Admiral Rickover's Navy Has Run Reactors at Sea for Seventy Years

The most demanding job interview in American history, and I say this with respect for every law school dean, was not held at a corporation or a university. It was held by a Navy officer who personally questioned the young men who wanted to run his reactors, one at a time, and who by one account personally interviewed every prospective officer considered for a nuclear ship, tens of thousands of them over the years. He also built the naval reactor program, which says it has operated reactors at sea for seventy years without a reactor accident.

That officer was Admiral Hyman Rickover. This post is a profile of a public career and an argument drawn from it. The argument is that nuclear safety is not a gift of luck or of physics. It is a product of culture, training and standards, and Rickover's navy is the best proof we have that a free country can build all three.

Defining the terms

Naval nuclear propulsion means ships and submarines driven by a nuclear reactor that makes steam to turn the propellers. A nuclear-powered submarine can stay submerged for as long as the crew can be fed, because it does not need air for combustion and does not need to refuel often.

A reactor accident, in the sense used here, means damage to a reactor core or a release of radioactivity from the reactor plant that harms people or the environment. It is different from the loss of a ship for other reasons.

Safety culture is the set of habits, rules and attitudes that determine whether people follow procedures, report mistakes and fix them. It cannot be photographed, and it is the thing that matters most.

The man and his charge

Hyman G. Rickover was born in 1900 in what was then part of the Russian Empire, in a Polish town, and came to the United States as a child. He graduated from the Naval Academy in 1922 and made his career in the Navy as an engineer. He rose to flag rank and served on active duty for an unusually long time, until he was retired in 1982.

In the years just after World War II, the question facing the Navy was whether a nuclear reactor could be shrunk to fit inside a ship. Many engineers thought it was a distant hope. Rickover thought it was an engineering problem that could be solved. He was sent to Oak Ridge in 1946 to work on a nuclear power plant, and from 1949 he was running naval reactors for the Navy and, through a second hat, for the Atomic Energy Commission, which gave him authority in both worlds. That dual role, one small office answerable for both the Navy's needs and the government's nuclear program, was unusual, and it was central to his success. Responsibility was not split between agencies that could blame each other.

From a paper reactor to a real one

There are two kinds of reactor in every engineer's imagination. The paper reactor is simple, small, cheap and quickly built, and it exists only in the mind of the designer. The practical reactor is the opposite: complicated, expensive, late and always full of problems nobody thought of in advance. The discipline that mattered was to treat the practical difficulties as the main subject and not as an inconvenience to be waved away later. Every engineer who has watched a project built on a slide deck collapse on contact with a welding shop knows what he meant.

The proof of the approach came quickly. The submarine USS Nautilus was launched on January 21, 1954, and commissioned that September. On January 17, 1955, she got underway on nuclear power. In August 1958 she became the first submarine to pass submerged beneath the North Pole.

Shippingport: the civilian proof

The same office built the first full-scale nuclear power plant for civilian electricity. The Shippingport Atomic Power Station, on the Ohio River in Pennsylvania, first produced electrical power on December 18, 1957, and the Nuclear Regulatory Commission describes it as the world's first full-scale atomic electric power plant devoted exclusively to peacetime uses. It was a pressurized water reactor supplied by Naval Reactors and Westinghouse, the same family of design that became common in American plants, and the Navy's work gave the civilian industry a tested template.

It is worth pausing on this. The conservative case for nuclear power often starts with the observation that the technology worked, early and well, under a disciplined institution. The pressurized water reactor, a type that became common in American commercial plants, was the type in use at Shippingport. When people say that nuclear power is untested, the answer is that the Navy tested it at sea, under the worst conditions, with young crews, for decades.

Rickover's rules, from the public record

How did he do it? The public record, drawn from his own testimony and writings and from the accounts of those who served under him, points to a small number of practices.

