The Back Page

"Dirty Bombs:" The Threat Revisited

By Peter D. Zimmerman

Peter D. Zimmerman

Ed. Note: This is the full version of the article by Peter D. Zimmerman. A considerably shortened version appeared as the Back Page in the print edition of the March APS News.

Many Americans first heard the term dirty bomb on June 10, 2002, when Attorney General John Ashcroft announced the arrest of Jose Padilla on the charge of plotting to detonate a device containing both high explosive and very radioactive material. In that announcement the attorney general used the following definition: "[A] radioactive ‘dirty bomb’ involves exploding a conventional bomb that not only kills victims in the immediate vicinity, but also spreads radioactive material that is highly toxic to humans and can cause mass death and injury."

On March 6 of the same year, the Senate Foreign Relations Committee held a hearing on the question of radiological dispersion devices (RDDs), the technical term for dirty bombs, and their ability to cause casualties and damage. Experts from inside and outside government testified that an RDD could cause economic harm but was unlikely to cause deaths or injuries beyond the area immediately destroyed by the high explosives used to spread the radioactive material.

Proper preparation for an incident of radiological terror requires an understanding of the real effects of an RDD attack, yet these two views of the effects are in direct conflict.

From the long list of known radioactive isotopes only a few stand out as being highly suitable for radiological terror. These are cobalt-60 ( ⁶⁰Co), strontium-90 ( ⁹⁰Sr) (and its short-lived daughter, yttrium-90), cesium-137 ( ¹³⁷Cs), iridium-192, radium-226, plutonium-238, americium-241, and californium-252.

Types of Damage

Deterministic Injuries. Radiation is said to cause deterministic harm if an individual can be identified who received a known exposure to radiation and became ill as a result. Such illness or injury can include classic radiation sickness (hematological effects, loss of appetite, vomiting and other gastrointestinal damage, hair loss, death) or radiation burns on the skin. In general, the threshold dose for deterministic injury is quite high.¹ Loss of white blood cells is detectable at a whole body dose of 25 rem in some individuals and in most at whole body doses in excess of 50 rem.² It is unlikely that the victim will report illness. Vomiting sets in at whole body doses between 100 and 200 rem and hair loss at about 300 rem. A dose of 400-500 rem is generally considered lethal to half the exposed population. However, prompt doses—those coming directly from external radioactive material—above 25 rem are exceedingly unlikely for most RDD scenarios.

Stochastic Injuries. Given common assumptions that any radiation dose, no matter how small, can cause harm and that the biological response increases with the size of the dose, it is conceivable that some individuals exposed to quite small doses of radiation might develop cancers of known types. Their risk of developing the disease can increase with increased radiation exposure (this is certainly true for whole body doses in the several 10s of rem range). This is a statistical calculation that cannot identify a specific cancer victim, even one known to have been exposed to radiation, and assert that his or her cancer was caused by the exposure. Approximately 2000 Americans in every 10,000 will die of cancer. It is impossible to identify a specific cancer victim exposed to radiation as the 2001st victim and to determine that the person would not have developed cancer had the exposure not occurred.

Economic and Psychosocial Damage. Economic and psychosocial effects are likely to be the most serious damage mechanisms from any use of an RDD. The fear of ionizing radiation is a deep-seated and frequently irrational carry-over from the Cold War. The threat of a radiological attack on the United States is real, and terrorists have a broad palette of isotopes to choose from. An RDD attack is unlikely to cause mass deaths, but it could cause tens to hundreds of fatalities under the right circumstances, and is essentially certain to cause great panic and enormous economic losses.

Sources of Material

Radioactive material suitable for use in a radiological dispersion device may be found, stolen, or purchased legally. The radioactive materials most likely to cause great harm, based only on their physical properties, are also ones that have significant commercial applications and are widely available. They are employed in thousands of different medical, academic, agricultural, and industrial settings around the world, including medical therapy, food irradiation, communication devices, navigation beacons, and oil well logging. This makes it extremely difficult not only to secure, but also to regulate these sources. The prevalence of these sources in the public domain, coupled with inadequate control and monitoring mechanisms, poses a significant threat to health and security, not only from the possible terrorist use of radioactive materials, but also from accidents.

