Case Study · Nuclear/Radiological · 1979
March 28, 1979. A stuck valve, misread instruments, and wrong operator responses combined into the worst nuclear accident in American history. The core partially melted. A hydrogen bubble threatened containment. 140,000 people left voluntarily. No one died. And the NRC rebuilt the entire regulatory system around what almost happened.
Three Mile Island Accident · March 28, 1979
Four in the morning. March 28, 1979. At the Three Mile Island Nuclear Generating Station near Middletown, Pennsylvania, the Unit 2 reactor was running at normal power. A mechanical failure — either electrical or mechanical, investigators later concluded — stopped the feedwater pumps that deliver water to the steam generators that cool the reactor core. Backup pumps activated automatically, as designed. But a valve that should have opened to allow backup cooling water in had been left closed during maintenance. The core began to heat.
Then a pressure relief valve opened — also as designed — but stuck in the open position after it should have closed. Cooling water began escaping from the system. An indicator light showed the valve had received the signal to close. The operators interpreted this as confirmation the valve was closed. It was open. They believed the reactor was over-pressurizing when it was actually losing coolant. They reduced cooling water flow — the wrong action, making the situation rapidly worse. Over the next several hours, the reactor core experienced a 45% melt. A hydrogen bubble formed in the reactor vessel, raising fears of an explosion that could breach containment. President Carter visited the plant on April 1 to demonstrate the situation was stabilizing. By April 9, the immediate crisis had passed.
March 28, 1979
Date
0
Deaths
~140K
Voluntary Evacuees
$973M
12-yr Cleanup
Nuclear
Type
The NRC's official determination is unambiguous: the approximately 2 million people in the surrounding area received an average radiation dose of about 1 millirem above background — less than a sixth of a standard chest X-ray, with no detectable health effects. More than a dozen independent studies over the following decades confirmed no adverse health outcomes attributable to the radiation. Governor Thornburgh had ordered the evacuation of pregnant women and children within 5 miles — a conservative and appropriate action given the hydrogen bubble uncertainty. Approximately 140,000 people left voluntarily. The cleanup took 12 years and cost $973 million. And the NRC rebuilt the entire framework for commercial nuclear safety — operator training, emergency planning zones, human factors engineering, regulatory oversight — around what almost happened on March 28, 1979.
The Science
Think of a nuclear reactor not as a bomb but as a very controlled heat source. The fission reaction produces heat; water carries that heat away to the steam generators; the steam drives turbines. The water doing this cooling is under pressure to raise its boiling point, allowing it to absorb more heat. The entire safety system is designed around one requirement: keep water flowing over the fuel rods at all times. If the cooling is interrupted for long enough, the fuel rods overheat. The zirconium alloy cladding on the fuel rods fails. The uranium fuel begins to melt. This is what happened at TMI. The cooling was interrupted — by stuck valves and misread instruments — and the core began to melt.
Despite a 45% core melt — more severe than initial estimates — TMI's containment building performed as designed. The containment is a massive reinforced concrete structure enclosing the reactor and primary cooling system, designed to prevent radioactive material from reaching the environment even in a severe accident. The World Nuclear Association account confirms: the containment held. The average radiation dose to surrounding residents — 1 millirem — is a direct measure of how effectively the containment structure performed under conditions more severe than it had ever been tested against.
One of the most alarming aspects of the TMI accident was a hydrogen bubble that formed in the reactor vessel as the damaged fuel reacted with cooling water. There was genuine scientific disagreement about whether the bubble could ignite — NRC engineer Harold Denton and plant officials publicly disagreed about its size and danger, contributing to public fear and confusion. The EBSCO account of the accident documents this communication failure in detail: the competing official statements, the uncertainty about the bubble's behavior, and the role of communication failures in amplifying public anxiety. The bubble was eventually resolved without incident, but its presence for several days — and the inability of officials to provide consistent information about it — shaped how the public experienced the event.
