The Goiânia accident,orphan sources,trust,and risk perceptions in the wider nuclear industry

Introduction

Orphan sources are dangerous radiological sources that have been lost, abandoned, or are no longer properly tracked. ¹ There have been a significant number of radioactive sources used globally, and many have become orphaned. The widespread nature of commercial radiation sources, used mainly in medicine and industrial settings, increases the probability that sources could become orphaned. Since 1995, there has been an annual average of fifty sealed radioactive sources that have been found by members of the public in the United States alone. ² Deaths are also not an uncommon feature of orphan source accidents. In 1982, for example, five people in Azerbaijan died after an individual found and carried a caesium-137 source in their pocket. ³ In 1984, in Morocco, eight people died after a worker took home a lost iridium-192 source. ⁴ The 1987 Goiânia orphan source accident, involving a caesium-137 source, was perhaps the most impactful orphan source incident that has ever occurred.

Goiânia occurred in Brazil just one year after Chernobyl, the accident that reminded the world of the potential consequences of severe radiological and nuclear accidents. ⁵ The scale of this orphan source accident was unlike any that had occurred prior. 112,000 people were monitored for exposure and 249 were found to be contaminated as a result of the accident. ⁶ Despite this, its long-term impact on the international public’s consciousness has evidently been negligible. Nuclear power accidents, however, evidently have a much stronger impact on the public’s perceptions of risk. A quick analysis of how subsequent orphan source accidents and nuclear accidents have been written about and what incidents they were compared to demonstrates the lack of a long-term impact of the Goiânia accident on the international public’s consciousness. Following the recent orphan source incident in Australia in which a Caesium-137 source was lost in the Western Australian outback, neither Cosmos Magazine, the BBC, nor The Guardian made any reference to Goiânia in their reporting, despite the incidents involving the same radioactive isotope. ⁷ The Guardian makes particular note of Dr. Edward Obbard’s description of orphan source accidents being “relatively common” and that the incident was being engaged with in a “positive” manner by the public with “an undercurrent of irony and mockery.” ⁸ Obbard, writing for the University of New South Wales, made one reference to “an infamous incident in Brazil” and provided a two-sentence overview of the accident with no reference to the name of the city the accident itself is named after. ⁹

Recent incidents involving nuclear reactors, however, are frequently compared to Chernobyl in particular. Reporting on the 2019 Nyonoksa radiation accident, the BBC compared the level of secrecy surrounding it to Chernobyl. ¹⁰ Defense News referred to concerns of a “small-scale Chernobyl incident,” and in an interview for PBS, Jeffrey Lewis, the director of a non-proliferation program, described the nuclear-powered missile involved in the accident as a “tiny flying Chernobyl.” ¹¹ When the Fukushima accident occurred in 2011, there were also many comparisons to Chernobyl in the media. ¹² An ABC report wrote that Hiroshima, Nagasaki, Three Mile Island, the Bikini and Muruoa atolls, and Chernobyl “each hold a sway in the public consciousness.” ¹³ Nuclear power and nuclear weaponry incidents are mentioned, but none made reference to orphan source accidents as events that “hold sway in the public consciousness.” Chernobyl is compared to consistently as a devastating and impactful event, which it certainly was, but so was the Goiânia accident. Yet, if and when the Goiânia accident is mentioned in international publications, it is not referred to with the same importance as nuclear accidents like the Chernobyl accident.

As demonstrated by the Goiânia accident, orphan source accidents offer the same threat of radiological injury as nuclear reactor accidents. While orphan sources will not explode or melt down like a nuclear reactor, the lingering danger of both kinds of accident is the threat of radiation exposure. It is the radiation risk that still makes Chernobyl and Fukushima not safe for long-term human habitation, and it was the radiation that caused harm in the Goiânia accident. Radioactive sources are used both in medical settings, which is where the orphan source that caused the Goiânia accident originated, and throughout industrial processes as gauges, measuring devices, and small power sources, and in the sterilization of food products. ¹⁴ The widespread and portable nature of many of these sources means that they pose additional risks that nuclear reactors do not. The frequency at which orphan sources are created in comparison to nuclear power incidents certainly demonstrates these risks. However, again there is a clear focus on nuclear power accidents over orphan sources accidents and comparatively few studies focusing solely on the Goiânia accident.

Like Chernobyl and Three Mile Island, which are much more closely studied, the Goiânia accident brought the issue of orphan sources under the wider nuclear industry’s focus, and yet evidently not the international public’s focus. The International Atomic Energy Agency (IAEA) made a number of regulation changes as a result of the Goiânia accident. ¹⁵ The former IAEA Director of Radiation, Transport and Waste Safety, Eliana Amaral, would go on to say that prior to the Goiânia accident, “regulations were weak” regarding commercial radiation sources and “there was no awareness that sources must be controlled from ‘cradle to grave.’” ¹⁶ The Goiânia accident certainly alerted the wider international nuclear industry to the dangers of orphan sources. There is a clear risk presented by commercial radiation sources becoming orphaned, and yet the global focus remains on nuclear power. Nuclear power accidents such as Chernobyl and Three Mile Island certainly left a lasting impact on global nuclear regulations, but so did Goiânia. Here, how the risk of orphan sources as a result of the Goiânia accident has been perceived will be analyzed. Safety changes made as a result of Goiânia will be taken into account as a reduction of the actual risk of an orphan source accident occurring, as opposed to the perceived level of risk. Utilizing surveys from both the United States and Brazil, whether or not Goiânia had an impact on the public’s perceptions of the nuclear industry’s risk will be analyzed.