The U.S. Nuclear Regulatory Commission has estimated that approximately one licensed U.S. source is lost every day of the year. These "orphan" sources have escaped proper control and their locations usually are unknown. An August 2003 GAO report states that from 1998 to 2002 there were over 1300 incidents in which sealed sources were lost, stolen, or abandoned in the United States.⁴

Theft of sources meant for field radiography is not unknown. Gamma ray cameras used in the field to check the integrity of welds weigh about 50 pounds and are roughly the size of a lunch bucket. They are quite portable and relatively valuable (they cost upwards of several thousand dollars). Other small or well-shielded sources are also vulnerable to theft by comparatively untrained personnel and pose very low risk from radiation exposure unless the shielding has been removed.

Two of the worst radiation accidents, the Goiânia tragedy and the 1984 Juarez, Mexico melting of ⁶⁰Co as scrap steel (from an abandoned and stolen teletherapy source), were the direct result of the theft of the radioactive material from abandoned radiation therapy facilities.

Other potential candidates that might be vulnerable for theft by extremely well organized and well-financed terrorist groups include "megasources" such as Russian radioisotope thermal generators (RTGs) and Gamma-Kolos seed irradiators.

By far the most likely route for terrorist acquisition of intermediate quantities of radioactive material (100-10,000 curies) is open and legal purchase from a legitimate supplier. Until some time after the World Trade Center and Pentagon attacks, regulation of radioactive sources was geared towards ensuring the safe use of the material by people and organizations presumed to be acting without malice.⁵ Inspections of facilities designed to hold moderate to large sources, such as those used in industrial radiography or teletherapy, rarely took place until at least six months after a license was issued and the source shipped.

Nuclear Regulatory Commission (NRC) officials report that they have begun the process of revising licensing regulations for acquisition of radioactive sources and that they have taken interim steps to determine that license applicants are unlikely to divert material to illicit uses. These steps have not yet been publicly described.⁶

The United States system of licensing of users of radioactive sources is fragmented between so-called Agreement States, which have been delegated by the Nuclear Regulatory Commission to regulate sources within their boundaries, and Non-Agreement States, which are regulated only by the NRC. Many observers contend that local regulatory authorities are better able to track users than is the more distant NRC. In the region surrounding the Nation’s Capital, Maryland and both Carolinas are Agreement States, while Virginia, West Virginia, Delaware, New Jersey, and the District of Columbia are not.⁷

A determined and well financed group feasibly could obtain even quite large sources openly. Additionally, many smaller sources are vulnerable to loss or theft. Finally, because very large and vulnerable sources exist in the former Soviet Union, a rigorous system of accounting for existing sources and detailed laws regarding the safe storage, sale, and shipment of these sources must be supported to ensure that accidental and intentional radiological incidents do not threaten American interests or security.

The 1987 Goiânia, Brazil Event

The tragic radiological accident that occurred in Brazil between 13 September 1987 and March 1988 is the closest event to a true RDD attack. While the parallels are not exact, study of the incident provides some insight into the possible progress of a case of radiological terrorism.

On 13 September 1987, two scrap metal scavengers broke into an abandoned radiotherapy clinic and removed a source capsule from the protective housing of a teletherapy machine. The International Atomic Energy Agency (IAEA) estimates that the source capsule contained 1375 Ci of cesium-137 chloride ( ¹³⁷CsCl) in soluble form. The capsule had been abandoned when the Instituto Goiâno de Radioterapia (Goiânia Institute of Radiotherapy) moved to a new location in the city two years earlier. The two thieves took it by wheelbarrow to the home of one of the men, a distance of half a kilometer. The same day both men were vomiting because, they assumed, of bad food they had eaten. The next day one of the men had diarrhea and a swollen hand.⁸

On 18 September the crucial event that precipitated the radiological incident occurred; one of the thieves punctured the 1-mm thick window of the source capsule, allowing the powder to leak out. That same day the assembly was sold to a junkyard owner, who had an employee take the apparatus to the junkyard by wheelbarrow and leave it in a garage. That night the junkyard operator, D.F.⁹ saw that the powder glowed blue. Intrigued by the glowing blue material, he took the capsule into his house to show it off to his family and friends.

D.F. also passed out fragments to his family. At this point several people sprinkled or rubbed the material on their bodies as they might have done with Carnival glitter.