1. Personal selection. Rickover interviewed prospective officers himself. The interview was a test of knowledge, composure and honesty, and he used it to find the people who would take responsibility. The Navy's nuclear officer pipeline kept his selectivity long after he left.
2. Individual accountability. He insisted that a person, not a committee, answered for each job. If a reactor plant had a problem, the question was who was responsible, and the answer had a name attached. Committees diffuse blame, and diffused blame is an invitation to carelessness.
3. Technical competence at the top. Those who managed nuclear work were required to understand it. He distrusted managers who relied on summaries without grasping the details.
4. Facing the facts. He demanded that problems be reported accurately and promptly, not softened for the chain of command. The culture rewarded bad news delivered early.
5. Training that never stops. Crews trained constantly, qualified formally and requalified, drilling on casualties until the responses were automatic.
6. Written procedures followed exactly, and a questioning attitude toward everything else. The rule was to follow the procedure and to stop and ask when the procedure did not fit the situation.
7. Learning from every incident. Problems anywhere in the fleet were studied and the lessons sent everywhere.

None of these is exotic, and that is the point. They are habits of a serious organization, and they are what separates a safety record from a safety slogan.

The record, and the honest caveats

The Navy says it has operated its nuclear-powered ships for seventy years, from 1955, without a reactor accident. That is a statement about reactors, and it needs two honest qualifications.

First, the Navy has lost nuclear-powered submarines. USS Thresher sank in April 1963 during deep-diving tests, with 129 people lost, and USS Scorpion was lost in May 1968 with 99 aboard. The Navy's court of inquiry tied the Thresher loss to the probable failure of a saltwater piping joint, which likely caused a reactor shutdown and loss of propulsion at depth, and Navy monitoring found no significant effect on the environment; the cause of the Scorpion loss was never established. The Thresher disaster led to the SUBSAFE program, which tightened standards for the materials and construction of everything that keeps a submarine's hull watertight. It is a good example of the culture at work: a catastrophe followed by a stricter system, not a cover story.

Second, a record of no reactor accident is a claim for a particular population of reactors run by a particular institution under military discipline. A military service can enforce rules that a civilian utility cannot, and it can select, train and discharge people in ways the private sector would find hard. The comparison to civilian plants is therefore a source of lessons rather than a proof that the same result would follow automatically. It is also a reminder that the Navy has its own incentives, and its record is documented largely by its own reporting. A skeptic is entitled to ask for independent verification.

Why seventy years is an argument and not an anecdote

A single quiet year proves little. A reactor can run safely for a year on luck. What makes the naval record notable is its length and its breadth. Seventy years is longer than most careers, longer than most corporations and longer than the political life of most programs. Over that stretch the Navy has put reactors in submarines, in aircraft carriers and in cruisers, under every kind of weather and every kind of strain, with crews rotated constantly and young enlisted sailors, many of them barely out of high school, standing the watch on the plant.

Think about what that implies. The people operating those reactors were not an elite of Ph.D. physicists. They were ordinary young Americans, selected and trained into excellence. If safety depended on exceptional genius it would not have lasted a single generation of sailors. It lasted because the system carried the standard from one crew to the next. That is the strongest evidence that culture can be built, written down and handed on, which is exactly what a conservative believes about the institutions that outlive the people in them.

There is a second implication, which runs against a common complaint. Critics of nuclear power say that a technology this demanding will inevitably be mishandled somewhere, sometime, by someone. The naval record is the standing counterexample. Demanding does not mean impossible. It means the work has to be done by organizations that treat the details as sacred, and a country that can produce such organizations for submarines can produce them for power plants, if it chooses to demand the same of them.

What civilian licensing could learn

The Nuclear Regulatory Commission was created in 1975, and its predecessor, the Atomic Energy Commission, had both promoted and regulated nuclear power, a conflict of interest that Congress ended by splitting the functions in the Energy Reorganization Act of 1974, which is why the NRC began work in January 1975. The civilian industry grew under a regime of prescriptive rules and long, expensive licensing proceedings, and critics on both sides of the debate have found fault with it: environmental groups for being too lenient, and builders for being slow and costly.

