Skip to main content
Nuclear physics

Nuclear physics

Lessons from Three Mile Island

10 Aug 2004

Three Mile Island: A Nuclear Crisis in Historical Perspective
J Samuel Walker
2004 University of California Press 315pp £16.95/$24.95hb

The emergency began at 4.00 a.m. on 28 March 1979. By 6.30 a.m. a dramatic increase in radiation levels indicated that the uranium core had been severely damaged, although this went unrecognized at the time. Some 30 minutes later, much of the core had melted and had begun to flow to the bottom of the steel vessel that holds the reactor core and its cooling water. By about 9.00 a.m. hydrogen that had been generated in the core collected in the reactor’s concrete and steel “containment” – and caught fire.

The emergency at the Three Mile Island nuclear reactor in Pennsylvania was caused by the reactor’s operators shutting down its emergency systems. Although they later regained control, the operators did so without understanding what was going on. Analysis, conducted many years later, indicated that the vessel could have failed at about 9.00 a.m. At this time only one barrier – the containment – stopped millions of curies of radioactive material from being released into the environment. While the chance of a containment failing is small – roughly 1% – we now know there are many ways it could fail, none of which can be predicted.

So from about 7.00 a.m. on that fateful day 25 years ago there was a small, but non-trivial chance of a major radioactive release occurring without warning. Fortunately, that did not happen and only about 15 curies of dangerous iodide-131 were emitted. In contrast, the 1986 Chernobyl disaster, which was also caused by operators closing off the emergency systems led to about 40 million curies of the isotope being released.

Although Chernobyl was a far bigger catastrophe, many of the efforts to prepare for an emergency at a nuclear power plant have been driven by the experiences gained during the response to the Three Mile Island crisis. This new book by J Samuel Walker, who is historian with the US Nuclear Regulatory Commission (NRC), provides a fascinating description of how the state of Pennsylvania and the NRC decided how best to protect the public and address their concerns. (In fact, they fell down on both counts – not only failing to take decisions but also communicating badly with the public.) However, the book does not, as the author claims, provide “the first comprehensive scholarly account of the Three Mile accident” because it does not cover all the facets of the emergency or how it was resolved.

The most effective way of protecting the public from the release of a large amount of radiation is to evacuate the area near the plant before the release takes place. If you wait until high levels of radiation are measured in the environment, you are guaranteeing that the public will be exposed to doses that could otherwise have been avoided. In the case of Three Mile Island, the operators declared a maximum (general) level of emergency at about 7.00 a.m. However, this did not lead to any action to protect the public; meetings on what to do were held instead.

These discussions dragged on for days, during which time the only advice was for pregnant women and pre-school children to leave the surrounding area. No attempts were made to administer stable (non-radioactive) potassium iodide, which can protect the thyroid to some extent from inhalation of radioactive iodine if it is taken before, or shortly after, a release of radiation. Although the federal government and private industry provided potassium iodide to local officials a few days later, it was placed in a warehouse where it could not be promptly distributed if needed.

But why did the NRC or the state of Pennsylvania not act? Today it is standard practice to have pre-determined criteria in place, based on plant conditions, that define whether a particular incident should be classified as an emergency, and, if so, what action should be taken. At the time of the Three Mile incident, however, no such criteria for taking protective action existed. The NRC did develop evacuation criteria, but only several days after the start of the accident.

State officials therefore relied on the NRC for advice. The NRC, however, did not have a defined role for itself during an emergency and was not prepared. Communication between the NRC, state officials and the site was very poor – normal phone lines were overloaded – and the NRC was not prepared to perform real-time assessments of accident conditions. In addition, the NRC was not used to making decisions promptly – it normally spent months or years reaching a conclusion. Indeed, early on in the emergency, the NRC was asked for advice on evacuation to which it never actually provided an official answer.

The emergency saw a continuous stream of public statements from “official” sources that were conflicting and confusing. The chief offender was the NRC. This facet of the response is described well in the book as a fascinating tale of how not to communicate with the public. NRC officials were making public statements from at least three different locations – the Three Mile Island site itself, the NRC headquarters in Bethesda, Maryland, and the Pennsylvanian state capital of Harrisburg. These statements were often speculative, uninformed, based on poor or incorrect information, unintelligible and conflicting.

The confused public not surprisingly became angry and ultimately lost confidence in public officials. Matters were resolved only when President Jimmy Carter designated Harold Denton – a senior manager within the NRC – as the sole source of information. The need to co-ordinate, hopefully from a single location, all official statements is now a fundamental concept of emergency preparedness. Fortunately, Denton turned out to be a good communicator, who was highly adept at dealing with the press and public.

Although this book will provide invaluable preparation for anyone who has to formulate an official response to an emergency, it does have several flaws. In particular, Walker does not say why neither the operators nor the NRC recognized that the core of the reactor had melted, despite numerous indisputable indications. Another omission is how the NRC overlooked the real threat of Three Mile Island – namely that tonnes of melted core could form an uncoolable mass that could destroy the reactor vessel.

Walker also does not discuss in detail how misleading information, poor procedures and training, control-room crowding and questions from the NRC all contributed to the operators’ confusion. All of these problems resulted from the belief that severe emergencies – such as a core melt – were so unlikely that they did not need to be considered when preparing for emergencies.

Overall, this is an interesting book that gives a fascinating description of the decision-making process during radiological emergencies and how one should communicate with the public. However, major problems with both issues continue to be seen to this day. Indeed, poor communication can create psychological problems that are in many ways more severe than the physiological problems that arise from accidental exposure to radiation. This book should therefore interest anyone who finds themselves at the centre of a technological emergency – scientists, journalists and public officials alike.

Related events

Copyright © 2024 by IOP Publishing Ltd and individual contributors