M.F.1, the wife of D.F., became ill with symptoms of acute radiation sickness on 21-23 September. Her mother, M.A. 1 nursed M.F.1 for two days, and then returned to her home outside Goiânia, taking " a significant amount of contamination" with her. M.A. 1 ingested 270 μCi of ¹³⁷Cs and received a dose of 430 rad. Although this is close to the lethal dose for half the population (LD₅₀), she survived. Over the next few days the rotating assembly of the source was disassembled by two of D.F.’s employees; both died having received estimated doses of 450 rad and 530 rad. W.P., one of the thieves, was admitted to the Santa Maria Hospital for 4 days and then transferred to the Tropical Diseases Hospital.

The saddest incident occurred on 24 September. Six-year-old Leide das Neves Ferreira (L.F.2 in the IAEA report) played with the colorful source powder, painted it on her body, and ate a sandwich while her hands were contaminated. She was massively internally contaminated (27 mCi) and received a 600 rad dose. She died on 23 October.¹⁰

The correct diagnosis of acute radiation sickness was made by Dr. P.F. of the Vigilancia Sanitária on 28 September after M.F.1 and G.S., an employee of D.F.’s, took the remnants of the rotating assembly to Dr. P.F.’s office at the clinic of the Vigilancia Sanitária. The two individuals, M.F.1 and G.S., carried the material in a bag and took a public bus to the clinic, thus contaminating the bus and exposing other passengers to the cesium.

The toll in Goiânia is staggering. In partnership with a team from the IAEA, Brazilian authorities monitored over 112,000 people in the city’s Olympic-sized soccer stadium for radiation exposure and sickness. According to the IAEA report on the incident, a total of 249 people were identified as contaminated by the Cesium-137, 151 people exhibited both internal and external contamination, 49 people were admitted to hospitals, with the 20 most seriously irradiated having received doses from 100 to 800 rads. The internally contaminated patients were themselves radioactive, seriously complicating their treatment. In the end, 28 people suffered radiation burns and five people died, including three men, one woman, and one child.¹¹

Eighty-five buildings with significant levels of contamination were found. Of these dwellings, seven were determined to be uninhabitable and subsequently destroyed; 200 people were evacuated from another 41 buildings.

Both patients and technicians spread radioactive contamination in Goiânia and even to Rio de Janeiro. For several days nobody remembered to decontaminate the ambulances used in Rio to transport victims from the airport to the naval hospital, which had the country’ s primary facility for the care of radiation sickness.

A total of 3,500 m³ of radioactive waste was collected and trucked to a temporary disposal site. Most of the original source material was recovered intact.

The radiological incident in Goiânia demonstrated the far-reaching consequences that a radiation incident, whether accidental or intentional, can cause.

What to Expect

Most RDD scenarios tend to focus on a device that uses high explosive to pulverize and disperse radioactive material. During the March 2003 International Conference on Security of Radioactive Sources, held by the IAEA in Vienna, Austria, it appeared that most of the world’s radiation protection authorities had adopted that simple scenario as the most plausible. Most seemed to accept the hypothesis that terrorists would be incapable of handling radioactive sources in relative safety or performing simple chemical operations on whatever radioactive material they might obtain.

While many terrorist groups are incapable of obtaining or using sophisticated technology, some are capable. We cannot rely on the premise that terrorists are unwilling to die attempting a devastating attack, for we know from experience that many are. Also, we know from Osama bin Laden’s videotaped comments about the September 11, 2001 attacks that terrorists will not necessarily know they are about to die. And while most terrorists may not be sufficiently imaginative or skilled to carry out such an attack, enough are to cause concern.

Radioactive material can be disseminated in the form of discrete sources. Some forms of isotopes can be dissolved in solvents and sprayed widely; still others can be burned or vaporized. Any complete plan to respond to an RDD must take into account all of the reasonable ways such a device might function, including those so stealthy that the population might ingest or inhale significant doses before an attack becomes apparent.

Generalizations about the RDD threat spectrum can be misleading. Possible devices range from a small package of explosives (< 100 kg) wrapped crudely around a comparatively small radioactive source (1-10 curies) detonated in a crowded area. At the high end of the spectrum up to several tens or hundreds of thousands of curies of material could be dispersed by a sophisticated device, the whole project requiring several physical and chemical processes to assemble and use the device effectively as a weapon.

