A caesium-137 radioactive source was left in an abandoned clinic in the city of Goiânia, capitol of the Brazilian state of Goiás. Scavengers took the massive device, gouged out the iridium window, which allowed high gamma radiation and a beautiful blue light to escape, and sold it to a junkyard owner. The blue glow of the small cylinder of caesium chloride that was inside the device made it appear very valuable, so it was broken into pieces and used for decorative or magical purposes, including rubbing on the skin. A six-year-old girl ingested some of it. At least 244 persons had measurable contamination of caesium chloride on or in their bodies, and 20 of them were seriously radiated. Ten of the 20 had extensive decontamination and radiation-damage treatment. To date four of these people have died. The cause of this incident appears to be lack of adequate accounting and inspection procedures for radioactive sources in Brazil. The organization charged with designing and executing such procedures promotes the nuclear industry in Brazil and has close ties to the military. Also, deficiencies were evident in the handling of the incident.

Herein we describe the details of the incident, the reporting of the press, the fear of the populace, the reactions of the various government agencies, and the deficiencies in maintenance and inspection of nuclear materials and x-ray facilities in Brazil. We also discuss in general terms the nuclear industry in Brazil.


Goiânia was the second of Brazil's planned state capitols (for the state of Goiás). Founded in 1933 in the Brazilian frontier, it's now on the edge of the country's coastal population belt, 210 kilometers west-southwest of Brazil's capitol, Brasília. At first glance, with its more than one million people, scores of tall buildings, many cars and trucks, a good airport with many daily flights, a federal university and large professional cadre and business elite, Goiânia appears to be a modern city. But, like other Brazilian cities, there is a wide educational and financial gap between the upper and lower classes and a small middle class. Because of this gap, as in many developing counties, modern high technology devices and marginal ways of living are juxtaposed. The “Goiânia Radiation Incident” (herein called GRI) is a tragic example of the clash between these ways of living.

When radiation therapy devices for cancer treatment became widely available in the decades following World War II, Brazil's well-educated medical elite were quick to obtain them. Goiânia now has two modern cobalt-60 sources, an electron accelerator and possibly other devices. GRI leads one to believe that some of these sources are located in obscure clinics in Brazil and operated by ill-trained individuals, with little accounting of their location and use and negligible inspection by a regulatory agency. The federal agency Comissão Nacional do Energia Nuclear (or CNEN, which will be discussed later) has regulatory jurisdiction over radiation therapy devices. This somewhat secret organization has close ties to the military, which, incidentally, vigorously promotes both civilian and military nuclear applications in Brazil. The Goiânia Radiation Incident (GRI), which involved an outdated caesium source, will probably result in the deaths of about ten people within a few years and may considerably shorten the lives of another hundred or more. At the time of writing, four people have died.

We think the word “incident” better describes this event than the word “accident,” because it was caused by the clinic owner deliberately leaving the source in an abandoned clinic and by the government failing to adequately regulate radiation sources, rather than by a malfunction of equipment or in the operation of equipment. The press stated that GRI was second only to the Chernobyl reactor accident in the Soviet Union in severity. Although there are some similarities in contamination, radiation released and the effects on humans, we see no similarities in the causes of the Chernobyl accident and GRI.

Herein, we describe the radiation therapy and diagnostic devices used in Brazil, including x-ray machines, and discuss CNEN, the government agency that regulates these devices. In addition to detailing a probable chronology of the relevant events involving the caesium source responsible for GRI, we discuss the roles of the press, government officials and medical and physics professionals in coping with the technical and sociological aftermath of GRI. Finally, we discuss some reforms for better federal controls of nuclear radiation sources, which were proposed by scientific organizations in Brazil before GRI.

Radiation Therapy Devices for Cancer Treatment in Brazil

The first caesium-137 radiation therapy device was imported into Brazil from the United States in the 1950s. In the 1960s, more modern cobalt-60 devices were imported. Reports indicate that no government organization knows exactly how many caesium-137 and cobalt-60 sources there are in Brazil, how many caesium sources are still being used, or what has happened to some of the caesium sources that were superseded by the cobalt sources. Of course, the entire world knows what happened to one of the cast-off caesium-137 sources: It was left in an abandoned clinic in an open building on an empty lot in the middle of Goiânia.