The United States is included in this analysis due to its global influence on the wider nuclear industry. While the history of commercial radiation did not begin in the United States, it quickly became an international leader in the industry during and after the Second World War. With the Manhattan Project underway, American scientists sought to understand how to sustain a nuclear chain reaction and therefore apply that understanding to a bomb. This process of understanding was facilitated by the world’s first nuclear reactor, Chicago Pile-1, tucked secretly under the stands of a squash court at the University of Chicago. ¹⁷ It should be noted, however, that the British were the first to commercialize nuclear power with Calder Hall, which ran from 1956 to 2003. ¹⁸ Despite this, the United States would go on to influence the foundation of the International Atomic Energy Agency, which monitors military and commercial nuclear projects in all United Nations affiliated countries. In 1953, U.S. President Eisenhower addressed the United Nations and made his “Atoms for Peace” speech, in which he proposed the world work together on the development of peaceful uses for the atom, particularly through nuclear energy. In 1956, eighty-one nations approved the IAEA’s statute, which was shaped by Eisenhower’s speech, and in 1957 the IAEA was officially founded. ¹⁹

In addition to having a history tied directly to that of the nuclear industry’s development, the United States also has an abundance of survey material regarding public opinion relating to the nuclear industry. The number of surveys that have been produced on this matter provide a much clearer view of trends over time specifically in the United States, thus making their results ideal for the analysis of the impact of specific incidents. Being relatively close to Brazil means that there is also a higher chance that the accident will have an impact on survey data. The proximity of the United States in comparison to Europe, for example, which also produces a large output of surveys focused on public opinions of the nuclear industry, means that the United States is an ideal nation to contrast with Brazil. Surveys from Brazil are, of course, also included as the nation directly affected by the orphan source incident utilized here as a case study.

Brazil along with Argentina and Mexico, were and still are the leading nuclear nations in Latin America. ²⁰ In 1958, Argentina was the first nation to operate a research reactor in the Southern Hemisphere, and they developed the first nuclear power reactor in the region, commissioned in 1974. ²¹ The ARCAL agreement was also in place at the time of the Goiânia accident, with the goal of fostering cooperation between Latin American nations on the topic of peaceful nuclear applications. ²² Between its introduction in 1984 and 1987, participating countries included Argentina, Bolivia, Brazil, Chile, Columbia, Ecuador, Paraguay, Peru, Uruguay, Venezuela, Guatemala, Costa Rica, Cuba, and Panama. ²³ So why not analyze their data? The foremost reason is they lack the quantity of data that the United States has readily available. Most surveys are recent, with one 2011 survey conducted in twenty-four countries, including Latin American countries, with each nation contributing at least 500 participants. This survey identified Mexico as one of the most anti-nuclear nations, with about 80 percent opposition to specifically nuclear power. ²⁴ This result, however, cannot be connected in any way to the Goiânia incident. In addition to there being a very large gap in time between this survey and the accident, Mexico is known to have issues with source control in general and Mexican officials have a reputation for denying or minimizing hazards, leading to strong lack of trust in those in power. ²⁵ This relationship between trust and perceptions of risk is very relevant to the study of nuclear incidents, but this is only an example of this correlation and has no major bearing on the study of the Goiânia accident.

Another study on Latin American nuclear development and Latin American perceptions of nuclear energy was published very recently, in 2024. Alejandro Loreto Alfaro, Florencia Renería, and Carmila Araujo took a sample of an unspecified size from the Latin American and Caribbean region and found that 70.4 percent of participants were strongly in support of nuclear energy and 18.5 percent were in support but had some reservations. ²⁶ However, while this data reveals the current state of opinions in Latin America, it does little to reveal what impact Goiânia may have had. Indeed, Latin America has only just started a push to develop nuclear power in the region outside of Argentina, Brazil, and Mexico as recently as 2025, even though the majority of Latin American countries are participants in the ARCAL agreement. ²⁷ These surveys present current views regarding the nuclear industry in Latin America, but the lack of additional surveys between recent years and the accident means that there is no trend to analyze and therefore connect to the accident and subsequent events.

Academically, the international conference held in Goiânia 10 years after the accident reveals that Latin American nations likely were not majorly impacted by the Goiânia accident. The reason behind this, however, cannot effectively be analyzed without more data on risk perceptions. Cuba, Peru, San Salvador, and Argentina were the only other Latin American nations outside of Brazil that contributed research to the international conference, and out of all of them, only Argentina put forward an article on the impacts the Goiânia accident had on their safety legislation. ²⁸ Even so, the author writes that the accident simply sped up the improvement of some of their regulations. ²⁹ Therefore, while Brazil is surrounded by other nuclear Latin American nations, the United States still provides much more comprehensive surveys and sources on the matter of the Goiânia accident and the nuclear industry as a whole.

