The accident last April at the Chernobyl nuclear power plant demonstrates that planning conducted at a national level alone cannot eliminate the risks posed to all nations by nuclear energy. In the aftermath of the Chernobyl accident, an attitude of "business as usual" will not sustain the atomic power industry worldwide.

The scope of the challenge to make nuclear energy production safer is even greater than that shown by the well-known accidents at Chernobyl and the Three Mile Island plant in the United States. Between 1971 and August 1984, two "significant" and 149 "potentially significant" mishaps occurred in 14 industrial nations outside the two superpowers. Even aside from the danger of accidents, the normal operation of nuclear power plants presents problems. These include the management of materials—plutonium and weapons-grade enriched uranium—which could be diverted for non-peaceful use by nations and terrorists, and the possibility of sabotage and military attacks on power plants. The potential damage from such actions includes radiological consequences far worse than those witnessed in the aftermath of the Chernobyl accident.

International institutions have not been oblivious to the challenges posed by the 382 commercial reactors presently operating in 26 countries. Emerging today are worldwide nuclear standards—albeit imperfectly applied—that attempt to treat each risk in a distinct manner. The Soviet accident and the international response it generated make this an opportune time to consider a more comprehensive approach.


At the time of the accident the Chernobyl energy complex consisted of four nuclear reactors, each producing 1,000 megawatts of electricity, and two additional reactors that were under construction. Situated on the Pripyat River, which feeds a reservoir of the Dneiper River above Kiev, Chernobyl is surrounded by an agricultural region that produces rye and dairy products. To the south is the Ukraine’s rich wheat belt. With the exception of Pripyat, a community of 49,000 built in the 1970s to accommodate construction workers and plant employees, and the town of Chernobyl, with 12,000 inhabitants, only villages with populations of 500 to 700 residents dot the nearby countryside. Chernigov, a city of some 210,000, lies 55 miles to the north; Kiev, the Soviet Union’s third-largest city with 2.3 million residents, lies 60 miles to the south.

Unit 4, the reactor involved in the accident, started operation in 1983. Like the other Chernobyl reactors, and ten similar plants—in some cases not as modern—near Leningrad, Kursk and Smolensk, Unit 4 is a RBMK (an acronym derived from the Russian initials for heterogeneous, water-graphite channel-type reactor), which uses graphite as a "moderator," i.e., to slow down neutrons and thereby sustain the nuclear chain reaction. The reactor’s design evolved from the smaller Obinsk plant built in 1954, one of the earliest Soviet models. The Soviets made the leap to the 1,000-megawatt (electric) RBMK reactor, the Chernobyl type, in 1974 when the first of these went into operation about 60 miles from Leningrad. A much larger 1,500-megawatt plant at Ignalino, Lithuania, went to full power in 1985; a twin is currently under construction on the same site, as are two other similar plants at Kostroma, 125 miles northeast of Moscow.

Before the Chernobyl accident, the European portion of the Soviet Union relied on the atom for one fifth of its electricity. Twenty-three graphite-moderated reactors throughout the country supplied 24,000 megawatts of electricity, or roughly half of nuclear energy’s 11-percent contribution to Soviet power generation. The remaining nuclear contribution came from pressurized, light water reactors (PWRs), similar in design to most plants in the West. PWRs use water, not graphite, to moderate and cool the reactor core. The Soviets intend to rely mainly on this PWR design to carry them into the 21st century, by which time they hope nuclear energy will satisfy over 30 percent of their electrical needs.

Graphite-moderated RBMK 1,000s, in comparison to PWRs, are physically larger, more complex in design and inherently more difficult to control. Some 2,500 hollow vertical channels run through the 12- by 7-meter, 1,700-ton graphite moderator. Within 1,659 of these channels are "pressure tubes" that house the reactor’s 30,000 fuel rods in bundles of 18 each. Cooling water circulates through the tubes; some of it boils, and the resulting steam drives two 500-megawatt (electric) turbines. Control rods inserted into 211 graphite channels manage the reactor’s power level. An automatic system adjusts the positions of the individual control rods, and a separate control system is able to shut down the reactor entirely. These mechanisms cannot, however, be described as a quick shutdown system to bring the reactor under control.

Graphite-moderated, water-cooled reactors have a peculiar feature—a "positive void coefficient"—that contributed to the Soviet accident by resulting in a sudden surge of power. In the water-moderated environment of a PWR, we find a negative void coefficient, which means that the formation of steam slows the generation of energy. In the RBMK, as in other graphite-moderated reactors, the formation of steam tends to enhance the chain reaction.

But the RBMK design also has advantages over PWRs. RBMKs do not require the costly and difficult-to-fabricate reactor pressure vessel or the complex and expensive steam generator system of PWRs. Since fuel channels are individually cooled in the RBMK, a break in one will not result in a loss of coolant for the entire core. Furthermore the channels can be refueled while the reactor is in operation, thus allowing the plant to be kept on line. PWRs, by contrast, must be shut down for refueling.

