This article discusses the real-world practicality of nuclear power. Neither reactors nor casks of spent fuel have the capability of going “prompt critical” like a bomb. The laws of nature prohibit it and engineers must make clear that facts of nature are not matters of opinion. The Chernobyl reactor, which failed so spectacularly in Ukraine in 1986, became for many a symbol of the dangers represented by nuclear reactors. But that is not warranted; such an accident is simply impossible for the kind of commercial reactors now being planned or built. From a public safety standpoint, the most important feature of our current reactors is that, after any event that ruptures the reactor coolant system, a large amount of water and steam would be violently swirling around inside the containment structure, even if containment structure has been ruptured. In the emerging context of realistically reexamining many long-held assumptions, engineers will find opportunities to drastically improve the way nuclear power plants are built and operated.


In a conversation about nuclear power, a friend of mine once asserted that a reactor had the potential to "go offlike a bomb." I told my friend that was simply untrue, and she smiled sweetly and said, "Well, that's your opinion."

And no, it was never a matter of opinion. Neither reactors nor casks of spent fuel have the capability of going "prompt critical" like a bomb. The laws of nature prohibit it. And, engineers must make clear that facts of nature are not matters of opinion.

As Josh Billings, the nineteenth century humorist, once wrote: "The trouble with most folks isn't so much their ignorance, as knowing so many things that ain't so."

There are many scary stories about potential nuclear power disasters that restrain people from embracing the nuclear renaissance. In order to calm these unwarranted fears, engineers have designed reactors that are unnecessarily complex and, as a result, far more expensive than they need to be. I suggest that any competent engineers-not just nuclear-can show from their own knowledge that these fears are not valid. Let's look at a few of them.


The Chernobyl reactor soured many people on nuclear power. But the predicted wave of radiation-caused deaths has not materialized.

Some people speculate that "nuclear waste" may pose some sort of unprecedented problem. But in practice, nuclear waste does not figure much in the daily business of operating a nuclear power plant. The main bulk of the radioactivity is bound up in the ceramic nuclear fuel elements. Every year or two, fresh fuel is put into the plant, and the "used fuel" is put into a large "swimming pool" for several years, until it has lost 99.99 percent of its radioactivity by natural decay. Then it is usually put into a dry fuel cask, and stood up on a concrete pad in back of the power plant. The U.S. President has stated that it poses no hazard in this form, and can be stored thus for decades, when it will then be recycled to produce more fuel-a process demonstrated in several forms decades ago. The radioactivity is never released into the environment. Even if it were, the anti-nuclear activist, Sheldon Novick wrote of fission products in his book, The Electric War, "it is difficult to see in what way they are any more or less hazardous than other poisons produced by industry."

U.S. Naval reactors can now operate for the life of the ship-about 30 years-without refueling. All the waste from that operation stays locked up within the fuel elements. It doesn't even distort the fuel. That's not much waste.

The Chernobyl reactor, which failed so spectacuLarly in Ukraine in 1986, became for many a symbol of the dangers represented by nuclear reactors. But that's not warranted; such an accident is simply impossible for the kind of commercial reactors now being planned or built.

The Chernobyl reactor was very different. Its graphite burned for ten days, pouring fission products directly into the stratosphere. There is no graphite in current commercial nuclear reactors. Also Chernobyl had no real containment. Its control rods actually increased pow-. er under some conditions when shutdown was called for. When reactor temperature increased, its power could also increase (rather than the opposite, as required in our reactors). Its safety circuits had been deliberately disabled by operators "for a test." And supervision of operator training and decision-making was inadequate.

Even so, the thousands of deaths initially predicted failed to materialize. Leukemia is the usual indicator of a radiation problem, but the leukemia incidence in the nearby population has not risen. Fifty of the many plant workers who stayed and fought the fire were heavily irradiated and died. The total public radiation deaths were limited to ten thyroid cases in children, according to research by the United Nations Scientific Committee on the Effects of Atomic Radiation.

Most of today's safety requirements were created Long before there was much operating experience. The Government's brand new regulatory body was anxious to prove its effectiveness under strong anti-nuclear scrutiny. No one dared oppose any new safety provision. Moreover, each new requirement was imposed equally on all competitors, so there was no financial incentive for anyone to resist any safety requirements; they just added to the approved cost of the plant.

But safety is not an independent variable. Safety is created by the interaction among a variety of factors such as materials, design, selection and training of personnel, attitude of management, safety culture, and regulators.

