Peter Lam's resume  reflects a lifetime of experience in the nuclear energy industry–including 20 years in the private sector, followed by 18 years as an administrative judge at the U.S. Nuclear Regulatory Commission .
He's a retired nuclear engineer with 110 published judicial decisions and more than 70 technical papers in industry journals and company publications. And he's considered an international expert on nuclear reactor safety and risk assessment strategies.
So nuclear opponents were stunned last year, when Lam revealed how the Fukushima Daiichi meltdown had changed his views on the importance of accident probabilities—a key tenet of America's nuclear safety policy.
In a presentation  before the California Energy Commission  in July 2011, Lam raised questions about the NRC's reliance on "likelihood calculations" to guide its safety and plant design regulations. He said the industry practice of not planning for statistically improbable accident scenarios—like the disasters that struck Fukushima—could be catastrophic and needed to end.
The Fukushima calamity involved two of the industry's five most dangerous but "extremely unlikely" nuclear events, Lam said. It also included multiple nuclear reactor core meltdowns—a "Black Swan" scenario never contemplated because it was deemed impossible.
"Probability dismissal is not an exact science. By [using] it, one can be very, very wrong," Lam told the commission.
The frankness of Lam's post-Fukushima assessment surprised and impressed Rochelle Becker, executive director of the Alliance for Nuclear Responsibility , a California group focused on nuclear power plant cost and reliability issues.
"I was absolutely amazed. This was not a man who I ever thought would say anything like that," said Becker. "He was a firm believer in probabilistic risk assessment until Fukushima, and he changed his mind."
Lam elaborated on those views in a March paper , where he endorsed beefed up oversight and safety enforcement and warned that "a simple dismissal [of extreme accidents] based on seemingly valid analysis should not be relied upon again."
His apparent willingness to reassess long-held industry beliefs convinced Becker to write a letter earlier this year supporting Lam's reappointment to the Diablo Canyon Independent Safety Committee . That panel has kept watch over the Central California nuclear plant since 1990. Lam was recently named committee chairman and confirmed to serve a second three-year term.
The committee is made up of three industry experts, and it has sometimes been criticized by Becker and others as toothless and too cozy with Diablo Canyon operator Pacific Gas & Electric Co . But Lam, 65, has thus far displayed a judicial temperament more open to opposing arguments and evidence. Recently, the California attorney general's office discussed the possibility of adding a similar committee  to focus on San Onofre, the troubled nuclear plant in Southern California.
In a telephone interview, Lam talked about his post-Fukushima views on the risks of nuclear power plants, the Diablo Canyon Independent Safety Committee and whether a similar committee would be useful for communities near other nuclear plants.
ICN: Fukushima has galvanized concerns about the safety of nuclear plants in the United States and around the world. What was your reaction when it happened?
Lam: One can plan for a lot of things, but things don't always happen according to what you plan for. ... Sure, I think everybody's doing the right thing globally to make sure if we're hit by a major earthquake, things don't collapse on us, and if we're hit by an earthquake followed by a tsunami, electrical equipment does not get wiped out.
But I also caution everybody ... let us make sure we are not fighting the last war. The next time we are surprised by a major nuclear reactor accident, it may not be a tsunami. Therefore, let us go back and look, and look, and look to see what else is out there. Nuclear power is an unforgiving technology. Everybody needs to recognize that.
ICN: When you look at a nuclear plant now, how do you assess risk?
Lam: I don't speak for my colleagues [on the safety committee] ... but in my opinion, there are only two ways. One is called deterministic. The other one is probabilistic. I, for one, use both.
Deterministic [risk assessment] is saying, to the extent possible, one should preset a lot of parameters for how to design, manufacture, install, operate and maintain the plant. These parameters should be predetermined from day one. That is the old way the U.S. NRC regulated [nuclear plants].
For example, say I'm building a hotel with a conference room. The room is large enough for 1,000 people, and I assume everyone weighs 250 pounds ... So maybe we build a room that can take 2,000 people weighing 250 pounds ... Now I have reasonable assurance that this thing is not going to collapse.
ICN: And probabilistic risk assessment [now part of the NRC's 'risk-informed and performance-based' approach to safety planning]?