The most attention has been given to the small, readily achievable dirty bomb, which may indeed be the most probable type of radiological attack. However, almost all experts agree that such an attack would be unlikely to cause mass casualties; rather it probably would cause great disruption and panic.

Very little analysis has been done on the maximum credible events, which have escalated from something resembling the Goiânia incident in Brazil in 1987 (2001 estimates) to present estimates involving hundreds of thousands of curies of ⁹⁰Sr or ¹³⁷Cs extracted from Soviet era devices. Almost certainly only a dedicated and well-financed group could pull off a maximum credible event. However, it is likely that some of the major international terror groups, including Al-Qaeda, have not only the resources to carry out such an attack, but also the willing martyrs, whose participation would significantly reduce the cost and complexity of an attack. It is very nearly impossible to disperse radioactive material from an explosively powered dirty bomb in such a way that victims externally absorb a lethal dose of radiation from the source before they are able to leave the affected area.

 Stealthier RDDs, not involving explosions, might actually cause deterministic radiation injuries in more people than would a bomb because remedial action might be delayed or because the RDDs might be designed to promote ingestion or inhalation. Even a small RDD is likely to do a great deal of real economic damage because of two principal effects: suspension of economic activity and long-term contamination of property, possibly resulting in its permanent loss.

Casualties

While many analysts have suggested that RDDs will neither sicken nor kill very many people, analysis of the Goiânia incident leads to a modification of this conclusion and to a caution: of the 249 contaminated victims of the Goiânia incident, 151 were contaminated internally. That is, they either ate or inhaled radioactive cesium, and the material was incorporated into their bodies. While the amounts ingested seem extremely small (Leide das Neves Ferreira, who died, was the most highly contaminated having consumed only 27 mCi), they were more than adequate to cause death or acute radiation sickness.

These minuscule quantities could be transferred from a hand with a little radioactive dust on it to the mouth with the kinds of simple gestures people make all the time. Thorough hand washing, before doing anything else, is probably among the most useful and time-urgent treatments. If the air remains dusty, hand washing may be ineffective, while dust masks become essential

Fortunately, there are drugs that can assist in purging the body of cesium contamination. The dye Prussian Blue is sold for this purpose under the trade name Radiogardase® by Heyl Pharmaceuticals in Germany. Prussian Blue was found very effective in Goiânia The national stockpile of products for use in the event of an emergency includes stores of Prussian Blue, but it would be appropriate for the U.S. government to ensure that the stockpile contains more than the amount needed to treat victims of a single, severe attack. The drug is far more effective given within 2-4 hours after exposure to cesium than it is later, so many geographically diverse storage sites are needed.

Not all of the internally contaminated patients at Goiânia participated in the events during which the ¹³⁷Cs was known to have been handled. Many people were victims of secondary contamination (they came in contact with persons who had been in direct contact with the source) or even tertiary contamination (there was an additional intermediate person or other vector between the internally contaminated victim and the radioactive source). It is known that many internally contaminated victims came into contact with the radioactive cesium in bars and restaurants.¹³

The Brazilian authorities moved to seal off the central area where contamination was known to be present. This action was effective in excluding human beings but not feral cats. Probably the fur of the animals became contaminated and they spread radioactive material beyond the central area.¹⁴

Because people might ingest or inhale radioactive material, it is not reasonable to assume that the human toll from a large RDD would be small or negligible outside the direct range of a dirty bomb blast. The U.S. should be prepared to cope with tens, hundreds, or conceivably thousands of victims of acute radiation sickness. Patients with internal contamination also pose a hazard to attending medical staff. The caregivers may be forced to limit their time with the patient or to work from behind shields or both.

Range of Sizes

A Small Device (1-100 Ci). This first case considers an unsophisticated RDD containing, at most, 100 curies of a gamma-emitting isotope such as ⁶⁰Co or ¹³⁷Cs dispersed by less than 100 kg of high explosive.

Regardless of how small the radioactive device, all areas that may have received some radioactive material will have to be evacuated and closed off for monitoring and decontamination. This is likely to include checking both interiors and exteriors of buildings for radiation. In all likelihood, such an examination would take several weeks or more to estimate the contamination over an extended area. During the initial monitoring period, it is nearly certain that all economic activity in the affected area would cease, in part because of the need to determine the extent of contamination, and in part because of the reluctance of the public to enter an area thought to be radioactive, no matter how small the dose rate. The period of mandatory evacuation resulting from the need to take precautions against even a very small device is certain to be several days, and could be many weeks or even months.

