Three Mile Island: An Unresolved Paradox

Abstract

There is biological evidence of injury to human beings due to the 1979 accident at the Three Mile Island Unit 2 nuclear power station. The evidence that exists may not have been given a proper hearing because authority is largely anchored to the community of engineering and the physical sciences.

Aaron Datesman is the primary author of this chapter.

On May 7, a few weeks after the accident at Three-Mile Island, I was in Washington. I was there to refute some of that propaganda that Ralph Nader, Jane Fonda, and their kind are spewing to the news media in their attempt to frighten people away from nuclear power. I am 71 years old, and I was working 20 hours a day. The strain was too much. The next day, I suffered a heart attack. You might say that I was the only one whose health was affected by that reactor near Harrisburg. No, that would be wrong. It was not the reactor. It was Jane Fonda. Reactors are not dangerous.

− Dr. Edward Teller, father of the hydrogen bomb, writing in the Wall Street Journal (Teller, 1979)

At the time of the TMI accident, I was living ... approximately four miles northwest of TMI. Concerning my experience following the accident at TMI: On Thursday, March 29, 1979, I was working all day with my son in our garage. The garage doors were open. That night when I took a shower, my face, neck, and hands looked like I was at the seashore and got burned real bad. I felt nauseous. My eyes were red and burning. I felt like I was looking through water. Friday morning when I got out of bed, my lips and nose were blistered, and my throat and inside my chest felt like fire. It tasted like burning galvanized steel. My son had similar experiences. He was 22 years old at the time.

− Affidavit of a resident living near Three Mile Island (Aamodt & Aamodt, 1984).

Becoming involved in the TMI research most certainly changed my life and research options. While emotionally trying I would do it all over again for what I’ve learned about science, academe, the courts, and the difficult situations of people fighting to overcome a system that exploits rather than serves them.

− Prof. Steve Wing, University of North Carolina, Chapel Hill; personal communication with the author, 2016.

The March 28, 1979, meltdown at Three Mile Island (TMI) Unit 2 in Pennsylvania remains the largest nuclear power plant accident, as well the largest industrial disaster, to take place in the Western hemisphere. As such, the topic is necessarily central to any discussion of the role of nuclear power − past, present, and future − in American society. While memory of the accident may be fading, a deep paradox concerning this event remains unresolved. The nature and resolution of this paradox should strongly influence decisions about the future of nuclear power in this country and the world.

The first two quotations opening the chapter illustrate opposing sides of the paradox. The physicist Edward Teller, inventor of the thermonuclear bomb and inspiration for the movie character Dr. Strangelove, asserted that no one could have been harmed by the accident at Three Mile Island.¹ Whatever his expertise, Dr. Teller was not present in Pennsylvania at the end of March, 1979. The observations and opinion of someone who was present at that time and place − whose own health, in short, was an indicator of the severity of the accident − were starkly different. How should the collision between authority and experience be resolved? It is the author’s view that the paradox illuminates a profound and worrisome failure of authority. Expert opinion in this instance privileged physical measurement over biological outcomes, and thereby may have failed to recognize and correctly identify injuries (both short- and long-term) due to radiation exposure at Three Mile Island.

By the time an order recommending evacuation for the most vulnerable was issued, two days after the accident began, the damaged facility had already released on the order of 20 million Curies (20 MCi) of the radioactive noble gas xenon-133 (Xe-133) into the environment. As shown in Fig. 6.1, most of the released activity was carried by the wind in a low plume traveling to the northwest. While it is common to read statements to the effect that “very low doses” resulted from the accident (Hatch et al., 1990) the release contained activity equal to that of 20 metric tons (44,000 pounds) of radium, the most radioactive element that occurs naturally. A far smaller amount of iodine-131, around 14 Curies, was also estimated to have been deposited in the ten-mile area surrounding the TMI facility. Some of that radioiodine was discovered in milk collected from local dairy farms.