Lastly, why look at this accident at all? One might suspect that because there is little research on the topic, it does not have the same international significance that a nuclear power accident might have or that it may be more a medical accident and thus irrelevant to the wider nuclear industry. The Goiânia accident can be viewed through the lens of a medical accident as the orphan source that caused it was from a medical context, but it is primarily an accident involving a radiological source and is therefore a nuclear accident. Indeed, the intersection of accidents being both radiological and medical have resulted in significant changes, especially in the early days of medical radiation. For example, in the 1930s, the Radium Girls were victims of poor workplace safety and a general lack of knowledge regarding the radium they were working with. ³⁰ Their suffering demonstrated that external and internal radiation exposure were distinctly different from one another in their effects. ³¹ This revelation was then reinforced by the death of Eben Byers in 1932, who consumed large quantities of the radioactive patent medicine Radithor. ³² Their combined tragedies resulted in radiopharmaceutical regulations and the U.S. National Bureau of Standards’ 1936 Handbook 20, which held the first government recommendation for occupational exposure at 0.1 roentgen per day. ³³ This recommendation was held through World War Two. ³⁴ One might view these accidents as either industrial or medical accidents, but they are also without a doubt radiological accidents that resulted in impactful change. Another set of examples are the Therac-25 linear accelerator accidents that occurred in the 1980s. These accidents were the result of software failures and a lack of physical safety checks, and caused multiple deaths with high radiation exposure. ³⁵ This accident is now commonly presented as a case study for computer science ethics classes. ³⁶ Again, these medical accidents, which undoubtedly resulted in lasting changes, also involved radiological sources, thus also making them nuclear accidents.

In 1967, Latin American nations, including Brazil, signed the Treaty for the Prohibition of Nuclear Weapons in Latin America and the Caribbean, and in doing so dedicated their nations’ nuclear development to peaceful applications. ³⁷ But, as pointed out by Bernadette Bensaude-Vincent, Soraya Boudia, and Kyoko Sato, there still remains a “destructive potential” for all nuclear technology. ³⁸ They go on to say that “disasters like Three Mile Island, Chernobyl and Fukushima reveal [that] nuclear violence is not bound to formal hostilities.” ³⁹ This direct comparison of nuclear technology with that of nuclear weapons under the singular term of “nuclear violence” reveals a significance to nuclear accidents outside of military applications. In an article soon after the accident, Gail Daneker and Jennifer Scarlott insist that in calling the Goiânia accident a “‘radiation’ accident” instead of a “‘nuclear’ accident,” officials sought to “decouple the accident from nuclear power in the public mind.” ⁴⁰ They go on to say that usually “officials go to great lengths to link the two—arguing that the safe handling of medical radiation shows that we can safely handle all nuclear technology.” ⁴¹ In saying this, Daneker and Scarlott are connecting medical radiation to the wider nuclear industry directly.

Combining the processes of thought from Bensaude-Vincent, Boudia, Sata, Daneker, and Scarlott, we can conclude that if the significance of nuclear technology accidents such as nuclear power accidents hold the same international importance as nuclear weaponry in terms of their potential for “nuclear violence,” then medical radiation incidents, such as orphaned medical radiation sources, also fall into this same level of importance. Indeed, after the Goiânia accident, in 1988, the United States looked into whether an accident like Goiânia’s could occur there and how they should respond to it and mitigate perceptions of risk. ⁴² Again, this reveals that there was international interest and concern regarding the Goiânia accident and the potential for major orphan source accidents. While in the Cold War context, nuclear weapons and nuclear power would have seemed much more significant than a medical radiation accident to global interests, the Goiânia accident was a nuclear accident and as such is no less significant internationally and is more than worthy of further in-depth study.

The Goiânia accident

Near the end of 1985, a private radiotherapy institute in Goiânia, Brazil, relocated to a new location, leaving behind a caesium-137 teletherapy unit. ⁴³ When the institute moved, it was meant to take all sources with it, or at least have the sources stored securely, and notify the appropriate authorities if the sources were left unsecured. ⁴⁴ In 1987, the caesium-137 source was eventually found in the abandoned institute by two members of the public and sold to a junkyard owner as scrap. The junkyard owner noticed that there were rice-sized grains of glowing material inside the source’s capsule and invited friends and family to look at it and take some. ⁴⁵ The junkyard owner’s brother took some home to his family. His six-year-old daughter played with it and rubbed it on her body so that she glittered before, with her hands covered in caesium-137, eating food. ⁴⁶ Within 10 minutes she was vomiting. ⁴⁷ Other family members, some putting it on themselves like it was “carnival glitter,” also began to become ill within 5 days following first contact with the material. The junkyard owner’s wife, recognizing it was likely causing their illnesses, took the source in a bag to the city public health department, the Vigilancia Sanitaria, telling the official she met that it was “killing her family.” ⁴⁸ It was placed on the desk of a department doctor who, out of concern, moved it to a chair in the department’s courtyard. A medical physicist was contacted and, upon arrival, they promptly borrowed a scintillation monitor. ⁴⁹ When switched on, the monitor immediately read the highest measurement on its scale. Assuming it was broken, the physicist got a replacement, only to return and read the same measurement. While the physicist was investigating, the department doctor became concerned enough that he called the fire department. The physicist arrived at the department just in time to stop them from throwing the source in a river. ⁵⁰

Previous incidents involving orphan sources failed to attract international attention as they were seen as domestic issues, but Goiânia garnered international interest. ⁵¹ Clean-up progressed quickly and was very well managed. ⁵² Information was spread about the incident and those concerned were instructed to go to Goiânia’s Olympic stadium for monitoring. 112,000 people were monitored, and 249 were found to be contaminated. Eight people were diagnosed with acute radiation syndrome and four died. ⁵³ The accident also sparked discussion on radiation source labeling, as the initial scrappers either did not see the radiation “trefoil” symbol or did not understand what it meant. The International Atomic Energy Agency’s report on the incident noted that the blue glow emitted by the caesium-137 source, in conjunction with human curiosity, “significantly affected the course of the accident,” and the trefoil clearly did not deter the scrappers. ⁵⁴ Attention was also given to the matter of source licensing and tracking as a result of the lack of proper management surrounding the orphan source, which contributed to the accident. ⁵⁵