Soviet authorities have been sensitive to the danger of nuclear accidents. The Soviets suffered a major nuclear disaster in 1957 in the Urals which may have cost hundreds of lives. Its nature remains a mystery to the West; it is widely believed not to have involved a reactor but rather a waste storage dump. But Russian reactors, like Western reactors, are known to have been subject to emergency shutdowns. We are not sure if these have involved the release of radioactivity, although Sweden, which was dusted with radiation coming from the direction of the Soviet Union in 1981 and 1983, is suspicious. As in Western reactors, ruptures of the large coolant-carrying pipes are assumed to pose the greatest hazard to Soviet plants, and therefore concrete-walled compartments—considered by Western experts to be adequate to the danger—have been built to shelter the conduits. In the event of a large pipe break, steam is channeled to two condenser pools of water in the basement of the plant. To make up for any loss of coolant, the Soviets have installed in the RBMKs an emergency core feedwater system. A leakproof cavity around the core, topped by a steel cover and situated in a heavy concrete-walled building with a thin roof, provides the final defense against the release of radioactivity to the environment. Some new Soviet PWRs have the more conventional heavy steel-reinforced concrete domes found in many Western containment buildings.

A report in a Ukrainian newspaper in March 1986 stated that the Chernobyl facilities were built with "defective materials," but Unit 4 had performed almost flawlessly until the accident, perhaps lulling the operators into believing the plant was impervious to disaster.

The Chernobyl accident resulted from gross operator incompetence—not entirely unlike that which resulted in the accident in 1979 at Three Mile Island—compounded by the managerial difficulties of the RBMK. Ironically, the origins of the Chernobyl accident lay in a decision to test safety systems. It is unclear whether officials in Moscow or those at the site authorized the test. The rationale was that all nuclear power plants, including Unit 4, rely on offsite power to drive pumps and other devices necessary to operate the installation when onsite generation ceases. A failure of such power could cut off the flow of coolant; a meltdown and release of radioactivity could then result even after a shutdown, due to residual heat. To compensate for this possibility, plants are built with emergency diesel generators that activate in the event of a power failure, but these diesels require up to 40 seconds to come on line. The operators wanted to determine whether the plant’s turbine generators, through their residual kinetic energy, could provide electricity for vital systems during the period between the loss of offsite power, or an internal power failure, and activation of the diesel generators.

In tests conducted at Chernobyl in 1982 and 1984, the reactor’s mechanical energy was exhausted before the diesel generators came on line, and voltage to maintain feed pumps declined too soon. The 1986 Chernobyl experiment was to test a new device to keep voltage up until the diesels were able to power the pumps. The operators, joined by the firm that had installed the plant’s electrical system, set in motion a chain of events that amounted to an accident simulation—giving little thought to the dangers. They chose April 25, when the reactor was scheduled to be brought down for maintenance.

Preparations for the test began at 1:00 p.m. with a 50-percent reduction of power generation and the shutdown of one of two turbines. At 2:00 p.m., operators took the first of several steps to block critical response systems, deactivating emergency core-cooling pumps to prevent their interference with the experiment. Shortly thereafter, the test was delayed for nine hours as the plant was placed back on the power grid to meet a new and unexpected demand for electrical generation. At 11:10 p.m. the reactor was taken off the grid and preparations for the test resumed.

From this point on, until the beginning of the experiment over two hours later, the reactor became increasingly difficult to control, due to the operators’ failure to observe standard low-power operating procedures. Problems began with a rapid decline in power—from a planned level of 700-1,000 megawatts (thermal) down to only 30 megawatts—well below the permissible safe operating level. To increase power, operators then withdrew the control rods, holding in reserve only six to eight rods instead of the 30 required by plant safety margins. According to plan, extra coolant was pumped into the core. But in the plant’s unstable low-power state, the water saturation of the pressure tubes increased the danger that small changes in temperature could cause extensive steam formation which might result in power surges. Because operators negligently continued to shut down more automatic safety systems to avoid their interference with the experiment, such surges became difficult to control. Although computer printouts warned operators just before the initiation of the test that the reactor was losing its emergency shutdown capability, the late hour, coupled with operator fatigue and impatience, led administrators to press ahead, knowing that if they failed to complete the test they would have to await the plant’s next annual maintenance.

At 1:23 a.m. on April 26 the operators cut off the steam supply to the turbine generator to simulate a failure of the plant’s internal support power. Coolant water from the pumps began to decline. With decreasing water flow and turbine generator inertia, boiling increased, due to the positive void coefficient. Power levels accelerated and the reactor became increasingly unstable and difficult to control. Thirty-six seconds into the experiment, operators attempted to drop the control rods, but not enough rods fell into place. By the time the minimum "safe" number of rods were finally in place, steam generation had diminished their ability to contain the chain reaction. In less than three seconds, power surged to 530 megawatts (thermal), accelerating even faster thereafter to the point where, American analysts believe, a "prompt neutron power burst" resulted in a rapid heating of the fuel, which in turn caused a massive steam explosion. A second explosion followed two to three seconds later, either because of the mixing of hydrogen or carbon monoxide with air or because of a second power surge. The detonations ripped through the thousand-ton, steel-plate encasement and the roof of the reactor, which were designed to withstand only two atmospheres of pressure, not the ten atmospheres to which they were now subjected.

One quarter of the graphite in the core was ejected in the explosions; 30 fires ignited in and around the plant, some flames leaping 30 meters into the air. One fire, on the roof of the turbine building shared by Units 3 and 4, raised the specter of another reactor accident. Fortunately, about 250 firemen who arrived soon after the accident extinguished all of the blazes by 6:35 a.m. except for the graphite fire burning within the Unit 4 core itself.