Fig. 6.1

Map of the area surrounding the Three Mile Island nuclear power plant in Middletown, PA. The colors indicate exposure to external gamma radiation due to radioactive xenon released from the auxiliary building vent stack (marked with the filled triangle) during the first thirty-nine hours of the accident. The most intense portion of the plume blew toward the northwest under conditions of steady wind late in the evening on Wednesday, March 28, 1979. This image is drawn from the author’s own work (Datesman, 2020)

While the release due to the accident should not be considered small, it is nevertheless correct that the absorbed doses to individuals were not (according to the conventional scientific understanding) alarming. The largest dose to any individual due to gamma radiation was estimated to be in the range from 0.7–2 milliSieverts (mSv), which is on the order of the annual dose due to background radiation.² Analysis revealed that the total exposure among the two million individuals living within 50 miles amounted to approximately 37 person-Sieverts. The dose would be expected to cause two additional cases of cancer mortality among the affected population. Quite aside from the ethical question centered upon two deaths, at this level of impact no adverse outcomes from the accident should have been observable by means of epidemiology.

Nevertheless, according to a recent review, several independent investigations reported that TMI accident emissions were associated with increased cancer incidence or mortality. The health endpoints³ included both lung cancer incidence and breast cancer mortality among women (Wilson et al., 2023). If the scientific prediction of no harm were correct, then the epidemiological findings demonstrating harm must somehow be wrong. On the other hand, if the epidemiological findings are correct, then how could the observed medical outcomes result from the low-level exposures that occurred?

There is a pat, but accurate, joke that begins “How do you find five different opinions about a controversial topic?” The answer is, “Ask three epidemiologists.” The TMI epidemiological results do not support an unambiguous interpretation, and therefore (like the joke) generate more heat than light. Fortunately, other areas of scientific investigation relating to the accident have the potential to be more illuminating. It is interesting to briefly examine this history.

Dozens of lives have been spent in activism because of the Three Mile Island accident. In some cases, these stories have been captured in records stored in the archives of the library of Dickinson College. The Dickinson archives regrettably do not include the papers of two of the most prolific activists involved with the Three Mile Island issue, Norman and Marjorie Aamodt. An electronic search in the library of the Nuclear Regulatory Commission using “Aamodt” as a prompt, however, returns a request to refine the search terms to limit the results to fewer than one thousand items.

The Aamodts, who were a married couple, possessed interesting backgrounds. Although in 1979 they lived on a cattle farm in Chester County, PA, they were both technically educated. In fact, they met while they were both employed at Bell Laboratories, which at the time was undoubtedly the world’s premier scientific research facility. The Aamodts became involved in the TMI issue through a classified advertisement seeking professional expertise in determining the cause of the accident. A lifetime of activism followed, including especially their service to the plaintiff’s counsel in the litigation in the United States District Court for the Middle District of Pennsylvania that became known as TMI Consolidated. The plaintiffs in the case were more than 2000 persons who believed they had suffered harm due to the meltdown at Three Mile Island.

Speaking at the 1984 Workshop on Three Mile Island Dosimetry, sponsored by the Three Mile Island Public Health Fund, Marjorie Aamodt made the following statement:

....I’m one of the women who did the study on the cancer deaths in the area northwest of the plant. And I would simply like to say that it is not just a matter of how many deaths, but of how much we can learn from the deaths. These people, I believe, were the true dosimeters at the time of the accident. (Beyea, 1985a).

The health survey initiated by Marjorie Aamodt reported cancer clusters in discrete locations northwest of TMI. Between the venue in which Marjorie Aamodt spoke (concerned with dosimetry, that is, a physical measure of exposure) and her statement (“people ... were the true dosimeters”), one finds a clear statement of the central paradox. How do we resolve a conflict between physical measurement and biological outcomes?