Adding "safety features" to mitigate unreal problems does not make a plant safer; in fact, it provides new paths to trouble. The European Pressurized Reactor is sold as a conventional nuclear plant, but it introduces many "advanced safety features" including a "core-catcher" and additional leakage barriers, for which realistic design analyses show no need. The only core-catcher added to a large commercial power plant (Fermi-i) broke loose, blocked coolant flow, and created the very problem it was intended to ameliorate, as described in the book, The Day We Almost Lost Detroit.


While many skeptics worry that the expansion of nuclear power will lead to an inevitable catastrophe, later analyses show that catastrophic release was not possible with current commercial reactors.

In the early days, General Electric argued that containment vessels should be vented right after a malfunction, before any radioactivity was released. But that sensible suggestion was hooted down. Containment leak-tightness became an increasingly significant cost item, despite tests showing that the steam and water always present in a nuclear accident quickly scrub out most of the radioactivity. Think of the cost saving if we could eliminate most of the heavy rebar-enforcement and tedious hermetic leak-tightness testing.

From the beginning, the nuclear community recognized that answering the nuclear risk question was a top priority, and they devoted unprecedented effort to doing so. They continue to examine every scenario they can think of that can possibly lead to trouble-closing the wrong valve, shutting off a pump at the wrong time, ete. And they try to make that scenario highly improbableproviding backup circuits and flow paths, additional sources of off-site electricity and water, etc. They are still at it, and the probability of a serious casualty in a nuclear plant is now remarkably low.

But improbable things do happen, and we have to ask, How bad can it be? What are the consequences of the worst possible accident?



The nuclear community has been reluctant to discuss that question. Its members have been told that discussing potential casualties just scares people. But, in 1982, a Nuclear Regulatory Commission laboratory calculated a hypothetical accident where the fission products are released into the air and remain suspended indefinitely in respirable form over the nearest highly populated areas. They assumed that even the smallest dose of radiation will cause cancer, and calculated that hundreds of thousands of people would die. The report, "Technical Guidance for Siting Critical Development," tabulated the death toll for each of the 130 nuclear plants built or planned, and the commission sent these calculations to the major news media, which added further fantasies. And, sure enough, people were scared. They had no way of knowing that such a situation cannot be created in the real world.

"Worst case" calculations are not meaningful if they are based on conditions that defy the laws of nature and the known properties of materials and processes. A physical chemist might calculate, for instance, that an iron fry-. ing pan put over a gas stove burner could burn to ashes. But real-life cooks find that's not a serious consideration. Field tests and more sophisticated analyses show that zirconium alloy fuel element cladding used in water-cooled reactors are also not as vulnerable to ignition as previous analyses had implied.

Probabilistic risk analysis is an important tool. Used properly, it can help reduce the probability of a serious accident, and that's important. But nothing can replace the knowledge that, when all else fails, the consequences of the worst realistic accident will be tolerable.

Ironically, before the Three Mile Island accident, the Electric Power Research Institute had started to explore how using realistic assumptions would limit the calculated consequences of the worst accident. "Realistic," as used here, does not exclude any possible malfunction of personnel or equipment. Researchers including Chauncey Starr, Milton Levenson, Ian Wall, and Frank Rahn pulled together a lot of data and analyses from laboratory work and from large-scale tests and experiments to measure the release and dispersion of radioactivity.

This information, conservatively interpreted and thoroughly documented, showed that few, if any, members of the public would die from the most severe realistic accident to a typical U.s. commercial nuclear power plant. (See, for example, M. Levenson, F. Rahn, "Realistic Estimates of the Consequences of Nuclear Accidents" and the 48 references in the paper, published by Nuclear Technology, Vol. 53, May 1981).

These studies showed that, in the worst realistic case, fuel elements will not get as hot as previously expected, and less radioactivity is released from molten fuel, and more slowly. In addition, deaths cannot legitimately be predicted by accumulating tiny individual radiation doses calculated for millions of people downwind.

They also showed that highly charged and reactive fission product particles rapidly clump together and accumulate in the water flows and settle in the sumps, are absorbed in steam and water, condense and plate out on cooler surfaces, and interact with other materials they contact. The water lost from the reactor cooling systems does not disappear; it does not all evaporate inside the containment building. It reduces the airborne radioactivity within the building many thousand-fold.

After September 11, 2001, with another 20 years of data accumulated, I arranged for 19 nuclear-expert members of the National Academy of Engineering to publish a Policy Forum in the mainstream, peer-reviewed journal Science (Sept. 20, 2002, p.1997), updating the 1981 report and concluding that the worst that can be expected is a few deaths off site, if any. The American Nuclear Society White Paper on Realism, followed up with more details, confirming the conclusion. (See: rsh/realism/WP-TableOfContents.htm.)