The battle to contain the release of radioactivity from this inferno—a fire that ultimately consumed about ten percent of the graphite—had only begun. It took almost two weeks to bring the situation under control. In the end, the plant released between three and four percent of the total core fuel inventory of radioactive materials: 15-20 percent of the radioactive iodine, 10-20 percent of the radioactive cesium and less than 5.6 percent of other radioisotopes, including three percent of the plutonium. One fourth of the release of radioactive material occurred within the first day; the remaining three-fourths was released in the following two weeks. The final solution was found in starving the graphite fire of oxygen. First, to stop the chain reaction, the Soviets dropped 40 metric tons of boron carbide on the core from helicopters. They bombarded the reactor with 800 metric tons of dolomite, followed by 2,400 metric tons of lead and 1,800 metric tons of clay and sand to afford shielding. To cool the reactor under the debris and prevent a meltdown through the reactor floor which could have contaminated the water table 50 feet below, Soviet engineers pumped in liquid nitrogen.

Notwithstanding the silence of Soviet officials at the time, policymakers in Kiev and Moscow were informed within hours. Major General Gennadi Berdov, the Ukrainian deputy minister of internal affairs, initially took command of efforts to control the situation 90 minutes into the accident, sealing off the immediate area around the complex with local militia. A team of doctors and technicians from Moscow reached the Chernobyl environs at 8:00 a.m. on April 26. But it was not until a senior delegation of government and party members arrived at the site that evening that policymakers at the highest level understood the seriousness of the disaster.

In the meantime the Soviets mobilized considerable resources to mitigate the consequences. Implementation of civil defense procedures in Pripyat was delayed at first because of an inadequate emergency plan and the presence of significant contamination on forested evacuation routes. Once communications with the inhabitants of affected areas were established, citizens were told to stay indoors. Officials felt that the shielding afforded by buildings provided the best defense, in light of their uncertainty about the location of early radioactive emissions. Volunteers went door to door distributing potassium iodide tablets to block thyroid intake of radioactive iodine (I-131), the most important immediate hazard. Soviet aircraft sprayed silver iodide through rain clouds moving toward the region to cause rain to fall before the clouds reached Chernobyl and thus prevent any precipitation from concentrating the fallout. Thirty-six hours into the accident, as radiation levels which were initially light in Pripyat climbed, 1,100 buses evacuated the population living around the plant. Ultimately, 135,000 people in an 18.6-mile radius around the reactor were relocated, some, including the citizens of the town of Chernobyl, nine miles away from the plant, as late as one week after the accident began. As the government further understood the gravity of the situation, it sent several hundred thousand children from Kiev, Byelorussia and the northern Ukraine to recreation camps. When summer ended, these children returned to their communities. Those from the original exclusionary zone did not return to their homes, however; they joined their parents in new and temporary housing near Kiev.

Radiation spread across Europe from the start. Days after the accident, prevailing upper-level winds swept radiation over the Arabian Peninsula, Siberia and eventually North America. Although the accident induced a good deal of concern and psychological trauma, the World Health Organization at the time determined that the contamination posed no immediate acute health hazard to any areas except the region directly around the reactor and some "hot spots" in the Soviet Union and Europe created by rain-bearing radioactive clouds. (After the accident, the Soviets estimated that half the released material fell within an "exclusionary zone" 18 miles in radius around the plant, posing immediate contamination problems in a zone extending at points 18 miles farther.) Nonetheless, the European Economic Community, recognizing that consumption of even lightly contaminated produce could increase the probabilities of cancers and thyroid maladies, banned all fresh food imported from nations within 1,000 kilometers of the accident, effectively barring agricultural exports from Eastern Europe (with the exception of East Germany, excluded at the request of the Bonn government). At the same time, public health officials, using different radiation standards, took a variety of steps to prevent consumption of homegrown leafy vegetables and milk. In countries bordering the Soviet Union, notably Poland, governments administered potassium iodide to children.


SOURCE: Simulation by the PATRIC computer model. Maps courtesy of Marvin H. Dickerson and Thomas J. Sullivan, Lawrence Livermore National Laboratory.

Firemen, other rescuers and plant personnel (176 operation staff and 268 construction workers from Units 5 and 6) suffered the most serious consequences of the accident. To date, the Soviets report that of the 203 persons hospitalized with thermal burns and acute radiation poisoning, 29 have died (the initial explosion took the lives of two others). An unknown number, perhaps thousands, of others were given medical examinations and released. At first glance, the fatalities seem rather low, certainly not the thousands that some speculated a major accident would produce. Indeed, the fatalities are modest compared to other industrial accidents such as the poisonous gas leak at a chemical plant in Bhopal, India, which took 2,000 lives in its immediate aftermath.


In some respects, namely the number of early fatalities (as distinguished from cancers that may not manifest themselves for years), the world escaped the worst that could have occurred at Chernobyl. There was an elevated temperature inversion over the plant the night of the explosion, which allowed the radioactive debris to rise quite high, thereby diminishing the density of its concentration in the atmosphere. Prevailing winds initially carried the radioactivity away from Pripyat. A less energetic plume, released at the base of the reactor during the daytime, when people would have been out on the streets, coupled with a steady drizzle which would have washed radioactivity down to earth, could have resulted in many early fatalities. Thus, generalizations drawn from the number of early deaths in the aftermath of Chernobyl must be treated cautiously.