Changes to safety that were introduced as a result of this accident focused on source regulation and labeling. With the IAEA’s involvement with the accident and its overall scale, it was “clear [to the IAEA] that a comprehensive international approach, rather than isolated national actions, [was] needed to cope with the problem of orphan sources.” ⁵⁶ The first problem local authorities and the International Atomic Energy Agency had to address was how the source became orphaned in the first place. The licensee of the caesium-137 source had violated the requirements of the license that allowed them to have it by abandoning the facility and not taking the source with them. ⁵⁷ Locally, Brazil introduced a system that required licensees to routinely submit reports on their controlled sources. ⁵⁸ The IAEA official accident report recommended that facilities that have a licensed radioactive source should be regularly inspected. ⁵⁹ This recommendation was put into place for UN member states by the IAEA in 1999 following a modernizing overhaul of safety systems for all member states. ⁶⁰ Indeed, after the accident, a few countries began inspecting facilities. According to the IAEA, “in one [search], eleven teletherapy units were found in old irradiation rooms partially demolished and in hospital courtyards.” ⁶¹ The International Basic Safety Standards No. 115 was introduced and was particularly instrumental in the widespread adoption of further orphan source related safety measures. ⁶² Brazil itself was instrumental in prompting this whole process of increased safety standards, requesting that the IAEA learn from Goiânia. ⁶³ The IAEA also developed a more easily understandable radiation danger warning symbol in response to Goiânia. ⁶⁴ Since the Goiânia incident and the introduction of these new IAEA regulations in the decades after it, there has been no orphan source incident similar to the scale of Goiânia.

The United States’ response

There are no surveys regarding U.S. public opinion specifically for this event. There are, however, a plethora of surveys that deal with nuclear power rather than orphan sources or fuel cycle facilities. These can be utilized to observe any large change in opinion that may be connected to Goiânia as part of the nuclear industry as a whole. It is, however, difficult to judge with any certainty how Americans felt specifically about radioactive sources before and after Goiânia. In using surveys that focus on nuclear power, it needs to be acknowledged that Chernobyl occurred the year prior to Goiânia. Therefore, opinions will be heavily influenced by Chernobyl and may overshadow any effect Goiânia had. That being said, Chernobyl may have also resulted in a larger reaction to Goiânia since it was connected to the nuclear industry. Hence, despite these difficulties, U.S. opinions before and after the Goiânia accident will be analyzed utilizing surveys that focus only on the nuclear power sector of the nuclear industry.

Prior to Goiânia, U.S. public opinion was very clearly negatively affected by the 1979 Three Mile Island accident. In a 1982 survey, 35 percent of respondents replied that they had “realis[ed] the dangers of nuclear power” where, prior to the accident, only 10 percent replied that they “fear[ed] a nuclear accident.” ⁶⁵ Additionally, in 1984, a comparison of five post-Three Mile Island accident surveys revealed that there was a growing tendency for host communities to oppose the construction of new nuclear facilities. ⁶⁶ So, prior to both Chernobyl and Goiânia, the American public were already experiencing a shift in their risk perceptions of the nuclear industry, and specifically nuclear power, in that they were more aware of the potential for an accident and therefore considered nuclear facilities less desirable.

After the Second World War, the term “Atomic Age” held optimism and advancement, but throughout the 1970s the meaning began to shift, reaching a tipping point with the Three Mile Island accident. ⁶⁷ Three Mile Island certainly was not the first nuclear accident in the United States, after all their SL-1 reactor exploded in 1961 and killed three men. However, SL-1 was a small test reactor located 65 km away from the nearest large population center. ⁶⁸ Three Mile Island was a commercial reactor located in a Pennsylvanian township only 15 km away from the Pennsylvanian capital, Harrisburg. ⁶⁹ In addition to this proximity to a large population, the poor communication from officials as the Three Mile Island accident progressed only served to sow confusion and panic and fuel a growing lack of trust in nuclear power and its safety, which is reflected in surveys. ⁷⁰

The poor way in which Three Mile Island was handled by nuclear authorities damaged trust in the nuclear industry. But trust in the nuclear industry had already begun to deteriorate prior to Three Mile Island. In particular, the Watergate scandal in 1972 greatly affected the general public’s appreciation of the government and associated agencies. Increasing environmental awareness and damage to the U.S. government’s image resulted in the Atomic Energy Commission, which regulated the U.S. nuclear industry, becoming increasingly scrutinized even before Three Mile Island. ⁷¹ Both before and after Three Mile Island, the trust that the general public in the United States had in their nuclear industry and associated authorities played a more important role in shaping perceptions of risk than the actual risk of a nuclear accident.