A complete and detailed description of the whole casualty process, and the scientific evidence supporting it, needs to be collected, documented, and promulgated to the scientific community at large to receive the wide peer review and acceptance it deserves. But the magnitude of overstatement in current safety models is more than enough to cover the various uncertainties, and to support the conclusion that a catastrophic consequence of any realistic nuclear plant event is simply not possible in the real world.

In this context, note that I am not claiming that all types of nuclear plants pose no public risk. The Chernobyl plant was certainly not risk-free. Each type of system must be evaluated separately.

Despite this hard evidence, we still require mass evacuation drills, involving prisons, hospitals. and factories. Such drills, under the panic-making and unwarranted fear of radiation, clearly pose more chance of harm than benefit.

An example of how such misinformation can cause significant harm was the Shoreham nuclear plant, where newly elected New York Governor Mario Cuomo refused to approve the fully developed evacuation plans, thus preventing the plant from operating. After permitting just enough low-level operation to complicate disassembling the plant, its $6 billion cost was passed on to the utility customers over a 30-year period. This was in 1989, after Three Mile Island and several studies, including EPRI's, showed no serious public risk.

According to David J. Allard, director of the Bureau of Radiological Protection in the Pennsylvania Department of Environmental Protection, initial and follow-up studies, including one conducted 20 years after the accident, showed no health effects due to the reactor meltdown. There was a greater concentration of fission products detected in the atmosphere over central Pennsylvania in the mid-1970s from a nuclear weapons test in China than there ever was from the Three Mile Island reactor in 1979.

It is ironic that the Three Mile Island reactor accident should become a symbol of the fear of nuclear power. The accident showed that concerns over the inherent safety of nuclear reactors have been exaggerated far beyond the facts. The accident that created 10 to 20 tons of molten fuel that slumped several feet down onto the bottom of the reactor vessel-initiating the dreaded China Syndrome-showed that, in the real world, the reactor vessel froze the fuel and stopped the journey to China at five-eighths of an inch. Small samples were machin~d out of the inside of the reactor vessel itself by MPR Associates to confirm this important fact.

The greatest disruption attributable to the accident was the inconvenience and cost of a voluntary evacuation, which in the end proved unnecessary. The other nuclear plant on the TMI site continued to operate reliably.

From a public safety standpoint, the most important feature of our current reactors is that, after any event that ruptures the reactor coolant system, a large amount of water and steam would be violently swirling around inside the containment structure, even if containment structure has been ruptured. That process has been extensively studied, and it does a tremendous job of scrubbing nearly all of the most dangerous fission products from the atmosphere, so that they never get out of the containment building.

The task of scientists is to consider all potentially relevant aspects of a situation, and show us its opportunities and dangers. Their product is untested knowledge. The engineer's job is to take that knowledge and use it to make useful hardware. Both tasks are essential.

When German physicist Werner Heisenberg, one of the greatest thinkers of the mid-twentieth century, first started contemplating potential uses of nuclear fission, he calculated in his head the critical mass-that is, the minimum amount of fissionable material needed to make a nuclear bomb. He wanted to be conservative, and the number he calculated was many tons of the uranium-235 isotope. Since this was impractically high, the Nazis never gave any serious thought to developing an atomic bomb. His calculation was not conservative; it was simply wrong.

Another example involves a "conservative" calculation in 1960 by the brilliant physicist, Edward Teller. He calculated that a submarine refueling accident could release a lethal cloud of airborne radioactivity extending several miles. He had described the idea in an article for Parade magazine and recommended that refueling be done at sea.

The article had come to Admiral Hyman Rickover for clearance, and he asked me if Teller was correct. I made a realistic calculation showing the maximum radiation dose would be an acceptable one-time emergency dose of25 rad at 100 meters.

Rickover, the architect of the nuclear Navy, invited Teller and the entire Atomic Energy Commission Reactor Safeguards Committee to meet aboard the latest nuclear submarine. I had discussed my calculations with the committee the night before, and they agreed with me. The next day, behind closed doors, Teller convinced the committee that he was right. So Rickover threw me into the lion's den.

The fmt time through, Teller was not convinced, because I used engineering units. So I struggled and got the same answer with unfamiliar physics units, but he still wasn't convinced. So he walked me through his approach, starting with the assumption that airborne radioactive particles are leaking from the reactor core. This hot cloud rises from a point near the water surface, diffusing outward in all directions. If this were in open air, it would form a slowly expanding sphere, irradiating shipyard personnel below.