During the evacuation period, small and undercapitalized businesses, such as small delicatessens, independent bookstores, and clothing stores, will suffer from diminished or even zero cash flow. In turn, small business owners will need to furlough or fire employees, will more than likely be unable to pay suppliers (who will then suffer cash flow problems), and probably will be unable to pay mortgages. Even with business interruption insurance, a wave of bankruptcies is likely to follow, unless the government steps in and offers subsidies to everyone from business operators to owners of buildings to mortgage holders. However, all commercial insurance policies sold in the United States appear to exclude damage from radiation. Residents living within the affected zone will also need to be evacuated and sheltered, adding to the already high economic cost associated with the RDD incident. It is unlikely that they will be able to return to their homes for weeks or months, if at all.

Furthermore, the streets in the affected area will require decontamination, as will the exteriors of buildings. Depending upon the location of air intakes and open windows, interiors may also require treatment. Unfortunately, there are no well-established technologies for wide area decontamination of modern built-up areas.

In Washington, D.C., an area the size of the National Mall could be affected by a simple dirty bomb— perhaps a few curies of material and a few kilograms of explosive—though the target would most likely be a government facility or a business or residential district, not just open space. More efficient RDDs relying on other means to disseminate the same amount of radioactive material could easily contaminate a significantly larger area.

A Large RDD (1,000 - 10,000 Ci). The response of the Brazilian public and government to the Goiânia incident and the effects of the radioactive material approximate the experience of an RDD event and are enormously instructive.

The majority of the damage done to the Goiânia region was caused by the nearly total cessation of economic intercourse with the rest of Brazil. The area’s primary business is agriculture. As a result of the incident it became impossible for farmers in the area to sell any of their produce to the rest of Brazil. In order to circumvent the boycott, local farmers took to labeling their products as grown in nearby, unaffected areas. The local government, hoping to "make lemonade" from the sour affair, changed its flag to include the trefoil that symbolizes radiation.

Super RDDs (> 10,000 Ci). It is difficult to predict the consequences of an attack using this much radioactive material, however, we can glean some information from previous incidents. The Chernobyl reactor fire, for example, released a large amount of material but injected most of it high into the atmosphere. In this case, an entire city, Pripyat, and a large agricultural area were abandoned and fenced to prevent unauthorized entry

It is not difficult to imagine sophisticated devices that could kill tens and sicken hundreds, and it is not impossible to envision devices that could be ten times as lethal. Nevertheless, an RDD is first of all an economic weapon. Cost estimates to restore lower Manhattan after the September 2001 attack range up to $40 billion plus loss of economic activity. The consequences of a large or super RDD might well be more costly.

Types and Mechanisms of Damage

Decontamination Levels and Economic Damage. On average, natural background radiation from all sources is 300-360 milli-rem/year including 0.2 rem/year from natural radon gas.

The intensity of cosmic radiation increases with altitude, so moving from Washington, D.C., to Denver, Colorado increases the background to 500 milli-rem/year. One might expect cancer rates in Denver to exceed those in Washington, D.C. because of the higher cosmic radiation background found at higher altitudes. As well, the Denver area is situated over uranium-bearing rocks, which provide a steady stream of radioactive radon gas. However, in reality, the cancer rates in the two cities are quite similar.

Acute Exposure to an RDD. How do background and medical procedures compare with doses from the kinds of sources likely to be used in RDDs? The dose rate from one curie of ¹³⁷Cs at one meter is 0.4 rem/hour.. Standing next to such a source for a year (8,760 hours) would result in 3,500 rem exposure, an amount almost 12,000 times the normal background dose and certainly lethal.

However, no victim of an RDD attack using explosively dispersed radioactive material will spend more than minutes or at most hours close to the source of radiation. The important thing to remember about exposure to a dirty bomb is that anyone who survives the initial bomb blast should have no problems leaving the area in time to avoid injury from external sources of radiation The most likely ways for an RDD to sicken or kill victims with radiation are by stealthy dispersal of radioactive material or distribution of lump sources that go undetected by the local civil defense authorities (something unlikely to be possible very much longer)¹⁵, or by use of a non-eplosive device that contains amounts of radioactive material sufficient to cause serious external irradiation (25,000 Ci or more would be a reasonable estimate) or that causes radioactive material to be ingested or inhaled, producing internal exposure.