In 1988, after both Chernobyl and Goiânia, an international collection of surveys, which included surveys from the United States, noted that as a result of the Chernobyl accident “there was an increase in the percentages considering nuclear power plants not very safe and even dangerous.” ⁷² De Boer and Catsburg go on to say that “a year after the Chernobyl accident the survey results [regarding perceptions of risk were] again very similar to the very stable answering patterns which were found in the previous years.” ⁷³ So, survey answers quickly returned to pre-Chernobyl patterns. It is very telling that, again, there is no mention of Goiânia, and they only mention the impact of Chernobyl. Similarly, another survey found that opposition to nuclear power reached 49 percent and opposition to having a nuclear power plant in one’s neighborhood had increased to 70 percent. But, like de Boer and Catsburg, Ortwin Ren also found that opinions had returned to pre-Chernobyl levels by 1989. ⁷⁴ People also became more certain about their opposition or support after Chernobyl. Although having no opinion in regard to nuclear power was decreasing after Three Mile Island, Gallup polls reveal a noticeable drop from 7 percent to 4 percent of respondents having no opinion between 1979 and 1986, in addition to a 60 percent to 79 percent jump in opposition. But these Gallup polls also revealed that Three Mile Island had a greater effect on U.S. citizens’ support of the construction of nuclear power plants. Chernobyl did, however, have an impact similar to Three Mile Island on opinions regarding nuclear power plant construction being cut back until greater restrictions were put in place. Opposition to construction jumped from 44 percent to 66 percent after Three Mile Island but fell to 55 percent in 1980, only to return to 66 percent after Chernobyl. ⁷⁵

Finally, the media storm surrounding Chernobyl likely strengthened American focus on the accident and the risks associated with nuclear power. Soviet secrecy meant that international media, including U.S. media, had no clear idea what was occurring, so at time simply fabricated numbers fueled by their “antagonism toward the Soviet Union.” ⁷⁶ One Los Angeles Times article specifically points out a New York Post headline that read “Late Word From Inside Russia: Mass Grave for 15,000 N-Victims,” which we, of course, know now to be completely false. ⁷⁷ This unrelenting focus on Chernobyl, driven by post-Three Mile island concerns and the media frenzy surrounding Chernobyl, even after the Goiânia accident had occurred, reveals that neither the Goiânia accident nor the resulting safety changes had any evident impact on U.S. perceptions of risk. Despite the return to pre-Chernobyl “answering patterns,” the U.S. public’s concern evidently remained more focused on nuclear power rather than orphan sources. Again, this is reinforced by the fact that there were no surveys focused on the opinions of the U.S. public in relation to orphan sources.

U.S. news media at the time presented the accident in Goiânia as an event of little importance to America. Marlise Simons, reporting for The New York Times, for example, described the fear of radiation in Goiânia 2 months after the accident as mainly affecting Brazil. The only mention of the relevance of the accident to the United States was that the issues that had “come up [in Goiânia],” according to Simons, “tended to elude American disaster models.” ⁷⁸ In American journals, the incident was simply recapped. An American Association for the Advancement of Science news article, written by Leslie Roberts, focused specifically on the effect on the Brazilian population and how Brazilian authorities handled the situation. Indeed, Roberts’ attention turned back to Chernobyl rather than Goiânia. She made particular note that U.S. doctor Robert Gale, who assisted in treating Chernobyl victims, was also involved in the treatment of Goiânia victims. ⁷⁹ Both public and academic news media avoided acknowledgment of any relevance the Goiânia accident had to U.S. orphan source incidents and safety.

Brazil’s response

Brazil, unlike the United States, was evidently concerned with the impacts of the Goiânia accident in that they did conduct surveys in response. Two surveys published in 1997 collected Brazilian opinions in direct relation to the Goiânia accident and the changes that were introduced afterwards. High school students from schools in Sao Paulo were surveyed regarding their knowledge of nuclear topics. Twenty percent had no knowledge about nuclear energy, and only 3 percent had heard of the concept of radioactivity. ⁸⁰ This lack of knowledge or awareness of the topic of radioactivity certainly reveals how easily an individual from Brazil might ignore the radiation warning symbol on nuclear sources. Again, this scenario contributed to the beginning of the Goiânia accident. Interestingly, 93 percent of those surveyed believed the proprietors of the clinic that caused the Goiânia accident were responsible for the accident, while 80 percent thought that radiation could have beneficial uses provided proper safety measures were followed, and 75 percent thought the use of radioactivity should not be abandoned. ⁸¹ This is remarkably different to the public in the United States who, after Three Mile Island, for example, put the blame on the nuclear industry, and their trust in the safety of the nuclear industry declined. Freudenberg and Baxter revealed in their 1982 survey that 34 percent of respondents believed that the power companies should assume the 26–50 percent of the burden of financial loss as a result of Three Mile Island and, additionally, respondents indicated that the federal government should take on the burden of another 26–50 percent. A fifth of respondents believed that the state government should be responsible for 25 percent of the burden. ⁸² This allocation of blame, reflecting the U.S. public’s opinions on who should assume financial responsibility, further aids in revealing the loss of trust that occurred after Three Mile Island. The Brazilian public, however, continued to trust their nuclear industry after the Goiânia accident, whereas the U.S. public lost their trust after Three Mile Island.

This preservation of trust may have been due to the levels of communication involved during the Goiânia accident in comparison to the Three Mile Island accident. U.S. officials communicating what was occurring during the Three Mile Island accident failed to provide a consistent story. The accident was first reported to the public, not by any government or regulatory official, but by a popular music radio station. ⁸³ Communication continued to filter through unofficial channels presenting conflicting information. ⁸⁴ The press secretary to a U.S. senator called the plant operator just hours after the accident and was told that they were “in the process of formulating a statement on the condition of the reactor.” In reality, there were no public information representatives on site. ⁸⁵ While a general emergency had been declared at the reactor, the area within the twenty-mile radius of the reactor was only notified of the emergency several hours after Harrisburg was told. The WHP, the hub of the Pennsylvanian emergency broadcast system, was never formally told of any emergency. ⁸⁶ Members of the State Bureau of Radiological Health said that they received a phone call from the supervisor of Unit 2 of the power station advising an evacuation, but then they received another phone call from a calm, unidentified individual who presented details of the power station’s radiation release and did not recommend evacuation. ⁸⁷ The governor, having just received two opposing recommendations, simply advised people within a ten-mile radius of the plant to remain indoors and keep doors and windows shut. Then, later that same day, pregnant women and preschool children were advised to evacuate from the five-mile radius around the plant. ⁸⁸ After the accident, the President’s Commission, formed in response to Three Mile Island, found that the “handling of information during the first 3 days of the accident resulted in loss of [the power plant’s energy company’s] credibility as an information source with state and local officials, as well as with the news media.” ⁸⁹