Another somber but less definable aspect of Chernobyl is the emotional distress reported by the Soviets as a particular problem among children. Most troublesome will be further illnesses and deaths. The Chernobyl reactor emitted two types of radiation that are of prime concern: iodine 131 and cesium 137. (Strontium 90 may also pose a problem, but scientific calculations have not focused on it.) Radioactive iodine concentrates in the thyroid and may be absorbed through breathing or food consumption. Fortunately, I-131 has a short half-life of eight and one-half days. Although it induces nodules in the thyroid rather easily, most of the nodules could be surgically removed. A small percentage, though, will be cancerous and fatal if not caught in time. Because I-131 quickly accumulates in the thyroid, and assuming that large populations were not administered potassium iodide soon enough or at all, we can anticipate that nodules may be the earliest, and quite possibly the most common, physical manifestation of the accident. Tens to hundreds of thousands of people may be affected.

Had Chernobyl contaminated large populations with hundreds of rems (a unit for measuring the biological impact of radiation), projections of the ultimate extent of the accident’s human costs might be relatively easy to define, given the present state of medical knowledge. But beyond the exclusionary zone within which some residents absorbed tens of rems, 75 million Soviet citizens and tens of millions of other Europeans residing in contaminated regions and eating radioactive food produced in contaminated soil will be exposed to well below five rems over a lifetime as a result of the Chernobyl explosion. No expert consensus about the physiological impact of exposure at this low level has been reached. Scientists express an extraordinary range of calculations about the long-term effects of Chernobyl, the low being 210 deaths and the high, one million deaths. To the extent that there is a consensus—and it is a quite shaky one—deaths are predicted in the range of 20,000 to 40,000 within the Soviet Union, with substantial numbers outside. Whatever the figure, nonfatal cancers will be of a similar magnitude and genetic effects may appear in future generations.

What accounts for these diverse estimates? What is the significance of the precise magnitude of the final casualty figure? The latter question may be easier to answer, political considerations being the "bottom line" in assessing a disaster with as many possible ramifications as Chernobyl. If the low figures prove correct, nuclear energy proponents will argue—indeed some already have argued —that atomic power poses problems not entirely distinguishable from those presented by other energy sources and toxic waste-producing industries with which we have learned to live. After all, following the Bhopal accident we have not abandoned the pesticide industry, nor have we tried to shut down coal-powered electricity-generating stations that emit sulphur dioxide and thus contribute to widespread respiratory problems for millions of people worldwide and threaten the stability of the earth’s climate by creating a "greenhouse effect."

On the other hand, if the moderate or higher casualty figures are correct, opponents will make a stronger case that nuclear energy creates an unacceptable risk. Sweden adopted this position even before Chernobyl; the incident only served to harden Swedish nuclear opponents in their opinions. Sweden intends to eliminate its use of atomic power by 2010. Several years ago, Austria decided not to license a newly constructed reactor, in effect discontinuing the country’s nuclear power program. Given the global nuclear energy policy debate, it is important to strive for better estimates of the human costs of Chernobyl.

Doing so will be difficult. Scientists are deeply divided on the biological impact of low-level radiation; they base their beliefs on vastly different assumptions, due to an absence of definitive data for the effects of low doses of radioactivity. We know that exposure to radiation from natural sources, such as uranium and thorium in the earth and cosmic rays in the atmosphere, is inevitable. Certain radioactive substances occur in the body. This "background" radiation, which is not necessarily harmless, varies at different points around the globe, but it is usually on the order of 100 millirems per year. This figure is used as the baseline for setting standards of man-made radiation. There is considerable controversy among scientists over the level at which man-induced radiation results in significant biological effects. Some argue that there is no threshold below which man-made radiation exposure is harmless, that the consequences are not reparable by the body’s defenses, that radiation accumulated over time has the same effects as a single, acute dose and that the effects of low doses are proportionate to higher ones. Another school holds that the biological risk is less below a certain threshold, because cells are able to repair sublethal damage.

To form hypotheses, scientists rely on studies of Japanese atom bomb survivors, Marshall Islanders irradiated by weapons-test fallout in 1954, uranium miners exposed to radon, doctors and patients using radiotherapy, and the results of animal experiments. Despite large study samples, researchers face difficulties in accurately measuring radioactive dosages and creating exactly equivalent control populations; they cannot always isolate radiation effects from those of other chemical or physical agents. They cannot easily establish a cause-and-effect relationship when there is a long period (30 years or more) between the exposure and manifestation of the effect, particularly in light of the fact that cancer "naturally" kills 20 percent of the West’s population. Scientists are groping for conclusions because the myriad of variables produces remarkably different results based on different interpretations of the risk level. The Soviets suggested that casualties from Chernobyl could at worst amount to two cancer fatalities per 10,000 people exposed to at least one rem of radiation, while International Atomic Energy Agency (IAEA) officials estimated fatalities at half that rate. Complicating matters will be the difficult assessment of measures taken to minimize exposure to radiation. Controlling the consumption of irradiated produce and scraping away affected soils or treating them with potassium fertilizers and calcium compounds to minimize cesium and strontium contamination may help ameliorate later health problems, but to varying and uncertain degrees.

The Soviets have promised to study intensively the 135,000 evacuees from the exclusionary zone around the plant and to share the information with specialists in the United States and elsewhere. This information, if tangible—a big "if considering that precise dosimetry may not be available—may help resolve the disputes that have risen over the human costs of Chernobyl.