A caesium radiation therapy device consists of a massive metal beam attached to thick concrete walls, from which extends a strong arm suspending an elongated sphere (about 60 cm diameter and 700 kg mass) on rotatable gimbals. Inside the sphere is a large amount of lead around a steel cylinder, which contains mostly lead (12 cm high, 30 cm diameter and 120 kg mass). In the center of the cylinder is a smaller cylinder (3 cm high, 3.6 cm diameter and 93 g mass) of caesium chloride powder held together by some non-radioactive material. A channel (“window”) filled with iridium extends from the caesium cylinder to the curved surface of the lead cylinder; it absorbs less of the caesium's gamma rays than the lead. The cylinder can be rotated so that the iridium window is aligned with a channel to the outside of the steel/lead cylinder when treatment is desired.

The caesium source contained 28 g of caesium chloride and 65 g of inert material when fabricated and emitted about 2000 curies [I curie=3.7 × 1010 photons/sec of gamma radiation:

Cs(55,137)->Ba(56,137) + e- + gamma ray (0.663 MeV)]

It has been estimated that at the time of GRI, due to caesium-137's 30-year half-life, there were 19 g of caesium in the source and it emitted 1370 curies. The steel/lead cylinder that encloses the caesium source apparently can be removed easily from the device and apparently would emit much gamma radiation through the window.

Brazilian Nuclear Policy: A Brief History

In the 1950s, nuclear energy was seen as a solution to the world's energy supply problem. Although petroleum was cheap, it was a non-renewable energy source, so an energy crisis was on the horizon. In nuclear energy man had achieved fantastic domination over nature; a few grams of a specific material could release as much energy as tons of coal. The enthusiasm was great, and pacific movements such as the Pugwash conferences were weakened by pro-nuclear government propaganda, which, despite the fear generated by Hiroshima and Nagasaki, emphasized the magical, peaceful side of this large energy source.

In Brazil the enthusiasm was not different. The first reactor acquired by Brazil in the 1950s was obtained through the US program “Atoms for Peace” and was installed at University of São Paulo. The second reactor was installed at the Federal University of Minas Gerais in Belo Horizonte and the third at the Federal University of Rio de Janeiro in the 1960s. Despite the fact that the first three research nuclear reactors were at large universities, they had very little connection to the academic departments of these universities.

The government agency CNEN was created in 1956 with the objective to develop nuclear energy in Brazil and institute strong controls on the export of nuclear materials. Monozoitic sands, for example, had been used as ocean ship return ballast by other countries. CNEN was largely controlled by the military after it seized power in 1964. At this point, Brazilian physicists and engineers discussed which type of reactor was best for Brazil. Some argued for a natural uranium reactor, in which the fuel could be supplied from within Brazil (this is what Argentina chose). Others argued for enriched uranium reactors, despite the fact that Brazil was not capable of producing enriched uranium. The military regime opted, at the end of the 1960s, to buy an enriched uranium 620-MW reactor from Westinghouse. It was installed at Angra dos Reis in the state of Rio de Janeiro and has never, to this day, operated properly. It was obvious to all concerned that the military wanted enriched uranium for other purposes.

In 1975 at the peak of the “oil crisis,” the government of General Geisel made an accord with Germany for the construction of eight 1300-MW reactors by 1990. The accord also provided for the possibility of Brazil acquiring sixty more reactors by the year 2000. A German company, KWU, in the Siemens group, became associated with the Brazilian government company, Nuclebrás, for fabrication of the reactor components and enrichment and reprocessing of uranium fuel in Brazil. Although only one of these proposed reactors is presently under construction, ground has been broken for the second one – both at Angra dos Reis, the site of the Westinghouse reactor. A third one is planned for the state of São Paula.

The Brazilian Society of Physics (Sociedade Brasileira de Física, or SBF) protested against the accord from the beginning, questioning the dimension of the program, the lack of previous debate, the small estimate of the hydroelectric power potential of Brazil, the dangers of nuclear energy and the vulnerability of the process of uranium enrichment by centrifugation. The reprocessed uranium and plutonium was suspected by SBF to be eventually used for building nuclear bombs. The Brazilian Society for Advancement of Science (Sociedade Brasileira para o Progresso do Ciência, or SBPC) also protested against the policies of CNEN.

During the Carter administration, the US decided against transferring this reprocessing technology to Brazil. The transfer had already begun, but after twelve years it has not proven to be economically viable. The transfer of this technology is dominated and controlled by KWU, and today it is certain that the accord does not favor the Brazilian domination that the country wants.

The economical hydroelectrical power potential estimated at that time was 100 GW, but has been revised to be 213 GW at present because of higher petroleum prices. This estimate perhaps should be increased in the future to include the possibility of transmission of hydroelectric power over long distances by superconducting lines from the Amazon. The initial estimate for investment in nuclear power was US $400 per kW, well below the current real cast of US $3,500 per kW, which is about three times the cost of traditional hydroelectric power investment.