Local authorities should be prepared to treat a number of cases of acute radiation exposure, but most hospitals, however, do not have specialized clinics for treating radiation injuries or contaminated patients.

Removal of external contamination can be accomplished simply by thorough washing. Internal exposure, poses far greater hazards to the victim, whose tissues are being continuously irradiated from the inside. Internal contamination can occur in many ways: Leide das Neves Ferreira rubbed the ¹³⁷CsCl material on her body and subsequently ate a sandwich that was believed to have been contaminated with material from her hands.

Patients who have been internally contaminated must be treated with due regard for the fact that they and their human waste are radioactive and that everything that comes into contact with them will become contaminated. They will require special facilities, and appropriate instrumentation to measure the degree of contamination. The medical caregivers will need to take precautions to prevent their own contamination. If the Brazilian experience is any guide, not all physicians will be willing to accept the risks attendant to treatment of internally contaminated victims.

Long-Term Exposure to Contamination From an RDD. Based on the linear, no-threshold model (LNT), it is often stated that exposure to low levels of radiation for long periods of time may lead to an increase in the death rate from cancer.¹⁶ Using the LNT assumption, exposure to one rem of radiation results in a 4 in 10,000 increase in the cancer death rate (the data largely come from atomic bomb survivors, whose results may not be typical of long-term exposure to low doses). About 2000 out of every 10,000 living Americans will die from all forms of cancer; it is not clear that an additional four cases per 10,000 could be detected and attributed.

The present U.S. standard is that the additional cancer risk to the general population from man-made radiation (other than for medical therapeutic uses) should not exceed one case per thousand people. The Environmental Protection Agency (EPA) has set a requirement that the increased dose for the general public above background for non-medical radiation should not exceed 100 milli-rem per year (0.1 rem).

Decontamination

Two decontamination standards have been set for cleanup after a radiation release. The Nuclear Regulatory Commission allows an additional 25 milli-rem of absorbed radiation dose per year, while EPA permits only 15 milli-rem per year; both figures should be considered in the context of the 300 mill-rem/year of background radiation always present. The limits on residual radiation after cleanup are the doses a person would receive who spent 24 hours a day, 7 days a week for 40 years in an affected area. Achieving such cleanup on a large scale in a populated area with many different kinds of buildings would be difficult.

Nevertheless, under present regulations and applicable laws, any building that cannot be decontaminated so that the dose rate from residual radioactive debris from any radiation accident is below the limits set by either EPA and NRC may not be occupied. Such a structure would have to be abandoned in place and fenced off, or razed and removed, with all materials going to a low-level radioactive waste dump.

Is the remedy reasonable in light of the actual risks? Obviously not. The decontamination limits of EPA and NRC, while satisfactory for a laboratory environment and spills of small radioactive sources, limit the increased cancer risk from a terrorist attack to far less than the increased risks accepted daily by virtually all Americans. If the current limits on residual radiation levels were maintained after an attack, even a small RDD, poorly dispersed, would require the leveling of large portions of a city for an uncertain, but certainly small, reduction in the long-term cancer rate.

It is plausible that relaxing the cleanup standard by a factor of ten would reduce the area that requires intensive cleanup and decontamination by the same factor. The process for changing the regulatory standards for residual radiation in the event of attacks on the United States should be explored immediately, and any necessary legislation should be prepared.

Direct Decontamination. There are no proven methods to decontaminate the exteriors of large buildings or to decontaminate large outdoor areas, other than to remove buildings and soil. Many experts label our only presently viable technology as "muck and truck," meaning that all one can do is dig up the soil, tear down contaminated buildings, and haul all of the contaminated material to a radioactive waste storage facility. Some improvements may, despite difficulties and costs, be on the way, and may be suitable (at least) for the highest-value and most symbolic targets.

At present, the direct costs of physical decontamination of large outdoor areas are difficult, if not impossible, to estimate in a credible way.

In a recent and very realistic Swedish exercise using instruments in cars, trucks, and aircraft to search for concealed sources, only about half of the sources were found by any given team, and some sources were not found by any of the search teams.17 This does not provide confidence that all sources distributed by a resourceful terrorist would be located, even after officials knew that a search was required. It also indicates that some contaminated areas might go undetected even if an explosive or other large-scale RDD were used.