Along with the human costs are still-uncertain economic ones. The Soviets place the direct cost near $3 billion. This includes lost electrical generation which, although mitigated in October 1986 by the resumption of generation at Chernobyl Units 1 and 2, will result in a five-percent reduction in Soviet energy generation in 1986-87, as similar RBMK plants are shut down for remedial improvements to be undertaken as a result of the Chernobyl accident. Also involved in assessing costs is the expense of relocating 135,000 people from the exclusionary zone and reimbursing them for their personal belongings.

The cleanup work was massive. To contain further releases, a concrete and metal "sarcophagus" now entombs the reactor. It will be in place for centuries and will require monitoring and continual reinforcement. To conduct heat away from the still-simmering core during the weeks after the accident, engineers pumped nitrogen into a basement vault now underlaid with a concrete slab to prevent ground water contamination. However, the danger of a lateral spread of radiation into the Dneiper River remains. To address this problem, the Soviets entered into a $23-million contract with two Italian firms for equipment to build a "diaphragm," four kilometers in radius and 100 meters in depth. To impede the dissemination of surface contamination by local winds, cleanup personnel spread polymerizing solutions over 150,000 square yards of land and removed tons of the most contaminated soil to a depth of five to ten centimeters for deposition in dumps.

It is unclear whether the Soviets include in the cost of the accident the expense of updating other RBMKs with modern computers; the cost of lengthening and increasing the number of control rods; the price of improving reactor monitoring and automatic shutdown mechanisms; the expense of boosting the capacity of coolant pumps; and the cost of replacing the core with more highly enriched fuel to provide greater stability. To these figures must be added tens of millions of dollars for oil to generate substitute electricity, oil that otherwise would have been exported. There are also projections of power brownouts in months to come, which will result in a loss of industrial capacity. In addition there are the modest costs of medical monitoring that will have to be conducted for decades ahead, and the price to Soviet agriculture of land rendered unproductive, of discarded produce, and of scrutiny for contamination that will vary in duration from months to years depending on the growth cycle of plants and patterns of radioactive redistribution by wind and rain.

The human and economic price of Chernobyl will take years to compute, during which time policymakers around the world will be forced to make difficult decisions about the future of nuclear energy in their respective countries. The benefits of atomic power will have to be weighed against the availability of other energy options, the risks of nuclear energy and the experience of other countries with reactors.

Even before Chernobyl, the nuclear industry worldwide was in a serious slump. Forecasts made in the 1970s of installed nuclear capacity in OECD countries through the 1990s were, by the early 1980s, reduced by 75 percent. Hopes for a burgeoning nuclear power market in developing countries failed to materialize due to the unavailability of capital and hard currency, rising construction costs, falling oil prices, slackening demand in industrial countries for electricity, and grass-roots anti-nuclear sentiment. Still, by 1990, 118 new plants in 17 countries will probably go into operation. At the present time, it does not appear that Chernobyl by itself has affected these plans markedly. What the accident may do is discourage countries from laying future plans for construction. Other considerations will include demand for electricity, the state of capital markets and the ability of reactor vendors to reassure consumers that atomic power facilities can be built safely and with features that avoid other strategic nuclear energy risks as well.


Clearly, Chernobyl is the most serious civil nuclear energy accident to date. It demonstrates that major accidents are possible because of inadequate reactor design and incompetent employees. It also illustrates that such incidents require heroic measures to combat them.

Notwithstanding more rigorous safety standards, reactors manufactured in the West have also experienced important mishaps. In 1984, for example, at one French reactor at Le Bugey, the failure of an electrical component and an emergency diesel generator, coupled with operator error, almost led to a major accident. In the United States, there was Three Mile Island; no early deaths resulted and, given the low amounts of radioactivity released, late-developing cancers—if there are any—will probably be limited to fewer than four cases, but the event highlighted the fragility of nuclear facilities. There were also several near-misses: the partial core melt in 1966 at the Enrico Fermi demonstration breeder reactor 30 miles from Detroit; the electrical cable fire at Browns Ferry, Alabama (which started because a technician used a candle to check for air leaks, and which resulted in several hours’ loss of control over the reactor); and the mishap at Toledo Edison’s Davis-Besse plant in 1985, caused by equipment failure and human error, which started a sequence similar to that at Three Mile Island. Whether any of these and other incidents could have progressed to the point of becoming a Chernobyl-like disaster in the absence of remedial measures, given dissimilar reactor designs, negative void coefficients and different containments, is debatable, but there is growing concern.

Serious questions have been raised about the safety of one American graphite-moderated plant that has no containment: the "N" reactor at the Hanford nuclear reservation in Washington State, a dual-purpose facility producing plutonium for the Defense Department’s weapons program and electricity for the northwestern U.S. power grid. Four other small plutonium production reactors at Savannah River, South Carolina, also have no containment features at all.

Besides reporting the 151 nuclear safety incidents noted above, the General Accounting Office has suggested that the outlook is questionable for assuring the safety of 73 plants either already in operation or on order in 17 developing nations, because most of these countries "lacked trained personnel to draft nuclear safety standards to train nuclear safety personnel." Assuring adequate standards will become more complex as South Korea, China, Argentina, India and other countries increasingly manufacture their own nuclear hardware and components rather than rely on imports. Developing countries as well as many developed states commonly recognize that they are unprepared to address a major nuclear accident alone.

But accidents are not the only problem for the nuclear industry:

Nuclear Sabotage. Between 1970 and 1984, 292 acts or threats of sabotage or diversion of nuclear materials took place worldwide. These incidents—including some bombings and arson—involved nuclear power installations, transport and personnel, and were perpetrated by terrorists, criminals and hostile employees. Recently, the number of such incidents has declined. Still, as a recent RAND Corporation report notes, "this should not lead policymakers into a false sense of security." Underscoring this admonition was the saboteur disabling of three of four offsite power lines feeding the Palo Verde, Arizona, nuclear energy complex in May 1986.

Neither this incident nor the hundreds of others around the globe have resulted in important radiological releases. Still there are ominous trends such as state sponsorship of terrorists which gives them access to sophisticated tools and training, and there is an apparent willingness by terrorists and their sponsors to take more lives than in earlier years. It is debatable whether a group would be willing to murder truly catastrophic numbers but sabotage of highly visible nuclear plants remains a possibility.

Too many countries are simply unprepared to address this threat. For example, U.S. regulations dating from the mid-1970s prescribe that reactor operations need not protect against more than one internal or more than three external attackers, against "enemies of the United States," or against attackers operating with anything more threatening than handheld automatic weapons or explosives in quantities small enough to be carried by hand (thus excluding protection against truck-bomb assaults). Power reactors, therefore, are required to have a minimum of only five guards. Though most facilities employ more, recent congressional investigation found that security around military production reactors and other sensitive defense establishments is a "shambles."

Military Attacks on Reactors. Since Iran (or, some speculate, Israel) initiated the first military attack on a nuclear reactor in 1980, when jet aircraft bombed Baghdad’s nuclear research facility at the outset of the Iran-Iraq war, reactors have been struck five times: Israel’s destruction in 1981 of the same Iraqi facility, three Iraqi attacks against Iran’s Bushehr installations along the Persian Gulf and one Iraqi strike against a small Iranian nuclear research center. Fortunately, none of these plants were operating at the time and nuclear releases were avoided. We cannot assume that releases will be avoided in the future.

Beyond Israel’s rationale, namely to preempt an adversary’s acquisition of atomic weapons, military attacks on reactors may occur for a variety of reasons: to cripple an antagonist’s industrial warfighting capability by depleting energy generation, to destroy the environment for military purposes through the release of radioactivity; or to make the war more costly for the adversary by destroying a valued capital asset (perhaps the rationale for attacks in the Iran-Iraq war). Then, too, an aggressor may bomb a nuclear power station accidentally, or the parties to a conflict might consider sabotaging a facility to escalate the war. Since the citizens of many countries have become acutely concerned about radioactive releases from nuclear power accidents, a belligerent could capitalize on this fear for purposes of intimidation. Unfortunately such threats are consistent with international law and are not explicitly opposed by the United States. Washington takes the position that an international convention prohibiting military attacks on nuclear reactors would not be meaningful in light of the defense requirements of the United States and its allies.

Conventional weapons presently in the arsenals of many countries are quite capable of breaching the hardest reactor containments or destroying critical equipment outside nuclear power plants. A wartime attack could produce radiological releases much greater than those from Chernobyl. War raises two distinct and additional dangers: first, remedial assistance to diminish escaping radioactivity might be less readily available than it was in the Ukraine, as civil defense needs would complicate the situation. Second, the contents of not just one but all reactors at the same site could be released.

Diversion of Nuclear Material for Weapons. Six countries have detonated nuclear explosives: the United States, the Soviet Union, Britain, France, China and India. Many analysts suspect that Israel and South Africa also have the bomb. Over the years, Libya, Iraq, Pakistan and other nations have voiced openly their aspirations to have nuclear weapons.

In the years to come the prevention of nuclear proliferation may be made more difficult with the introduction of nuclear reprocessing, which will increase the worldwide plutonium supply considerably. Extraction of plutonium from spent fuel—with its nuclear weapons applications—now takes place in reprocessing plants in France, Germany, India, Britain, the Soviet Union and the United States. By the end of the 1990s, 350 to 400 metric tons of plutonium will have been separated from spent fuel in the West. If history is any guide, we will find that, as this material enters nuclear commerce, the opportunity for nations and terrorists to divert poorly safeguarded stocks will increase. In the United States alone, 9,000 pounds of plutonium are already missing from the books. Although a large fraction may have been lost due to accounting errors or the manufacturing process, the lack of accountability portends a troubled future. The same problem applies to exported, highly enriched uranium, which the United States has had trouble tracking.


For decades, policymakers have attempted to lessen the global risks of nuclear energy. Proposing the ambitious "Atoms for Peace" program in 1953, President Eisenhower called for the creation of the International Atomic Energy Agency. Although established as a promotional body, the IAEA gradually acquired a monitory function to verify that nations were not diverting sensitive materials for atomic weapons production; addressing other risks, including plant safety, was a lesser priority.

Through the mid-1960s, the IAEA was an ineffectual policing organization. Moscow, its allies and some developing nations strongly opposed its regulatory functions. The U.S.S.R. repeatedly denounced the concept of safeguards, i.e., onsite inspection of nuclear materials, as a "spider’s web, which would catch in its threads all the science and scientists in the world." India also opposed the agency:

If safeguards are applied by the Agency to those states which cannot further their atomic development without the receipt of aid from the Agency or other member states, the operations of the Agency will have the effect of dividing member states into two categories, the smaller and less powerful states being subject to safeguards, while the greater powers are above them. This will increase, rather than decrease, international tension.

However, the United States and its allies persisted. Roll-call votes became a regular component of the debate, with the West usually prevailing in the effort to establish safeguards.

By the time the IAEA was operational at the end of the 1950s, the United States had signed some 40 bilateral nuclear cooperative agreements. These agreements allowed American inspection of U.S.-supplied nuclear materials to ensure their peaceful use. However, many importers did not wish to accede indefinitely to U.S. scrutiny. With the additional personnel burden that inspections placed on the United States, U.S. officials increasingly looked favorably upon the IAEA as an enforcement vehicle, adding impetus to creation of an agency inspectorate.

By 1959 the Soviet Union was reevaluating its own position, spurred by China’s nuclear development, which the Soviets supported until they recognized its threat to themselves. The Soviet Union then became both a supporter and innovator of nuclear safeguards; it limited exports to only light water reactors that did not use weapons-grade material; it required all spent fuel to be returned to the U.S.S.R. to minimize the danger of plutonium diversion by importers; it forbade East European recipients to build their own enrichment and plutonium reprocessing installations.

The tightening Soviet export criteria did not initially include a role for the IAEA, but as the United States increasingly surrendered safeguard responsibilities to the agency and as West European nations overcame their own concern that the organization would be used for industrial intelligence, the Soviets became more supportive. The IAEA expanded its safeguards to cover reactors of all sizes in 1964 and to reprocessing and fuel fabrication plants in 1966 and 1968. The IAEA enforcement staff has grown from three inspectors in 1961 to 250 today. The agency now devotes over one third of its budget to more than 2,000 inspections conducted each year, compared to less than four percent of total expenditures devoted to safeguards in 1962.

The willingness of nations to allow extra-national entities to scrutinize a most sensitive industry is remarkable indeed. Warren Donnelly, a senior analyst of nuclear affairs for the Congressional Research Service, noted: "[IAEA] safeguards are unique. It is the first time in history that sovereign states have invited an international organization to perform inspections on important installations within their territories."

For all its progress, this international nuclear regime has been subject to increasing stress. In 1974, India detonated a "peaceful nuclear device" manufactured with equipment imported from abroad and indigenous materials; the plutonium was produced in a Canadian-supplied reactor using heavy water from the United States, and the device was powered with Indian uranium. India’s action was disruptive, but it did not prove as damaging as some feared. New Delhi did not go on to build up a nuclear weapons arsenal.

Yet the event had both negative and positive consequences: it provoked Pakistan to undertake development of a bomb, but also encouraged nuclear suppliers to convene talks that resulted in export guidelines in 1977. When Pakistan and South Korea tried to acquire reprocessing plants from France, the United States prevented consummation of these arrangements. In the process, Washington stimulated European sensitivity to nuclear trade (although it failed to prevent West Germany from exporting to Brazil reprocessing and enrichment technology). Alerted by the Soviets to a planned South African nuclear detonation in 1979, the United States and other governments pressured Pretoria to desist.

The most dramatic and controversial effort to prevent nuclear proliferation occurred in June 1981 when Israel destroyed Iraq’s Osirak reactor, which Israel suspected was a guise for developing an Iraqi nuclear weapons program. The Israeli action stimulated concern for the vulnerability of nuclear plants to military bombardment, but also had political consequences. IAEA members voted to withdraw Israel’s credentials to attend the agency’s annual general conference in 1982, in part because of the attack, in larger measure because of Israel’s invasion of Lebanon. The United States then walked out of the meeting and suspended its support for the agency. Although the matter of Israeli participation in the IAEA was resolved within a few months and the United States reaffirmed its commitment to the agency, the organization’s ability to deal with external strains was brought into question.

Still, the IAEA has been able to cope with this and other challenges. Throughout the 1970s and 1980s it steadily expanded the number of its safeguard agreements with Non-Proliferation Treaty signatories. To address the issues of nuclear safety and security, the agency formalized a set of voluntary guidelines and established standing expert committees—albeit with limited resources—to assist nations requesting advice. Complementing these efforts are bilateral, regional and multilateral ventures. The United States has nuclear safety cooperation agreements with a number of countries. It also works with the membership of the Organization for Economic Cooperation and Development’s Nuclear Energy Agency to share safety information. Argentina and Brazil recently entered into an agreement whereby each will warn the other of mishaps; they also have discussed measures to minimize nuclear proliferation, thereby helping to reinforce the Treaty of Tlateloco, which prohibits nuclear weapons in Latin America. Euratom continues to verify the peaceful and safe development of the atom in Western Europe. Finally, in 1985 parties to the NPT held their third review conference. As opposed to their experience in a previous meeting in 1980, in 1985 the signatories reached a formal consensus in support of the nonproliferation regime. This consensus will be reviewed in 1990, and tested in 1995 when the NPT is due to expire and parties are given the option to extend it.

In sum, the international nuclear energy regime is battered but still viable. In the 1950s few thought it would get off the ground; in the 1960s few believed that safeguards could be implemented in a world of sovereign nations. Yet the regime has made remarkable progress, despite the strains experienced in recent years. Ideally, Chernobyl should provide the impetus to reinforce a new regime, based on common interest. No nation wants a nuclear accident; no nation wants nuclear weapons used on its territory; no nation wants sabotage or military destruction of its plants. Save for the phasing out of nuclear energy, there are few options. There will always be risks, but the risks can be minimized by authoritative institutions ready, willing and able to apply preventive medicine.