Lam: Probabilistic is, 'alright, outside of these predetermined parameters, there's something we have not thought about.' Of those we have not thought about, let us stack them up in terms of these things: One, how likely are they to occur? Two, how major are the consequences? And three, how effective are the remedies? All three need to come together.
For example, a meteorite comes in once every 5 million years, and nothing can defend against that, so I'm not going to consider that. Ah, but the flooding, once in 100 years, you'd better design against that. It's all a balance of how effective [the prevention measure] is and how cost-beneficial it is.
Looking at Fukushima, all they needed to do to avoid disaster was to elevate one or two diesel generators above the 50-foot tsunami level. And the cost, if they were to do it, would probably have been $10 million to $30 million. I'm just guessing, but let's say it's $100 million. If you knew about the tsunami coming, you would have spent the $100 million. Now [cleanup is] going to cost $100 billion, easy. So it's a matter of, 'Well, tell me what's going to happen? How much is it going to cost?" (Editor's note: The Japan Center for Economic Research put the accident's cost at between $71 billion and $250 billion.)
To be fair, a lot of things fall into that category besides a tsunami. And that's where probabilistic risk assessment comes in. What about a hurricane? The water ... would it hurt my diesel generators? Would it collapse my building? The building may not collapse on you, but if something else collapses and falls on a cable and cuts it, you are in equal jeopardy. What about a solar flare? These are the types of deliberations we need to go through.
ICN: Some people believe it's impossible to adequately protect the public from nuclear mishaps, and that belief forms the basis for opposing the technology altogether. What do you think?
Lam: They raise an interesting point. Are there any alternatives to this technology, so that if an accident were to happen, it would not cause so much harm? I think they're probably coming.
I would say that one should be really open minded as to what the future will bring, with or without nuclear power. And before we get there, our number one task is whatever nuclear power plants we have running, let's make sure they are safe.
ICN: Has the independent safety committee for Diablo Canyon affected things like safety culture at the plant?
Lam: I think the committee may play a part in enhancing that safety culture since three of us go down there nine times a year. We meet all the senior managers; we meet the rank and file, middle managers and first-line people. They probably know that the committee is there. They would not hesitate to tell us things that don't get appropriate attention.
The [NRC has] a program for whistleblowers. But perhaps we are a little bit more informal. We hear things. I usually meet with the site vice president behind closed doors. I would definitely let him know what I hear unless the people who told me things request confidential agreement. My job is to make sure, to the extent that we can contribute, that the plant operates safely.
ICN: As you know, your committee colleague, Robert Budnitz, was asked by the attorney general's office about a committee for San Onofre . Have you been contacted by people from elsewhere about setting up a similar committee?
Lam: There are some private discussions elsewhere, yes. But not officially. There are people who wish to remain anonymous who talked with me about setting up an independent committee. But I would not put much weight on it as an indication that there is a strong interest elsewhere.
ICN: So are other areas missing out by not having an independent safety committee?
Lam: One can make an argument both opposing and proposing it. The proponent's argument is, 'a nuclear power plant typically generates $2 million to $3 million a day of revenue. If a committee, for less than $1 million a year, which is half a day's revenue, if it can improve reactor safety, then on a cost-benefit ratio, it's certainly worth it.' The opponent's argument would be, 'come on, I have the NRC, I have the Institute of Nuclear Power Operations, I have the Electric Power Research Institute, and I have in-house nuclear oversight board.' Really, both sides have an argument.
I do not know any other forum that provides this type of venue for public discussion. The NRC usually does public outreach at different locations. I do not think they do it three times a year. Furthermore, the NRC, as a regulator, would be hard pressed to have a free-ranging discussion as we do in a public meeting.
If you are member of the public, you probably would opt for an independent safety committee, because at least you can come to talk to the committee, once, twice or three times a year, or even tour the plant—and also get feedback from three committee members.
ICN: So you are not taking a position on it?
Lam: No. Each locality needs to make their own opinion based on the facts and based on their perception. If it improves public participation, if it improves confidence among members of the public, or for that matter, if it increases opposition—I think that could be healthy.
Slides from Lam's July 26, 2011, presentation: