back in service within a week."

That didn't sound right to either Governor Thornberg or to the NRC, and it quickly turned out that those reports were vast. The NRC tried to provide and did provide information based on what it knew. The problem for the NRC, for the public and for the governor was that the information kept changing, and people didn't know exactly what it was. They didn't know exactly how serious it was.

So they issued press releases based on what they knew, but oftentimes, within a couple of hours, they knew something different than what they had put out publicly a short time before. So communication was a problem, not because anything was secret or the NRC or the utility or the governor was trying to hide anything. It's just that information about the condition of the plant was almost a complete mystery.

They simply didn't know the extent of the damage, what had caused the damage, and what was going on inside the reactor. They knew that there were very high levels of radiation inside the containment building, extraordinarily high, and so they realized they had something serious, but they didn't know the cause.

And it took at least a couple of days for them to figure this out, that what they had on their hands was a loss of coolant accident. Then they started figuring out a way to bring that reactor back to a stable state. And that was not obvious. There was no obvious solution for dealing with that because of the damage that the reactor and the safety systems had suffered.

Jowanza: How did the public perceive the communication, whether it was from the governor or the utility or NRC? Was it a really chaotic period?

J.S. Walker: Well, the public was concerned with good reason to be concerned, but they were amazingly calm. And that's one thing that I deal with in my book was that there was concern and they weren't sure what to do. And some people left their homes and went elsewhere fairly quickly within a couple of days.

So there was growing concern and there was growing recognition that we have something very serious on our hands. But the people of Central Pennsylvania were amazingly calm and amazingly rational, didn't have people panicking and doing crazy things that people do when they panic. And partly I think that's a result of their conservative nature, but more of faith in Governor Thornberg, number one.

And also faith in the NRC's head of staff at the plant who was sent up there on Friday. The accident occurred on a Wednesday. Friday, the NRC sent a large contingent of experts up there. And the head of that effort was a man named Harold Denton, who was a high-level official at the NRC, and who quickly won the confidence of Governor Thornberg, which wasn't easy to do at that point because he had had unfortunate communications with the NRC and with the public.

And the reasons were: First, Denton was credible. Second, he was honest. If a reporter asked him a question and he didn't know the answer, he said, "I don't know. This is what I think, but we're in a situation we've never had before, and this is my best judgment, but I can't be certain, and we're doing the best we can."

And just Denton's down-to-earth mannerisms and his frankness were reassuring to people. That was extremely important to keep people from panicking and to reassure them, at the very least. No one was making any guarantees about what might happen, to reassure them at least that people were working on it and that they were doing the best they could to figure out what had happened and to make certain there was no significant release of radiation.

Jowanza: It sounds like, from your perspective, maybe that was one of the best things that happened, that there was institutional readiness and preparedness for this public response.

J.S. Walker: Well, yeah, Denton became a hero in Central Pennsylvania. The NRC was not viewed in general as being heroic, but Denton was. And the faith that they had in him and that Thornberg had in him were essential. But the NRC was not terribly well prepared. It wasn't until a couple of days after the accident occurred that their professional public affairs officers arrived on the scene to explain the situation.

This caused some problems because junior staff members were making assessments about the damage or lack of damage they perceived, which turned out to be false. And so the NRC was not totally prepared for the kind of accident that occurred, in part because they never expected an accident like this to happen.

Lasting Impact and Policy Changes

Jowanza: How would you say that the Three Mile Island disaster changed the way that nuclear energy and nuclear policy were regulated in the U.S. after this disaster?

J.S. Walker: Well, after the accident, there was a whole lot more emphasis placed on human factors. That was the first and most significant improvement made. There were also equipment problems, and the accident revealed a lot of things that no one had really expected.

A thorough reexamination was conducted to identify possible causes of accidents and the factors most likely to occur. And so the whole field of risk assessment was reassessed for nuclear power after Three Mile Island, and conclusions were reached. Many improvements were made. The industry has improved its safety record and its performance record dramatically since Three Mile Island, partly as a result of Three Mile Island.

What was also important was that there was recognition of what the effects of the fear of a severe accident might be. And one of the good aspects of the accident was that despite the damage done to the pressure vessel, even though you had hundreds of tons of fuel rods that melted and that gathered in the bottom of the pressure vessel, there was not a significant release of radiation at Three Mile Island.

And that's what distinguishes Three Mile Island from Chernobyl and Fukushima, which did have large releases. That did not happen at Three Mile Island. Three Mile Island is the worst accident in the history of commercial nuclear power in the U.S.. Still, it didn't come close to the damage that was done and the amount of radiation, vast amounts of radiation that were released at Chernobyl and Fukushima.

Jowanza: It sounds like human factors really is the biggest lesson learned and kind of the legacy of it.

J.S. Walker: Yeah, and there was a clear recognition after the accident that we have to give the operators the information that they need and policymakers the information they need to make decisions. And if you read the transcripts of phone calls and meetings and what was said at press conferences and elsewhere, the great difficulty was that they didn't have enough information about what was going on in the reactor.

The Evacuation Decision

Jowanza: I'm fascinated by this communication and back and forth between NRC and the governor here. As far as an evacuation is concerned, was that an NRC decision, or is that entirely on the governor's side to decide?