Before Chernobyl, attention focused on minimizing the dangers of nuclear weapons proliferation. Some suggested ways to discourage states from acquiring nuclear weapons include expanded sales of conventional arms, strengthening alliances, creating nuclear-free zones and establishing multilateral fuel banks, reprocessing plants and spent-fuel storage facilities. To address safety questions, calls are also heard for the development of a new generation of inherently safer reactors, which could come on line in a decade or two.

But in the wake of Chernobyl, pessimism seems to dominate. The principal IAEA remedies enacted in the aftermath of the incident are two conventions: one mandates that states suffering accidents which may result in "an international transboundary release" notify nations that "are or may be physically affected," either directly or through the IAEA. The second agreement encourages international assistance for nations suffering nuclear power accidents. Although valuable and negotiated remarkably quickly—the impetus of Chernobyl overcoming earlier opposition—these agreements are inadequate, akin to closing the barn door after the horse has escaped. They will not prevent future Chernobyls. The IAEA is mindful of this fact. It is looking to enhance safety through a modest increase in its budget, creation of study groups to examine such matters as the interface of nuclear operators and their machines and the expansion of expert committees that can advise nations about safety. But even this would not be enough, a fact acknowledged when West German Chancellor Helmut Kohl called for an international institution to prescribe safety standards and verify implementation, and when Hans Blix, director general of the IAEA, suggested binding, minimum safety standards.

These last recommendations should be viewed as a point of departure to address nuclear energy risks comprehensively. Had the Chernobyl accident resulted in a very large number of early fatalities rather than potentially numerous later fatalities, and had it occurred in a country subject to intense press scrutiny, perhaps there would have been impetus for greater international regulation. While the Soviet Union displayed remarkable willingness to expose itself to international review and criticism in the IAEA post-accident review conference last August, its openness should be considered only the beginning of an intense effort to examine more effective modes and methods for managing nuclear energy.

The question remains: how far will countries go to increase national and multilateral scrutiny of nuclear energy? It is not enough to have adequate controls in one nation and not in its neighbors. Given the concern over state sovereignty, it is probably unrealistic to expect foolproof international remedies. But because nations share the common goal of minimizing nuclear risks, I have proposed several multilateral alternatives which I call International Nuclear Reviews (INRs), distinguished by a formal structure and incrementally increasing authority, that address the whole range of nuclear risks. The diverse efforts of such a comprehensive approach could mutually reinforce one another to better ensure a benign nuclear regime.

The most elementary form would require the IAEA to publish a set of guidelines, portions of which would be relevant to exporters, importers and domestic manufacturers. These guidelines would require definition of the contemplated nuclear project, its economics, the operator’s disposition toward nuclear weapons, the facility’s vulnerability to sabotage and military attack, safeguards against diversion of sensitive nuclear materials, and the operator’s ability to run the plant and to manage materials safely.

The assessment also would have "action options" —suggested remedies, incentives and sanctions that might be applied by IAEA member-states should an operator be found out of compliance with INR guidelines. The checklist in this least restrictive INR would not be enforceable but it would serve as a standard of conduct, just as IAEA guidelines on physical security and nuclear safety already do.

More elaborate INRs, to be adopted as confidence in such a regime grew, would upgrade the expert assessment of proposed facilities, define the obligations of suppliers and importers and even make compliance with the impartial findings mandatory. This would be consistent with the call of Morris Rosen, IAEA director of safety, for "development of a clear and universally acceptable approach to safety guided by an international body composed of prominent exporters [that] could alleviate national and international safety concerns, and . . . also positively influence public opinion." In the most ambitious form of INR, the IAEA would license all nuclear facilities, whether imported or domestically produced.

Would such an incremental but comprehensive approach work? Certainly all the stages of INRs are more comprehensive, and most are more authoritative, than the institutions addressing nuclear risks today. In Chernobyl the most powerful INR might have prevented the accident by, at a minimum, requiring installation of quick-starting emergency diesel generators (which are available in other countries), thereby eliminating the rationale for the ill-fated "safety" test in the first place. Efforts of this sort are not cost-free. Nations would have to accept greater international scrutiny of their nuclear programs. But this burden is not entirely foreign to today’s nuclear regime, which embraces IAEA and NPT nuclear safeguards. Considering the dangers now evident, it is reasonable to argue that the cost involved is nominal compared to the consequences of another Chernobyl—or worse.

You are reading a free article.

Subscribe to Foreign Affairs to get unlimited access.

  • Paywall-free reading of new articles and a century of archives
  • Unlock access to iOS/Android apps to save editions for offline reading
  • Six issues a year in print, online, and audio editions
Subscribe Now
  • Bennett Ramberg is Senior Research Associate at the Center for International and Strategic Affairs at the University of California, Los Angeles. His most recent books include Nuclear Power Plants as Weapons for the Enemy: An Unrecognized Military Peril and Global Nuclear Energy Risks: The Search for Preventive Medicine. He would like to thank the following people for helpful comments: Steven Aftergood, Jan Beyea, Michael Brenner, Frank Congel, Warren Donnelly, Robert Gale, Daniel Hirsch, Joseph Pilat, Gordon Thompson and Charles Shapiro.
  • More By Bennett Ramberg