Timeline
01
March 28, 4:00 AM: Feedwater pump failure. Backup systems activate, but a maintenance valve is left closed. A relief valve opens correctly, then sticks open. Instruments show it closed. Operators believe the system is over-pressurizing and reduce cooling water. The core begins to heat. Within hours, partial meltdown is underway. Plant operators and NRC are alerted but initial assessment underestimates severity.
02
March 30: Gov. Thornburgh orders evacuation of pregnant women and children within 5 miles. A hydrogen bubble forms in the reactor vessel — its behavior and stability are uncertain. NRC official Harold Denton publicly disagrees with plant operator assessments. About 140,000 people leave voluntarily. Competing official statements amplify public fear beyond the actual radiation exposure, which is less than a chest X-ray for those nearby.
03
April 1: President Carter visits TMI control room — a deliberate demonstration that conditions are stabilizing. The hydrogen bubble is being reduced. April 9: The immediate crisis is declared over. The containment has held throughout. Unit 2 is shut down permanently. NRC launches its investigation. The average dose to 2 million nearby residents: 1 millirem above background — less than a chest X-ray.
04
1979–1993: Twelve-year, $973 million cleanup. 100 tonnes of damaged fuel removed. NRC implements sweeping reforms: operator certification, Emergency Planning Zones, human factors engineering review, INPO established, control room redesign standards. More than a dozen independent health studies find no adverse effects from radiation exposure. The industry is permanently changed by what almost happened.
Human Decisions
What worked
Despite a 45% core melt and a hydrogen bubble that threatened to complicate containment, the TMI-2 containment building did what it was designed to do. The DOE's official account confirms no deaths, injuries, or detectable health effects resulted from the radiation release. The containment is the safety system that makes nuclear power viable in populated areas — and TMI tested it under severe conditions and the design held.
Rather than ordering a full-scale general evacuation — which officials feared could cause more deaths from traffic accidents and stress than the radiation exposure — Governor Thornburgh ordered a targeted evacuation of the most vulnerable populations within 5 miles. The EBSCO account documents the deliberation behind this decision: officials weighed the actual radiation exposure against the risk of evacuation panic. The decision held: no deaths resulted from the radiation, and no mass evacuation panic occurred.
What the accident exposed
The NRC investigation found that operators had not been trained to recognize and respond to the specific failure sequence they encountered — a Loss of Coolant Accident (LOCA) masked by misleading instrument readings. The operators were trained on normal operational procedures and on common emergency scenarios. They were not trained on the instrumentation failure mode that made the open valve appear closed. The post-TMI operator certification requirement — mandatory training on the full physics of the reactor system — directly addressed this gap.
The public communication failures at TMI — NRC officials contradicting plant operators about the hydrogen bubble's size; initial uncertainty about whether to recommend evacuation; conflicting statements from the utility and the regulator — produced fear and voluntary evacuation behavior whose effects (stress, displacement, economic disruption) exceeded the measurable radiation impact. Communication failure during nuclear events can itself become a public health problem, independent of the radiation.
Before TMI, no formal Emergency Planning Zones existed for communities around nuclear plants. There were no pre-scripted evacuation routes, no potassium iodide distribution programs, and no standardized community notification procedures. The recognition that a community of 140,000 could voluntarily evacuate in response to an accident — and that this response needed infrastructure to prevent chaos — was a direct output of TMI's aftermath.
The compound effect
Three Mile Island is simultaneously the most severe nuclear accident in U.S. commercial power history and a demonstration that the containment design can work even when every other system fails. The accident killed no one and exposed the surrounding population to less radiation than a chest X-ray. It also revealed, in documented detail, that operator training was insufficient, that control room design was inadequate, that emergency planning for surrounding communities was nonexistent, and that the regulatory framework was reactive rather than proactive. The reforms that followed TMI — mandatory operator certification, Emergency Planning Zones, INPO standards, human factors engineering — made the U.S. nuclear industry measurably safer. The cascade lesson is not that nuclear power is inherently catastrophic. It is that the safety of complex technological systems depends on the quality of the human and institutional systems surrounding them — and that the near-miss is sometimes the most valuable event in the long-term safety record.