While the communication of the Three Mile Island accident was ineffective and contradictory, the Goiânia accident had comparatively excellent communication surrounding it. The Brazilian authorities notified the IAEA very soon after the accident was discovered, utilizing the 1986 Vienna Convention on Early Notification of a Nuclear Accident and the Convention on Assistance in the Case of a Nuclear Accident or Radiological Emergency. ⁹⁰ The act of quickly contacting the IAEA allowed them to secure international help from U.S. experts, and equipment from France, West Germany, Hungary, Israel, the Netherlands, and the United Kingdom. ⁹¹ Additionally, instead of communicating complex information directly to the public, coordinators for the accident communicated directly with journalists and they explained the basics of radiation’s applications and effects. ⁹² A pamphlet on “what you should know about radioactivity and radiation” was made, and 250,000 copies were distributed. In addition, a 24-hour telephone service was set up for any concerns or inquiries regarding the accident. ⁹³ With this effective communication, it would seem trust was preserved in Brazil after the Goiânia accident, while after Three Mile Island, trust in the United States was evidently damaged.

This does not mean, however, that Brazilian perceptions of risk were entirely unaffected by the accident, but as fears began to spread about radiation, Brazilian regulatory authorities quickly realized they needed to both inform the Brazilian public of any real risk from radiation sources and also investigate radiation source inventories for danger. ⁹⁴ Shortly after the accident, an inspection program was put into place to investigate all radioactive installations. ⁹⁵ In the decade after the accident, strict control of all nuclear installations and radiation sources in Brazil was practiced. ⁹⁶ In another 1997 public survey, 80 percent of respondents considered applications of radiation indispensable and 81 percent had heard about radioactivity. ⁹⁷ However, it should be noted that, despite their greater general awareness of radioactivity, 35 percent still did not know what the radiation warning symbol meant. ⁹⁸ This issue would not be rectified until the introduction of the new radiation warning symbol in 2007. ⁹⁹

Despite their continued trust in their nuclear industry, the Goiânia accident had left a definite impression on the minds of the Brazilian public. Ninety-two percent of respondents in one of the 1997 surveys remembered the Goiânia accident. The surveyors, Wieland, Steinhauser, Xavier, and Unterbruner, noted that the population “had vivid memories of the occurrences and in several cases, expressed a high degree of emotion.” ¹⁰⁰ Recent reporting in the Brazilian media reveals the long-term impact the accident left on the Brazilian population. As of 2025, a Netflix series dramatizing the Goiânia accident is in production. ¹⁰¹ However, it has generated some controversy in that it is being shot in Sao Paulo rather than Goiânia. The city’s Municipal Culture Council published an open letter that stated that having the series shot in Goiânia instead would “do justice to [their] history” and that the accident “belongs to Goiânia, to its people, to its scars.” ¹⁰² An Estadoa article describes the “marks” left behind by the accident, such as the junkyard where the capsule was opened now being a hole “dug eight metres deep and concreted,” remaining a “great void since no work can be done there for safety reasons.” ¹⁰³ Another Estadoa article mentions that 1,141 people are still monitored by the Assistance Centre for Radiation Accident Victims. ¹⁰⁴ In 2007, sixty people gathered in front of the Municipal Theatre in Sao Paulo to mark the twenty-year anniversary of the Goiânia accident. The people dressed in black and lay on the ground in the presence of a person dressed as Death to represent the “risks of nuclear energy.” ¹⁰⁵ The article goes on to describe the problems survivors endure, such as “cancer, genetic defects, psychological sequelae, and prejudice.” ¹⁰⁶ At this event, Greenpeace and SOS Mata Atlantica called for anti-nuclear energy sentiment, as at the time the federal government announced that the Brazilian nuclear program would resume. ¹⁰⁷ Even with the impact the Goiânia accident had on the Brazilian populace, orphan sources are evidently not the primary concern for Brazilian anti-nuclear activists. Similarly in academia, in Luiz Pinguelli Rosa’s article, “History of Nuclear Power in Brazil,” Goiânia is referred to in passing, the article focusing instead on the waste storage repository resulting from the event. ¹⁰⁸

Academically, there has been consistent, but seemingly not widespread, Brazilian interest in the Goiânia accident throughout the years since the accident took place, which also serves to demonstrate the imprint the accident left in the minds of the Brazilian population. In 1990, 3 years after the accident, J.J. Rozenthal, C.E. de Almeida, and A.H. Mendonca, from the Brazilian National Nuclear Energy Commission (CNEN), wrote an invited paper for the IAEA on initial steps the CNEN took to minimize the population’s exposure to radiation. This article further highlighted the excellent communication employed by Brazilian officials, noting the importance of “a well understood chain of information and command,” and describing that workers were encouraged to explain what they were doing to locals and “thus gained the people’s confidence and raised the credibility of official statements.” ¹⁰⁹ In the same year, L.A. Vinhas of the Institute of Energy and Nuclear Research in Brazil published an article with the IAEA on how sites were decontaminated, highlighting effective protocols for the workers’ personal protection. ¹¹⁰