J.S. Walker: Well, it was the governor's decision, and it was not a presidential decision. President Carter was very much involved. He gave Denton at the site a red phone, and Denton had orders to call the White House twice a day to explain what was going on, what he knew, and what the prospects were.

But in the end, it was not a decision of the president. It was not a decision of the NRC. It was a decision of the governor. And Governor Thornberg, it was Thornberg's power and his decision of whether or not to order an evacuation. And that's a life and death decision because he had to decide whether or not to order a mandatory evacuation of a population surrounding the plant, perhaps out to 20 miles or not.

And if you don't order an evacuation and the plant fails and you have a major release of radiation, if the containment structure fails, then people are going to lose their lives. On the other hand, if you do order an evacuation, within a 20-mile radius, you've got 13 hospitals, a population of 600,000, a prison, and the difficulties of executing that kind of evacuation, people are going to get killed that way.

And we've had plenty of examples of that since Three Mile Island, where evacuations have caused serious loss of life. So Thornberg was faced with this excruciating decision, and he kept saying, "I don't have enough information. I don't have enough information to go on." And he didn't.

The Hydrogen Bubble Crisis

J.S. Walker: And the great crisis and source of most concern by the third day of the accident was that a hydrogen bubble had formed in the upper parts of the pressure vessel. And the concern first was the hydrogen bubble might inhibit cooling because it was in the way of efforts to cool the core. But then the concern became, well, what if the hydrogen bubble is explosive or flammable?

And if that happens, can it cause inflammation or in worst case, an explosion that would rupture the pressure vessel? And then you've got this ruptured or this very hot core in a containment building, and that increased the chances of a breach of containment by uncomfortable proportions.

From the high levels of radiation in the pressure vessel, this through the process of radiolysis, which breaks down water into hydrogen and oxygen. So you had a lot of hydrogen that was created in the pressure vessel, and you also had oxygen, and no one knew how much oxygen, and you need both to have a flammable or an explosive condition.

Eventually, the crisis of the accident, the five days of crisis, ended when experts determined that there was no possibility that there would be enough free oxygen to create a flammable or an explosive condition. And then, at least for the acute part of the crisis, it ended.

Jowanza: This was something that required five days of investigation rather than a quick answer?

J.S. Walker: Yeah, that's a good question. One problem that you had was that all the indications were destroyed by the accident. But no way was built into the plant to measure free oxygen because no one anticipated that you would have an accident that would release vast amounts of hydrogen and perhaps some amounts of free oxygen.

The condition was unanticipated, and it was a source of great worry, as the extent of its danger was unknown. So you're dealing with a worst-case scenario where you might lose a pressure vessel. If you lose a pressure vessel, then that increases the chances that containment could fail.

Lessons for Industrial Safety

Jowanza: What lessons should people who are operating in an industrial context carry forward with them from the lessons of Three Mile Island?

J.S. Walker: Yeah, that's always a good question, and the answers in some ways are obvious. You can't be complacent. In some ways, the NRC and the industry thought that they had solved all the problems of reactor safety, and they never really anticipated the kind of accident that occurred, the series of errors that occurred, compounded by operator error, that caused the accident at Three Mile Island.

And I guess another lesson is to overdesign as much as possible. There's always a question of cost versus safety considerations, and that's a difficult balance. But one reason that the pressure vessel did not fail at Three Mile Island was that during the 1960s, the AEC, the NRC's predecessor, had imposed stringent guidelines, very strict rules, I should say, on how sturdy the pressure vessel should be.

And the industry, which was used to building pressure vessels for coal plants, said, "No, this is ridiculous. The AEC should not impose these kinds of requirements, which are much stricter than those for coal plants." And the AEC said, "Well, yeah, but you could have a catastrophic accident at a nuclear power plant that you can't have at a coal plant, and so you are going to do this."

And one reason the pressure vessel held up at Three Mile Island was that it had an extra margin of safety. So those are always tough calls for engineers and for designers, but it's essential to recognize what you're dealing with and what the risks are and how you assess those risks.

Common Misconceptions

Jowanza: What would you say are some of the most misunderstood aspects of Three Mile Island?

J.S. Walker: Well, yeah, I can give you the biggie, and the biggie is that numerous books and television programs claimed erroneously, completely erroneously, that the plant was on the verge of an explosion that was going to blow the roof off the plant, and that despite knowing that, the NRC allowed President Carter to visit, knowing that an explosion of the roof off the containment building would contaminate an area the size of Pennsylvania and endanger the life of the President of the United States.

That's not at all what anyone worried about. No one was concerned about that who was in a knowledgeable position. They were anxious, as I mentioned earlier, about losing the pressure vessel, and that would increase the chances of a breach of containment. But no one thought, with good reason, that something would happen to blow the roof off the containment building.

So there's been a lot of erroneous reporting about what the dangers were and very exaggerated reporting about what could have happened at Three Mile Island. And it's simply wrong to say that's what people worried about or that's what could have happened. It just was not going to happen.

The full conversation is available on Industrial Risk: Beyond the Blueprint podcast.