Economic Impact

The economic impact on a major metropolitan area from a successful RDD attack is likely to equal and perhaps even exceed that of the September 2001 Al Qaeda attacks in New York City and in Washington, D.C. The estimated cost to return the lower Manhattan area to the condition prior to the September terrorist attacks was in excess of $30 Billion. The immediate response costs exceeded $11 Billion.18

Much of the private cost of recovery from the September 2001 attacks was paid by insurance. That would not be the case following an RDD attack, because radiation is a specifically excluded risk in virtually all policies written in the United States. The government will have to step in to subsidize economic recovery after an attack, or some form of insurance reform will have to occur before an attack, in order to facilitate economic recovery.

Conclusions and Recommendations

Radiological dispersion devices pose a unique threat to the United States. While an RDD attack is unlikely to cause mass fatalities, it is apt to cause mass panic and great economic damage. There remain many uncertainties in the spectrum of responses.

Responses to an attack are complicated by jurisdictional issues. Some sources are regulated by the Nuclear Regulatory Commission, while others are controlled by state agencies. The NRC and the Environmental Protection Agency have significantly different cleanup standards. Finally, the plume from an explosively driven RDD is likely to cross city, county, and even state lines and require a high degree of cooperation among unrelated organizations in the face of likely mass panic. A great deal of additional effort to pre-plan local responses is required.

The following specific recommendations should be implemented:

1. The Department of Energy weapons laboratories, in cooperation with other agencies and institutions, should identify, test, and deploy technologies that will enable rapid cleanup and decontamination of buildings, vehicles, and people.
2. To reduce economic disruption, the permitted level of residual radioactivity after cleanup from an attack (not ordinary radiological accidents) should be raised by a factor of ten. If this requires legislation, the Administration should develop a bill and send it to Congress. Acceptance of the increased levels of residual radiation will require a program of public education about the risk that should begin soon.
3. Because Americans cannot presently obtain any sort of insurance to cover radiological terrorism, it is all but certain that even the smallest of attacks will result in economic catastrophe for the victims. The cost of cleanup, even if feasible, is likely to be too great to be borne by individual owners and businesses. Even if an individual homeowner or a specific business can afford to decontaminate a dwelling, store or factory, it is not good public policy to push the costs onto a few and to abandon the many who are unable to afford restoration.
4.  Just as the Federal government provides subsidies for flood insurance, so it should also provide some form of national insurance against nuclear terror. There is ample additional precedent. The Price-Anderson Act already provides insurance in the event of a nuclear accident caused by a licensed company or facility acting within the terms of its license.
5. It is likely that raising the permitted level of residual radioactivity in order to reduce the area requiring intensive decontamination will reduce the property values in the affected zone. This can be offset by a one-time direct payment by the federal government to the property owners. Some kind of legislative or regulatory remedy should provide relief to Americans in the wake of an RDD attack.
6. NRC licensing rules should continue to be toughened.
7. The United States should stockpile Radiogardase® in sufficient quantities to treat at least 1,000 victims in each of ten cities for at least one month. The medication should be available in any city within 2 to 4 hours after an attack.
8. Programs to recover orphan sources in the United States and abroad should be fully funded on a continuing basis. A one-time cleanup of known sources will not protect against sources orphaned in the future.
9. Very large radioactive sources, particularly those used in the former Soviet Union, should be retired and replaced with benign technologies.
10. Where feasible, non-radioactive technologies such as X-rays and accelerators should be substituted for radioactive sources
11. An appropriate program of public education about the high probability of surviving an attack without serious injury or additional risk of cancer should be instituted in a timely manner.

Radiological attacks against the United States are a matter for urgent concern, but not for panic.