In 1997, at the international conference in Goiânia focused on the accident, out of the fifty articles presented, half were from Brazil. These articles focused on a range of topics, including follow-up studies on the impacts, the legislative lessons learned, proposals for further protection for radio diagnostics patients, proposals for waste repository sites, and studies into soil contamination. ¹¹¹ In 2001, A.S. Paschoa, J.J. Rozenthal, and A. Tranjan Filho detailed the decision-making behind constructing the waste repository for the waste left behind by the Goiânia accident. They again highlighted the importance of “credibility” in lowering the public’s perceptions of risk. ¹¹² In 2002, Telma Camargo Da Silva published an article focusing on the memories of survivors and how these memories form a “politics of remembrance,” which she believes is important for illuminating “social suffering that had importance health implications” for survivors. ¹¹³

Academic interest remains niche but continues into recent years. In 2018, L.C.J. Silva and F.B. Razuck focused on the importance of learning from the Goiânia accident, noting that “it is important to highlight the radiological accident as a tool of Knowledge Management in the field of nuclear science.” ¹¹⁴ As a final example, also in 2018, Elizabeth M. Pontedeiro, Paulo F. Heilbron, Jesus Perez-Guerrero, Jian Su, and Martinus Th. Van Genuchten published an article reassessing the water infiltration through the soil of the Goiânia waste repository site using the HYDRUS-1D software, noting that the chance of the caesium-137 waste’s movement to groundwater is unlikely over the time period of 600 years. ¹¹⁵ From these examples, it is clear that Brazilian academic interest in the Goiânia accident has remained consistent since 1987. It would seem that the impact of the accident remains strong in the minds of Brazilian academics as well as the general public.

Despite this strong imprint in the Brazilian public’s consciousness, it would seem that they still considered the risk of the radiation sources minimal enough that the majority felt that the continued use of radiation sources was “indispensable or very important/important.” ¹¹⁶ The Goiânia accident was significant to the Brazilian public, but their perceptions of risk remained positive after the accident. It is evident the Brazilian public’s positive risk perceptions came about because of the Brazilian government’s effective communication skills and their subsequent increases to safety as a result of the accident.

The United States’ history of orphan source accidents

In the United States, the public was unaffected by both the accident and the resulting safety changes. The first reason for the lack of a reaction from the U.S. public is that they had not had an orphan source accident on the same scale as Goiânia. The United States had plenty of orphan source incidents occur, but none nearly as severe as Goiânia had been. In 1999, for example, it was reported that the United States Nuclear Regulatory Commission received 200 reports of lost radioactive sources and devices. Additionally, these were only the instances when it was noticed that a source was lost and a report was made. ¹¹⁷ There were likely many more orphan sources simply not reported to the U.S. Nuclear Regulatory Commission. Orphan sources such as these did cause minor accidents. ¹¹⁸ In 1979, for example, an iridium-192 industrial source was left at a job site in California by accident. ¹¹⁹ One worker, unaware of what it was, placed it in his pocket and received a high dose of radiation to the tissue underneath where it sat. ¹²⁰ In the medical sector, as an additional example, a patient receiving an internal radiation therapy in 1992 had a radioactive source left in their body. ¹²¹ The patient died from acute radiation exposure and ninety members of the public were exposed to the source. ¹²² Accidents such as these, while damaging, tragic, and sometimes fatal, were on a much smaller scale in comparison to Goiânia, where 249 people were exposed to the source and four died, and so they evidently did not influence the public on any major scale. ¹²³

Many sources end up in metal scrap yards as well. Since 1983, radiation detectors have been installed at metal mills and scrap processing centers to investigate all the metal that comes in. ¹²⁴ Between 1983 and 2002 in the United States, over 500 radioactive sources were reported to have been found in scrap. ¹²⁵ One of the worst orphan source accidents to occur in the United States actually began in 1984 in a Mexican scrapyard. Scrapyard workers in Juarez opened up an unmarked capsule and spilled 6,010 pellets of cobalt-60. ¹²⁶ Scrap metal and cobalt-60 were spread over the scrapyard, resulting in a number of exported metal products becoming contaminated. ¹²⁷ Six hundred tons of this reused contaminated steel was then shipped to the United States. ¹²⁸ Eventually, a truck containing some of the contaminated metal accidentally passed through the entrance to the Meson Physics Facility at the Los Alamos National Laboratory in the United States. Once there, the load triggered a radiation detector. ¹²⁹ It was estimated that approximately 4,000 people across Mexico and the United States were exposed to radiation as a result of this incident. ¹³⁰ So, the United States’ most notable orphan source incident was in fact a Mexican accident that was spread over the border into New Mexico. The United States only became involved because the accident was discovered once contaminated materials had entered the country. Thus, the United States itself has had no orphan source incident close to the scale of Goiânia’s and the worst incident that did occur began in Mexico and was simply discovered in the United States.

While Brazil had reason to be concerned but trusted their regulatory bodies as a result of their handling of the accident, the United States had no reason for concern. This also indicates that the Nuclear Regulatory Commission in the United States was trusted with monitoring sources. There was little to no indication of public outcry or protest in regard to orphan sources in the United States. Unlike Three Mile Island, the public had not been given any reason to be alarmed by the prospect of an orphan source. They simply did not perceive any great risk as a result of potential orphan source accidents. The lack of any major orphan source accident in the United States was one reason why the Goiânia accident in Brazil and subsequent safety changes did not have an impact on the U.S. public’s perceptions of risk.