The Aamodt survey was an important motivation for a larger, more rigorous epidemiological investigation (including 130,000 persons out to a distance of ten miles) funded by the TMI Public Health Fund. The investigation was undertaken by researchers from Columbia University, under the supervision of the well-regarded epidemiologist Mervyn Susser. Prof. Susser possessed a notable background. He and his wife, Zena Stein, were anti-apartheid activists in their native South Africa, which they left in 1956 due to their political beliefs. Susser went on to become chairman of the division of epidemiology at Columbia in 1966, where he was one of the first epidemiologists to examine the AIDS epidemic when it emerged in New York City during the early 1980’s. His profile is not consistent with that of a man willing to act as a toady for a harmful industry.

The results of the Columbia study are often portrayed as providing no evidence for health effects arising from the accident at TMI. This interpretation is not quite correct. In fact, the Columbia researchers found a discernible increase in the incidence of lung cancer among the affected population. Controversy in this matter arises not from the finding itself, but rather, from its interpretation. The Columbia investigators asserted that radioactive emissions from Three Mile Island could not have been causative (Hatch et al., 1990), in part because of the low dose of radiation received.⁴ They did not assert that no excess incidence of cancer had been found.

The incorrect impression of the Columbia results has taken root, especially because the plaintiff’s attorneys in TMI Consolidated engaged their own expert to re-evaluate the evidence. The expert was Steven Wing, an epidemiologist from the University of North Carolina (UNC) at Chapel Hill and author of the third quotation opening this chapter. Prof. Wing had previously analyzed mortality data among occupationally exposed workers at Oak Ridge National Laboratory. His experience at Oak Ridge left him with a sense of deep skepticism regarding the government/industrial/scientific nexus surrounding nuclear weapons and nuclear energy. Due to his earlier encounter, Wing was initially hesitant to become involved with the TMI litigation. He changed his mind based on the quality of the Aamodt cancer cluster survey, as well as the commitment and reasonableness demonstrated by these two remarkable individuals.

Working with the data collected by the Columbia researchers, Wing and the UNC team reached, for the most part, similar results. Susser wrote precisely this in a letter published in the journal Environmental Health Perspectives following the publication of the UNC investigation:

Our results and those of Wing et al. differ in no important respect. Our conclusions do differ: we saw no convincing evidence that cancer incidence was a consequence of the nuclear accident; they claim there is such evidence. (Susser, 1997)

It is the author’s opinion that this statement is a valid summary of the Columbia-UNC controversy. Unlike the Columbia researchers, the UNC team was willing to assert that TMI accident emissions could have caused the increased incidence of lung cancer. Their willingness to make this controversial claim seems mostly to have been due to three factors. The UNC team a) made no assumption that the doses were “low level”, b) analyzed the data with more granularity, and therefore were more strongly convinced by the dose response for lung cancer, and c) gave more weight to anecdotal evidence of radiation exposure.

In short, two well-credentialed teams of epidemiologists using the same data reached similar results, but nevertheless interpreted those results in diametrically opposite ways. The reader might be excused for concluding that epidemiology is not a rigorous scientific discipline. However, in the author’s opinion, there is a more subtle lesson. The conclusions to be drawn from epidemiology can only be as robust as the underlying understanding of the physical and biological mechanisms connecting exposure to harm.

The insight Marjorie Aamodt expressed at the workshop in 1984 is therefore valuable. If epidemiological results cannot be interpreted reliably due to a possibly deficient physical understanding, a biological dosimeter − that is, a yardstick by which to directly measure the impact of exposure on a living organism − might shed light instead. Such a yardstick does exist; in fact, its nature and use had already been described by two scientists from Oak Ridge National Laboratory in 1962. The relevant scientific field is known as cytogenetics, that is, having to do with the structure and function of human chromosomes. Chromosome “aberrations” due to DNA misrepair are indicators of the severity of exposure to ionizing radiation.

Marjorie Aamodt was not alone in her assessment. The Advisory Panel on Health Research Studies established by the Pennsylvania Department of Health (DOH) in 1979 initially called for a program in cytogenetic dosimetry as one of several recommended investigations. Citing “uncertainty surrounding causes of DNA strand breakage,” however, the panel later reconsidered its position (Wilson et al., 2023). It does not appear that the cytogenetic investigation recommended by the PA DOH was ever conducted.