In the 1980s, more people began to discuss the problems of nuclear energy in Brazil. The erosion of the military dictatorship, the process of democratization of the society, the lack of success of the Angra reactor, the Three Mile Island accident and, more recently, the Chernobyl accident set the tone for the debate. During the reign of the last military president (1984), a massive campaign for direct election of a president peaked, and pressure on politicians made them promise a re-evaluation of the Brazilian nuclear program. At the beginning of the present civilian government, a commission to evaluate the Brazilian nuclear program was formed. In August 1986 this commission presented recommendations for several modifications of the program. Nothing has changed.

The central point of the criticism of SBF and SBPC was the fact that CNEN had the function of implementing the nuclear energy program and, at the same time, was responsible for regulation of nuclear installations. CNEN is connected to the President through the Casa Militar, the president's military policy-making body, and therefore has very little incentive to satisfy public demands. The hole at Serra do Cachimbo (see below), secret CNEN bank accounts and the nuclear submarine program still today do not have satisfactory public explanations because of the argument of national security.

The government ignored the recommendations of SBF and SBPC, and the principal recommendation of the commission, mentioned above, that was recently formed to evaluate the Brazilian nuclear program. That recommendation was that a new commission should be created with the function of standardization, licensing and inspection of nuclear facilities.

The Fear of the Populace, The Role of the Press and Reactions of the Governments after GRI

The news that something very dangerous related to nuclear energy had happened spread through Goiânia and throughout Brazil. People immediately compared it to Chernobyl. Panic began to envelop the city. The official information was contradictory and showed a large lack of technical knowledge about how to control the situation. The government of the state of Goiás used all means of communication to avert panic. Because of the citizenry's lack of confidence in the government, which they believed to consistently deceive them, rumors flew.

Within a few days of the public announcement of GRI, an undetermined number of people evacuated Goiânia, and the American Field Service removed its United States exchange students. Many people who had planned to come to Goiânia for various reasons canceled their plans, and concerts were canceled because of the fear generated by the national and international press. News reports compared GRI to Chernobyl, Three Mile Island and other nuclear accidents. After the initial panic the city became relatively calm, because there were few visible signs of the effects of GRI, despite the continuing news of a severe problem in Goiânia.

The public fear spiked again when knowledge became available about the radioactive waste of GRI. For example, a Goiânia dentist asked one of the authors (LDR) if he was afraid to live in Goiânia. The federal government quickly decided that the waste would be deposited in a 300-meter deep hole at Serra do Cachimbo air force base in the state of Pará in northeast Brazil. This hole, according to Sociedade Brasileira de Física, has similar characteristics to the holes the US uses for nuclear bomb tests; however, the Brazilian military has always stated that they are not planning to build nuclear bombs. When the news of the hole appeared, a large religious festival in Belém, the capitol of Pará, quickly became a massive demonstration against nuclear waste being deposited there. The federal government immediately withdrew its decision. After this every other state of Brazil indicated that it would not accept the GRI waste. The president of the country then stated that the Congress would have to decide what to do with the waste, despite the fact that CNEN had, before GRI, selected several sites around Brazil for longterm storage of nuclear waste. Effectively, the state of Goiás was left to decide by itself what to do with the nuclear waste.

The citizens of Goiás didn't want the waste, including the radioactive bodies of the dead, stored in their state. At the first funeral a group of protesters threw stones at the lead casket. For a day local motorists blocked the road to the waste storage site, twenty kilometers from Goiânia, where about 350 cubic meters of GRI waste was eventually deposited. The government of Goiás indicated that this was a temporary site for the waste.

People in other states did not accept food and clothing products from Goiás for several weeks after GRI, due to contamination fears. It was reported that the clothing and other industries in Goiânia suffered reduced sales of about 40% during the months of October and November. It was also reported that the governor of another rice-producing state urged Brazilians to avoid buying rice from the state of Goiás. Travelers from Goiânia were occasionally refused rooms in hotels in other states of Brazil. Government officials tried to use television to alleviate these fears. The governor of Goiás started spending his weekends near the waste deposit site to indicate that there was no reason to be afraid of contamination.

About one hundred miles from Goiânia is the lovely hot-springs town of Caldas Novas. Two months after GRI began, the occupancy rate at the many hotels there was reported to be 14% of normal. On a casual drive around the town, one author (LDR) noticed that the usual many large tour buses from Rio de Janeiro and São Paulo were not there.