Radiation warning symbols

The recognizability of radiation warning symbols was another concern raised after the Goiânia accident, which resulted in the introduction of a new symbol for dangerous sources. The victims of Goiânia were unaware of what the radioactive “trefoil” symbol meant and so these warnings were useless in altering the course of the events. ¹³¹ It was recommended, in the international conference marking the ten-year anniversary of the Goiânia accident, that a different symbol be used that should be more generally understandable. The example given at the time indicated it should have a “skull of death” and perhaps the words “danger [to] life” to signal the danger of the source. ¹³² It was not until 2007, after much study on how to best visually convey danger, that a new symbol was introduced and applied to sealed sources that could cause death or serious injuries. ¹³³ The IAEA surveyed eleven different countries and across India, Brazil, and Kenya only 6 per cent recognized what the original trefoil meant. The new symbol did indeed include a “skull [to] death,” which was situated under the original trefoil symbol with rays emitting down from it. Next to the skull was a person running away, all on a bright red background. ¹³⁴

The lack of impact on the perceptions of risk in the United States implies that the American public were not concerned with recognizing radiation warnings. With the concept of radiation being incorporated into much of the U.S. media during the Cold War and the prevalence of the fallout shelter symbol, it would be very unlikely for a U.S. citizen to not recognize the radiation warning symbol. ¹³⁵ Therefore, by 1987 and the Goiânia accident, the American public would not have the same concern over not recognizing the radiation symbol as had occurred in Goiânia.

The United States had a strong involvement in the history of these warning symbols and they were particularly recognizable to the public as a result of Cold War nuclear tensions. Indeed, the radiation warning symbol originated in the United States, with its invention and first use in the University of California’s Berkeley Radiation Lab in 1946. ¹³⁶ By the late 1950s, the American National Standards Institute formally codified the symbol with the option of coloring the symbol’s trefoil either with the original magenta color it was designed with or black, which is now more common and internationally used. ¹³⁷ While this shows the United States’ strong connection to the international radiation warning symbol, U.S. citizens were unlikely to come across this symbol in their everyday life, but they were more likely to see fallout shelter signage. In 1969, with the Cold War looming over the United States, Congress voted to spend more than 169 million dollars on nuclear fallout protective structures for the civilian population, and over a million fallout shelter signs were produced. ¹³⁸ These became a common sight to U.S. citizens during the Cold War and were extremely similar in design to the standard radiation warning symbol. ¹³⁹ Instead of three black or magenta blades on a yellow background, the fallout shelter sign consisted of three yellow triangles, encased in a black circle, on a yellow background, with the words “fallout shelter” beneath it. ¹⁴⁰ The fallout shelter symbol, which bore great resemblance to the radiation warning symbol, likely would have increased the recognizability of the radiation warning symbol to those in the United States who may not have encountered it before. Indeed, the Office of Civil Defense originally intended to use the warning symbol, but felt it was inappropriate as a fallout shelter represented safety and the warning symbol represented a hazard. ¹⁴¹ The relationship between the warning and fallout shelter signs was strong and therefore the recognizability of the warning symbol in the United States was likely strong too.

Conclusion

The Goiânia accident and resulting changes to safety regulations had little impact on the U.S. public’s perceptions of risk. The accident was ultimately irrelevant to the American public and therefore did not generate any concern. This is made evident by the lack of surveys on the matter. The American public were far more concerned with the Chernobyl accident that occurred a year prior and nuclear power accidents in general as a result of their experiences involving Three Mile Island. This lack of concern for events in Brazil was due in part to the fact that the United States had no major orphan source incidents. The public trusted their regulatory bodies to find and dispose of orphan sources. The fact that there had been no prior accidents on the same scale as Goiânia meant that the American public could trust that their Nuclear Regulatory Commission were competent in dealing with source distribution and orphan sources. In comparison, Brazil, which was directly affected by the accident, experienced an unchanged risk perception. As mentioned before, this was the result of effective communication techniques that fostered trust between the public and the government. The responses of both the Brazilian and U.S. publics further demonstrate the relationship between trust and risk perception. The United States’ public had no reason to distrust their regulatory bodies in their management of orphan sources, and so the accident had no effect on their perceptions of risk. Additionally, the United States’ history in the development of the radiation warning symbol and the widespread use of the very similar fallout shelter symbol meant it was unlikely that a U.S. citizen would not recognize the symbol. Therefore, unlike the Brazilian public, the U.S. public did not need to be concerned with this possibility.

The clear lack of concern about commercial radiation sources and the possibility of orphan sources reveals much about the nature of trust in perceptions of risk. The nuclear energy sector of the overall nuclear industry has a perception of risk shaped by damaged trust and the looming impact of the Three Mile Island accident’s disastrous communication. Officially, Three Mile Island had no real measurable health impact aside from stress-related illnesses, and yet it had a lasting impact on perceptions of risk regarding nuclear energy. ¹⁴² Goiânia, on the other hand, did have victims affected directly by exposure to radiation, but concerns related to orphan sources are nowhere near as prominent as nuclear energy. Communication related to this event was extremely effective at informing the public about the risk and what to do if they were concerned. Here we see the lasting effect of trust on risk perceptions in these two sectors of the nuclear industry. Whether trust in nuclear energy can be restored is a topic of great interest considering the growing need for alternative energy sources.