Endnotes

1. Samuel Glasstone and Philip J. Dolan, The Effects of Nuclear Weapons, 3d ed. (Washington, DC: U.S. Departments of Defense and Energy, 1977), 578, et seq.
2. The customary unit by which "absorbed dose" of radiation is measured in the United States is called the roentgen after Wilhelm Roentgen, who discovered x-rays. Since not all radiation is equally effective at inflicting damage on humans, a "quality factor" has been introduced to include the damage. The quality factor for beta and gamma radiation is taken as 1 and that for alphas and neutrons as 10. The rem or "roentgen effective in man" is a measure of effective absorbed radiation dose and is corrected for the fact that alpha radiation and neutrons are far more damaging to a human being than are betas and gammas. Hence, a rem is equal to the dose in roentgens times the quality factor. The rest of the world uses the metric unit "gray" (Gy) for absorbed dose and "sievert" (Sv) for equivalent dose. One gray equals 100 rads; one sievert equals 100 rem.
3. "Nuclear Security: Federal and State Action Needed to Improve Security of Sealed Radioactive Sources," United States General Accounting Office, August 2003, GAO-03-804.
4. Private communication with Abel Gonzalez, head of waste disposal and security, IAEA, and Richard Meserve, then-chairman, U.S. Nuclear Regulatory Commission.
5. Private communication with Richard Meserve.
6. Private communication with Joel Lubenau.
7. See www.nrc.gov/what-we-do/state-tribal/agreement-states.html
8. Most information taken from The Radiological Accident in Goiânia (Vienna: International Atomic Energy Agency, 1988). Accounts of the incident vary in their details, but it is felt that the official IAEA report forms the best available basis for subsequent analysis. Much of the text in this section is an abbreviated paraphrase of parts of the IAEA chronology, 22–29.
9. The IAEA report identifies people only by initials. Other reports use full names.
10. Alex Neifert, accessed at
    www.nbcmed.org/SiteContent/MedRef/OnlineRef/CaseStudies/csGoiânia.html
11. Where numbers in the text differ from those in the official IAEA report it is because the authors have obtained more recent and complete information from participants in the response to the event.
12. "Al Qaeda pursued a ‘dirty bomb,’" The Washington Times, October 17, 2003.
13. Department of Homeland Security Working Group on Radiological Dispersal Device (RDD) Preparedness, Medical Preparedness and Response Sub-Group. 5/1/03 version.
14. The Radiological Accident in Goiânia, International Atomic Energy Agency (Vienna: 1988), chapters 5 and 6.
15. Telephone interview with Dr. Carlos Nogueira de Oliveira of the IAEA, 12 August 2003. Dr. Nogueira headed the Brazilian government’s response in Goiânia.
16. Interview with Dr. Nogueira de Oliveira .
17. A transient stroke lasts only a few minutes and occurs when the blood supply to part of the brain is briefly interrupted.
18. "Sensors May Track Terror’s Fallout," The Washington Post, June 2, 2003, A1.
19. The linear, no-threshold rule is conservative. No firm threshold has been established. Many scientists believe that the LNT rule is not supported by experimental evidence and that there is a threshold dose and dose rate below which the body is able to repair itself.
20. R.R. Finck and T. Ulvsand, "Search for orphan gamma radiation sources: experiences from the Barents Rescue 2001 Exercise," in Security of Radioactive Sources, International Atomic Energy Agency (Vienna: 2003), pp 123-138.
21. "Financial Impact of the World Trade Center Attack," prepared for the Finance Committee of the New York State Senate, January 2002.
22. "State Farm won’t cover nuke losses." USA Today, February 27, 2003 (web edition, http://www.icnj.org/SiteDocuments/News/2-28-2003/statefarmwontcovernukelossesusa_2-28-2003.htm). This article describes the addition of a nuclear exclusion (including radioactive contamination) to State Farm auto insurance policies in addition to existing exclusions on homeowner and commercial policies. The Chubb Group of insurance companies excludes radioactive contamination from, inter alia, its marine insurance: http://www.chubb.com/businesses/cargo/ocean_marine/oceancargo/basic.html. See also "The Insurance Implications of Terrorist Attacks," Independent Insurance Agents of America, web version at: http://vu.iiaa.net/Docs/Terrorism/ TerrorismWhitePaper.pdf.

Acknowledgments

Discussions with Hans Binnendijk, Don Cobb, Charles Ferguson, Abel Gonzalez, Steven Koonin, Edward Levine, Joel Lubenau, Carlos Nogueira de Oliveira, Harry Vantine, and Greg van Tuyle have been especially important to the development of our thinking and to shaping this paper. Cheryl Loeb's contribution has been crucial to this project. See National Defense University publication Defense Horizons, #38. January 2004.

Peter Zimmerman has held positions with the US State Department and the Senate Foreign Relations Committee. He is now professor and chair of Science & Security and the director of the MacArthur Centre for Science & Security Studies at King's College, London.