The accident at Three Mile Island in 1979 produced an institutional response that borrowed directly from Rickover's world. The industry created the Institute of Nuclear Power Operations in December 1979, in response to the recommendations of the commission that investigated the accident, and it is funded by the nuclear industry. Its first chief executive was a retired Navy vice admiral, Eugene Wilkinson. The idea was to bring peer pressure and a culture of excellence to civilian plants, because the regulators alone could not supply it. That is a case where a lesson from the Navy was applied outside it.

Three further lessons are worth drawing, and they apply to government more broadly.

• Speed and discipline are not opposites. The Navy took about nineteen months to take Nautilus from the keel laying in June 1952 to launch in January 1954, and the early naval program moved quickly because it had a clear mission, a single accountable office and strict standards. Civilian licensing today often has more process and less accountability. The goal should be rules that are exact and a process that is fast, as argued in We Raised the Empire State Building in About a Year, So Why Not a Power Line.
• Accountability beats paperwork. A thick binder of compliance is no substitute for a person who answers for the result. The same principle runs through the case against the administrative state in Madison Called It the Very Definition of Tyranny, and Agencies Hold All Three.
• Safety is a product of standards, not of fear. The numbers behind that claim are laid out in Count Deaths per Unit of Power and Nuclear Beats Nearly Everything Else: nuclear power, operated well, is among the safest ways to make electricity.

The strongest objection

A fair critic will say that Rickover was a difficult man who ran a closed system, and that his methods cannot be copied by a civilian industry in a free society. His sharp elbows and his habit of overriding objections are part of the record, and no one should hold him up as a model of how to run an office. His critics note too that a program with unlimited access to the best young talent in the Navy is not comparable to a utility with a budget.

The objection is correct about the man and wrong about the lesson. We are not recommending his temper. We are recommending his standards: accountability, technical competence, honest reporting and constant training. Those travel. They are the stuff of the institutions that conservatives admire, from the Marine Corps to a good small-town hospital. And the claim is modest. It says that where a free republic has built a high-reliability institution, it did so by insisting on those habits, and that civilian nuclear power, which needs all of them, could use more of them.

Why this belongs in a conservative case for nuclear

There is a larger point here about what American conservatism says it believes. It believes that institutions are shaped by character, that standards matter, that responsibility should rest on identifiable people, and that excellence is a moral achievement as much as a technical one. The naval reactors program is a living example. It was not built by a mandate or a subsidy. It was built by an insistence on getting the details right, in a country that was willing to let a stubborn engineer do it.

It also connects to the strategic argument in Admiral Rickover Built the Navy That Made Nuclear Deterrence Survivable, that the fleet he made possible has kept ballistic missile submarines at sea as a deterrent for generations. A power that can stay hidden under the ocean for months is a power an adversary cannot easily threaten, and that is a form of peace through strength built on engineering. The case for a strong fleet more generally is in Six Arguments Against a Strong Navy, and Why Each One Sinks.

The case in brief

1. Hyman Rickover built naval nuclear propulsion, from Nautilus (launched 1954, underway on nuclear power in 1955) to the fleet that has operated reactors at sea for about seventy years without a reactor accident.
2. The Shippingport Atomic Power Station (1957) showed the same design could produce civilian electricity, and the pressurized water reactor became the commercial standard.
3. The method was personal interviews, individual accountability, technical competence, honest reporting and endless training, which are habits of culture and not of luck.
4. The losses of Thresher (1963) and Scorpion (1968) were not reactor accidents, and Thresher led to the SUBSAFE program, which shows the culture correcting itself.
5. Civilian nuclear power borrowed some of this through the Institute of Nuclear Power Operations in 1979, and licensing could learn more about speed, discipline and accountability.

What you can do: read your own state's utility commission page on any nuclear plant it licenses or has a stake in, and ask what training and reporting standards the plant's operators are held to.