Three Mile Island

What It Covers

On March 28, 1979, a cascade of mechanical failures, operator errors, and regulatory shortcomings at Pennsylvania's Three Mile Island Unit 2 reactor nearly caused a full-scale nuclear meltdown. The episode traces the accident's timeline, the hydrogen bubble crisis, the contested health impacts, and how the event permanently reshaped U.S. nuclear energy policy.

Key Questions Answered

• •Cascading failure design: The TMI-2 accident exemplifies sociologist Charles Perrow's "normal accident" theory — complex systems with tightly coupled components will inevitably produce catastrophic failures regardless of operator competence. Understanding this framework helps engineers and regulators design systems where single-point failures cannot trigger irreversible chains of consequences across multiple interdependent subsystems.
• •Indicator vs. confirmation problem: A critical design flaw at TMI-2 was that the relief valve indicator light confirmed the valve received a close command, not that it actually closed. Engineers should distinguish between command-confirmation and state-confirmation signals in safety-critical systems — a lesson directly applicable to any control interface where assumed states drive high-stakes decisions.
• •Operator training gap: NRC post-accident investigations found operators were trained to run normal plant operations but received no meaningful emergency scenario training. The Kemeny Commission concluded the NRC's own lax training requirements — not just operator error — caused the delayed response. Regulatory bodies must mandate scenario-based emergency drills, not just procedural competency certifications.
• •Institutional optimism bias: Nuclear industry regulators, operators, and companies shared a collective overconfidence in reactor safety that caused them to dismiss early warning signs, falsify leak test results for weeks, and minimize the accident publicly in real time. Organizations managing high-risk systems should institutionalize adversarial review processes specifically designed to counter optimism bias in safety assessments.
• •Radiation exposure context: Multiple independent studies — EPA, NRC, Union of Concerned Scientists — found peak radiation exposure during the 1979 accident and subsequent cleanup reached approximately 0.98 rem on-site, roughly one-tenth of a standard chest X-ray. While disputed by some community members reporting symptoms, no study established statistically significant causal links between TMI-2 radiation release and measurable population health outcomes.