"But it's not a whole sphere, Edward," I protested. "Half of your sphere is under water. The airborne particles are all above water. Your radiation dose is high by a factor of two."

"We're just trying to get a conservative answer," he grumbled. But I pressed on. There were other factors of similar magnitude, and the final answer came down to about what I had calculated.

"Do you agree with that?" he asked the committee. They did, and he withdrew the article.

This doesn't prove I'm smarter than Teller or that engineers are smarter than physicists. I don't know any engineers who can do what Teller did. The point is that both scientists and engineers have essential, complementary roles to play.

The lesson I draw from this example is that persons performing safety reviews must be very careful to distinguish real problems from imaginary ones.

So, what are the most urgent tasks today in nuclear power? First, is the question of how smoothly the regulatory process will play out. The surest way to answer that question is to pick a type of plant with the fewest unanswered regulatory questions, and push one or more through the regulatory process as quickly as possible. More advanced plants can be worked on at the same time, and filed for regulatory review later.

That process will quickly reveal the second type of problem: recent initial bid prices for new plants are shockingly high. Normal negotiation should have some impact, perhaps bringing loan guarantees and some form of no-carbon credit. But skillful engineering could bring more fundamental improvement.


Fear of a nuclear accident like the one at Three Mile Island (below) kept the Shoreham reactor (above) from commercial operation.

The nuclear industry has demonstrated decades of nearly flawless performance and safety worldwide. Nuclear plants do not need more safety features. They need to be simpler and less expensive to build and to operate, so that we can maintain that excellent record. We need to build hundreds of them, as quickly as possible.

Is there reason to believe that nuclear power can be made cost-competitive with gas and coal? Yes, indeed. . Many plants that have been running long enough to write off much of the original construction cost are now operating profitably year after year. In fact, the governments of Finland, Sweden, and Belgium, and even the Attorney General of Connecticut have complained that nuclear plants are so profitable, and so free of the uncertainties facing non-nuclear competitors, that they are making "unearned profits" and should be fined a windfall profits tax of some sort.

In the emerging context of realistically reexamining many long-held assumptions, engineers will find opportunities to drastically improve the way nuclear power plants are built and operated.

As we look at the challenges posed by the unknowns and uncertainties of burning biofuels instead of returning them to the soil, or storing billions of tons of carbon dioxide and other harmful products offossil fuels, or dealing with the unpredictabilities of solar or wind power, we must ask: Suppose we learn to solve all their problems; why would it not be simpler and more sensible to just build a few more nuclear power plants, that we know from decades of experience meet all our requirements?

Quick Reverse

One myth that sometimes is used to present a case against nuclear power plants is that they are baseload only, and not maneuverable enough to be used for load following. That's absurd. The flexibility and responsiveness of a nuclear power plant are dramatically demonstrated in a working nuclear-powered attack submarine. Admiral Hyman Rickover used to.challenge crews to see how fast they could change power level.

Admiral Hyman Rickover (below) is known as the "Father of the Nuclear Navy" for his role in ushering in nuclear-powered submarines, such as the USS Nautilus, shown arriving in New York in 1958.

Admiral Hyman Rickover (below) is known as the "Father of the Nuclear Navy" for his role in ushering in nuclear-powered submarines, such as the USS Nautilus, shown arriving in New York in 1958.

Close modal

They selected the biggest, strongest sailor aboard, and passed the order, "Crash back!", a maneuver that takes' the ship from "All Ahead Flank!" to "All Back Emergency!" The sailor spun the ahead throttle closed, bringing the reactor from 100 percent power to zero. He then. immediately opened the astern throttle as fast as possible, restoring the reactor to 100 percent power.

The boat shook violently, shrieks of various stressed equipment mixed with the clatter of coffee cups, operating manuals, and other miscellany sliding off normally horizontal surfaces, as the vessel shuddered to an emergency stop. The reactor plant operator watched placidly as the reactor temperature dropped a few degrees, which automatically raised the power smoothly to the required level. He could have pulled the reactor control rods a notch to restore the temperature, but he didn't need to. The electrical plant operator and the steam plant operator were equally relaxed, as the ship's propeller reversed and the plant accommodated itself to the new conditions.

Ifs hard to imagine the need for any greater flexibility than that. Unlike a combustion-fired power plant, which has to warm up and cool down very carefully to avoid thermally shocking the system, the temperature swings in a water-cooled plant are quite moderate, and impose no restraints on the operators.

The greatest disruption attributed to the Three Mile Island accident was the inconvenience and cost of voluntary evacuation.