Nevertheless, a human cytogenetic (that is, biological) investigation was eventually performed. In 1994–95, a Russian scientist named Vladimir Shevchenko twice visited Central Pennsylvania while engaged as an expert witness for the plaintiffs in TMI Consolidated. His participation was prompted by Norman Aamodt, whom he had met at a scientific conference in Geneva, Switzerland in 1994. Dr. Shevchenko had trained in ecology in the Soviet Union. His specific area of expertise involved assessing damage to forest ecosystems due to radioactive contamination. His work had taken him⁵ to the sites of multiple radiological disasters across the Soviet Union, including western Siberia near the Semipalatinsk test site, the area near the Mayak site where plutonium was manufactured, and to Chernobyl in Ukraine, where a reactor lacking a containment structure exploded in 1986. Norman Aamodt had the privilege of spending many days shuttling Dr. Shevchenko around the area around Harrisburg, PA, looking at trees.

Cores were taken from more than eighty trees in Pennsylvania, which were sent back to Russia for analysis. It was Shevchenko’s professional scientific opinion, conveyed in an official report filed with the District Court for the Middle District of Pennsylvania, that damage to trees indicated exposures in the range from 2000–10,000 mSv in locations northwest and west of TMI where the plume of xenon had been most intense. Shevchenko was not alone in his assessment. The American scientist Dr. James Gunckel, who was a world authority on modifications of plant growth and development induced by ionizing radiation, had expressed a concordant opinion about a decade earlier. Dr. Gunckel had examined deformed plants such as those shown in Fig. 6.2. He wrote the following in 1984:

Fig. 6.2

An example of a mutated plant observed by local resident Mary Osborne. The image here was included in the records of the 1985 Workshop on Three Mile Island Dosimetry. From (Beyea, 1985b)

I have carefully examined a few specimens of common plants collected shortly after the accident at TMI and compared them with specimens collected more recently. The current abnormalities are probably carried forward by induced chromosome aberrations ... it would have been possible for the types of plant abnormalities observed to have been induced by radioactive fallout on March 29, 1979. (Aamodt & Aamodt, 1984)

The results with plants are significant because they contradict the conventional explanation offered by the authorities. A psychological explanation for some of the adverse health impacts experienced by human beings, as the authorities assert, is not implausible. Both the accident, and the subsequent evacuation, were very traumatic events in the lives of the individuals affected. The psychological explanation, however, is incompatible with the observation of injury to trees.

Dr. Shevchenko did not confine his on-the-ground investigation to plant life. He also interviewed people living nearby. In the same locations where the structure, growth, and health of trees indicated exposure to ionizing radiation, he wrote, “the residents in these areas felt at the time of the accident unusual events about their health,” including.

...skin redness and rashes, nausea, inflammation of the eyes, metallic taste, inflammation of respiratory ways, diarrhea, anal bleeding, hair loss, interruption of the menstrual cycle, pain in the joints, and others. (Shevchenko, 1995)

The symptoms in humans, Dr. Shevchenko wrote, were consistent with radiation sickness resulting from an exposure in the range of 1000 mSv. The assessed doses to trees are higher than the doses to human beings both because the xenon plume was elevated, and also because the living portion of a tree (the bark and leaves) is external, and therefore unshielded.