Various persons accused the media of exaggerating the problem. We think that the communication problem was largely due to a lack of timely, adequate, organized and accurate information supplied by CNEN to the press. There were many contradictory spokesmen instead of one authoritative spokesman, so the press had to decide who was correct. Here is an example of the poor communication between the technicians and the public: When a spokesman was asked when the radiation would disappear, he said “never,” because of the asymptotic nature of radioactive decay. One newspaper then carried the headline “The contaminated area will not be habitable for over 100 years.” The information was correct but incomplete, and was not properly communicated to the public. A month after GRI began, the press began consciously indicating that life in Goiânia had returned to normal, which helped rapidly diminish public anxiety. However, people outside the state of Goiás remained fearful for several more months.

Fear of the unknown is a normal – and useful – characteristic of man. In an age when dangerous materials are routinely used, leaders have a strong obligation to assure that use of these materials follows well-tested procedures and that the populace is educated about dangers. Only then will ill-founded fears be conquered.

The Future for the Victims of GRI

Most of the victims of GRI are poor. In exception are the junk dealer and his wife; reports indicated that the wife worked very hard at making the business a success. Some of the victims who owned things have now lost all or part of them due to caesium chloride contamination. Irrational fear or anger will probably make it difficult for them to find housing and employment in the future, while their lives may be in danger of retributive action. However, the junk dealer has tried to capitalize on his instant fame; for example, he asked reporters for money in exchange for his picture and asked for (and got) a visit from a famous actress.

A group of professors in Brazil plans to study the victims for several years, but we know of no long-term funds that will give such a study the longevity it requires. There are large needs for more than study; there are needs for adequate medical care, housing and employment opportunities for the victims. When the excitement of GRI fades and Brazil's many other crises supplant it in the public eye, none of these necessities will be supplied. We suspect that even the proposed study will fade away. In fact, when this article was completed in February 1988, GRI appeared to have faded into the background of Brazil's other problems.

Brazil, or some international organization, should now apply for dedicated long-term international financial help to properly care for and study the victims of GRI. Their plight is the result of international trade in dangerous high technology materials promoted by developed countries. The knowledge gained from learning how to properly care for the victims and from a long-term study of the victims could be of great benefit to the entire world.

Medical X-Rays in Brazil

From the author's personal experiences of living in Brazil and from informal discussions with other residents and medical personnel involved with medical x-rays, we offer the following general conclusions about the use of medical x-rays in Brazil.

There is very little, if any, regular inspection of medical x-ray devices in Brazil. Although there are programs to train operators for these devices, it is probable that, in many cases, persons who have had no formal training operate the devices. There appears to be no federal or state organized procedure for monitoring the exposure of the operators. We believe that it's standard practice to give patients unnecessary x-rays in Brazil.


The Goiânia Radiation Incident was the most serious event involving a medical radiation source ever recorded. At the time of this writing, four victims have died and many more lives will be drastically shortened. It is likely that the future for the GRI victims will be very bleak unless some extraordinary action is taken.

The late arrival of the rainy season and the actions of M.F.1 may have saved many other people from becoming victims.

Brazil should promptly institute a tough licensing and inspection regimen for medical radiation sources, including x-rays. A new regulating organization with no ties to the nuclear industry or the military should be given strong power to regulate nuclear materials in Brazil.

A major effort should be made to locate all nuclear radiation sources in the world. The effects of GRI, due to negligent behavior, should put fear in the minds of all the world's leaders and remind them what chaos could be caused by one of these sources due to purposeful behavior.

Schematic diagram of the dispersal of caesium-137 in the accident in Goiânia

The persons most highly contaminated in the accident in Goiânia, listed by site of exposure and family membership. Estimates of cytogenic data of doses incurred (in grays) and information on those admitted to hospital and on the four fatalities are also given. (A thick black border indicates who was hospitalized; the other individuals were treated as out-patients.)

Plan showing the location of Goiânia. Rio de Janeiro is located 1348 km to the South East.

Cross-sectional diagram of the radiation head of a teletherapy machine similiar to the one from which the source assembly was removed in Goiânia, showing the rotating assembly for the source capsule.

Cross-sectional diagram of an international standard capsule. Such a capsule of radioactive caesium chloride was broken open in the accident in Goiânia. The source was compacted to a coherent mass and sealed within two stainless steel capsules.

A schematic of Goiânia showing the locations of the principal sites of contamination.

Contamination pathways for caesium.

Schematic of a teletheraphy machine similiar to the one from which the source assembly was removed in Goiânia