What Changed
Before TMI, no standardized emergency planning framework existed for communities near nuclear plants. After TMI, the NRC required all operating nuclear plants to establish Emergency Planning Zones — the 10-mile plume pathway EPZ and the 50-mile ingestion pathway EPZ — with pre-scripted community notification procedures, evacuation routes, potassium iodide distribution plans, and exercises conducted with state and local emergency management. Every community within 10 miles of a U.S. nuclear plant now has an emergency plan that did not exist before March 28, 1979.
The NRC implemented mandatory operator training and certification requirements that required operators to understand the full physics of the reactor system — not just operational procedures. The industry established the Institute of Nuclear Power Operations (INPO), a non-regulatory oversight body that conducts regular evaluations of all nuclear plants against high performance standards. Remaining in good standing with INPO is a condition of continued operation that creates industry-wide pressure toward safety culture. The World Nuclear Association confirms: more than a dozen independent health studies found no adverse effects. The safety culture reforms that followed TMI are a primary reason.
About 93 commercial nuclear reactors operate in the United States today. The safety framework surrounding them — Emergency Planning Zones, operator certification, INPO evaluations, human factors engineering standards, control room design requirements — traces directly to what was absent on March 28, 1979. The NRC's assessment is that the accident's aftermath "significantly enhanced U.S. reactor safety." For the communities that live within 10 miles of those 93 reactors, TMI's legacy is the infrastructure that now exists specifically to protect them if the worst happens again.
If It Happened Today
What You Can Do Now
About 93 commercial reactors operate across 28 U.S. states. If you live within 50 miles of one, some of these actions apply directly to you. The key lesson from TMI: the infrastructure now exists to protect you. Use it.
The NRC's Emergency Planning Zone system divides your risk into zones. Within 10 miles, you would receive direct guidance to evacuate or shelter in place in a General Emergency. Contact your county emergency management office to request the site-specific emergency plan, learn your evacuation route, and ask if potassium iodide is distributed to households in your zone. This is the infrastructure TMI created for you. Use it before you need it.
Find your local nuclear EPZThe 140,000 who self-evacuated from around TMI did so with no official guidance — many may have driven toward rather than away from the plume depending on wind direction. Today's EPZ system provides zone-specific evacuation or shelter-in-place guidance based on real-time plume modeling that the public can't access. Following zone guidance is safer than self-evacuation. If told to shelter in place when your instinct says leave, trust the system.
Nuclear emergency response guideTMI exposed 2 million people to 1 millirem above background — less than a chest X-ray. The fear and voluntary evacuation produced by communication failures during the crisis had greater documented health effects (stress, displacement, accidents) than the radiation. In any nuclear emergency, the relevant question is the official dose estimate, not whether radiation is detectable. Radiation is always detectable at some level. The question is always whether the dose is harmful.
Understanding radiation dosePotassium iodide (KI) protects only the thyroid gland, only against radioactive iodine (I-131), only when taken at the right time — before or shortly after exposure. It does not protect any other organ or against other radioactive materials. Take it only when directed by authorities. If your county distributes KI tablets (many within the 10-mile EPZ do), keep them accessible but don't take them preemptively — they can cause side effects, particularly for people with thyroid conditions.
Potassium iodide guideIf you live within 50 miles of a nuclear plant and a significant release occurs, state agriculture and health officials will issue guidance about locally produced food — particularly fresh milk, leafy vegetables, and garden produce. Radioactive iodine and cesium can deposit on crops and concentrate in dairy. Following ingestion guidance during the recovery period, not just the acute emergency phase, is the protection the 50-mile EPZ was designed for.
Nuclear emergency preparednessNext step
The nuclear emergency preparedness guide covers EPZ lookup, evacuation route planning, shelter-in-place procedures, KI use and timing, understanding radiation dose, and how to stay informed during a nuclear emergency.
Nuclear emergency preparedness guideSources