Dr. Shevchenko additionally coordinated a sizable investigation that drew upon experts from disparate fields (including botany and ecology, physical dosimetry, immunology, and cytogenetics) from within the Russian scientific establishment. The final report of the investigation he oversaw is troubling, and almost completely unknown. This was the finding as expressed by Shevchenko:

In the cytogenetic report the data on the level of dicentrics⁶ in residents living around TMI is compared to the results of cytogenetic investigation of populations exposed to irradiation .... approximately the same frequencies of dicentrics were found out in residents of the areas around TMI and the residents of a number of regions in Russia most suffered (sic) from the action of ionizing radiation. (Shevchenko, 1995)

Summarizing the findings concisely, this is what Dr. Shevchenko found: the level of biological damage among the persons examined in Central Pennsylvania was comparable to that discovered among members of the Altai population in Western Siberia who were severely exposed to fallout from an atomic bomb. The finding is consistent with observations of damaged and deformed trees and plants, anecdotal information gained from interviews, and evidence of immunological deficiencies (among other insights), but at the same time seemingly inconsistent with the observation that the TMI exposures were about the same as the annual dose due to background radiation.

The comparison returns the discussion to its central theme, the contradiction between physical measurement and biological outcomes. In the author’s opinion, it is likely that the following two observations are both simultaneously correct:

• 1. The absorbed dose to any individual around TMI was small (less than about 2 mGy). The conclusion is anchored to physical measurements, most significantly those made using electronic devices known as thermoluminescent dosimeters (TLDs).

• 2. The biological impact to the most-exposed individuals was severe (in the range of 600–1000 mSv). The conclusion is supported by biological outcomes: anecdotal evidence consistent with radiation exposure, the outcomes of the investigations performed by Russian scientists, and (although not conclusive) epidemiology.

It is of course legitimate to view the results of the investigations coordinated by Shevchenko − which took place in the context of a legal proceeding, and which were never published in the peer-reviewed literature − with a reasonable degree of skepticism. Might they be mistaken, incorrectly interpreted, or manipulated in some way? The most sensible response to the concern is to replicate the investigation. Because some chromosome aberrations are stable over time − that is, they exist for the entire lifetime of the exposed person − this possibility remains.

Recognizing the possibility, the authors of this book (along with other collaborators) have undertaken just such an investigation. A karyogram from our 3MILER RUN (Three Mile Island Low level Exposure to Radioxenon: a Re-assessment Using New cytogenomics) investigation is shown in Fig. 6.3. The results of our preliminary investigation should be available by late 2024.

Fig. 6.3

A karyogram (an image of the complete set of chromosomes from one person) from the 3MILER RUN investigation. The study participant to whom this karyogram belongs lived within ten miles of TMI in 1979. Because the investigation remains blinded as this book goes to press, the authors are not aware whether this individual was exposed to radioactive xenon. The total number of chromosome aberrations (marked with circles in the image) appearing in hundreds of such karyograms belonging to one individual may provide indication of biological harm due to radiation exposure. The technique is similar to that used for the 1994 cytogenetic analysis, although technologically more refined

The paradox embedded in the story of the Three Mile Island accident, it has been argued, represents the dichotomy between physical measurement and biological outcomes. Proponents of nuclear power technology are, for the most part, anchored in the community of engineering and the physical sciences. Because the culture is technophilic, and often quite callous regarding questions of risk, and because the weight of power and wealth are on the side of government and industry, the side of the dichotomy anchored to the physical understanding has dominated official perspectives. Overall, the power imbalance acts to exclude the observations of affected persons regarding their own health, in favor of a mere physical theory.

Regarding the future of nuclear power, the lesson of the Three Mile Island accident in the present day ought to be that questions of harm from low-level exposure are not yet settled. If the degree of harm resulting from “low-level” exposures is in fact not negligible, then additional engineering controls will be necessary to mitigate these previously unrecognized or discounted risks. The necessary changes would increase the already uncompetitive costs of nuclear power technology.

Summary Points

• 1. Injury caused by exposure to ionizing radiation may be assessed using methods anchored either in physical or biological understanding. There is significant biological evidence of severe harm due to the low level TMI exposure.

• 2. A cytogenetic investigation named 3MILER RUN, conducted by the authors, is underway at the time of publication.

• 3. If biological indications that the TMI exposures were harmful were in fact correct, the additional engineering controls necessary to construct new nuclear power stations to an acceptable level of safety will almost certainly be prohibitively expensive.