Three Mile Island Accident

Three Mile Island Accident

The Three Mile Island accident, a partial nuclear meltdown of the Unit 2 reactor at the Three Mile Island Nuclear Generating Station in Pennsylvania, stands as the most severe accident in the history of U.S. commercial nuclear power plant operations. Occurring at 4:00 a.m. EST on March 28, 1979, on Londonderry Township, Dauphin County, Pennsylvania, near Harrisburg, Pennsylvania, the incident resulted in the release of radioactive gases and iodine into the surrounding environment. Despite the severity of the event, official assessments indicate that the small radioactive releases had no discernible health effects on plant workers or the public. On the International Nuclear Event Scale (INES), the accident is classified as Level 5, signifying an "Accident with Wider Consequences."

Accident

The cascade of failures that led to the partial core meltdown in TMI-2 commenced in the non-nuclear secondary system of the reactor. This initial malfunction was exacerbated by a stuck-open pilot-operated relief valve in the primary system, which permitted a significant quantity of water to escape from the pressurized, isolated coolant loop. The mechanical failures were compounded by the operators' initial inability to recognize the situation as a loss-of-coolant accident (LOCA). The training and operational procedures in place at TMI left the personnel and management ill-equipped to handle the escalating crisis. Further complicating matters were design flaws, including poorly designed controls, an overwhelming number of similar alarms, and a critical lack of equipment to indicate the coolant inventory level or the status of the compromised relief valve.

The accident irrevocably heightened public apprehension regarding nuclear safety, precipitating new regulatory measures for the nuclear industry and contributing to a decline in the construction of new reactors. Activists within the anti-nuclear movement voiced grave concerns about potential long-term health consequences in the region. While some epidemiological studies suggested a statistically significant increase in cancer rates in the vicinity of the plant following the accident, other studies found no such correlation. Establishing a definitive causal link between the accident and cancer rates remains challenging due to the inherent complexities of such studies. The cleanup of TMI-2, initiated in August 1979, concluded in December 1993, at an estimated cost of approximately $1 billion (equivalent to$2 billion in 2024). Notably, the companion TMI-1 reactor was restarted in 1985 but was retired in 2019 due to operational losses. However, it is slated to return to service in 2027 or 2028 as part of an agreement with Microsoft to power its data centers.

Background

On the night of the incident, the TMI-2 reactor was operating at 97% power, while the adjacent TMI-1 reactor was offline for refueling. The sequence of events leading to the partial nuclear meltdown began in the early hours of March 28, 1979, at 4:00:36 a.m. EST, originating in TMI-2's secondary loop, one of the three primary water/steam circuits in a pressurized water reactor.

The initial trigger for the accident occurred approximately eleven hours prior, during an attempt to rectify a blockage in one of the eight condensate polishers. These sophisticated filters are crucial for purifying water in the secondary loop, preventing mineral buildup in steam generators and minimizing secondary-side corrosion. While blockages are not uncommon, the standard method of clearing them with compressed air proved ineffective. Operators then resorted to forcing compressed air into the water, intending for the water's momentum to dislodge the resin. This action, however, inadvertently forced a small amount of water past a stuck-open check valve and into an instrument air line. This infiltration ultimately led to the shutdown of the feedwater pumps, condensate booster pumps, and condensate pumps around 4:00 a.m., resulting in a turbine trip.

Reactor overheating and malfunction of relief valve

The cessation of feedwater flow to the steam generators severely reduced heat transfer from the reactor coolant system (RCS) [25], causing a rapid increase in RCS temperature. As the coolant heated and expanded, it surged into the pressurizer [26][27][28], compressing the steam bubble at its apex. When the RCS pressure reached 2,255 psi (155.5 bar), the pilot-operated relief valve (PORV) actuated, releasing steam through piping to the reactor coolant drain tank [29] located in the containment building basement. The RCS pressure continued to climb, reaching the reactor protection system's high-pressure trip setpoint of 2,355 psi (162.4 bar) just eight seconds after the turbine trip. The reactor automatically tripped, causing its control rods to descend into the core under gravity, thereby halting the nuclear chain reaction and the heat generated by fission. Nevertheless, the reactor continued to produce decay heat, initially equivalent to approximately 6% of its pre-trip power output. With the turbine no longer consuming steam and without feedwater supply to the steam generators, heat removal from the reactor's primary water loop was limited to the minor steam generation within the secondary side of the steam generators, which was then vented to the condenser via turbine bypass valves [30][31][32].

Upon the trip of the feedwater pumps, three emergency feedwater pumps automatically activated. An operator noted their operation but failed to observe that a block valve in each of the two emergency feedwater lines was closed, effectively preventing emergency feed flow to both steam generators. The valve position indicators for one valve were obscured by a yellow maintenance tag. The reason for the operator's oversight regarding the second valve remains unclear, though one theory suggests his physical stature may have obstructed his view. It is possible these valves had been left closed following a surveillance test conducted two days prior [34][35]. The closure of these valves constituted a violation of a critical Nuclear Regulatory Commission (NRC) rule, which mandates reactor shutdown if all auxiliary feed pumps are out of service for maintenance. This oversight was later highlighted by NRC officials as a significant failure [36].

Following the reactor trip, secondary system steam valves operated to decrease steam generator temperature and pressure, thereby cooling the RCS and causing the primary coolant to contract [Thermal expansion]. This contraction, coupled with the coolant loss through the open PORV, led to a drop in RCS pressure and, after a peak 15 seconds post-turbine trip, a decrease in pressurizer level. Concurrently, 15 seconds after the turbine trip, the coolant pressure had fallen to 2,205 psi (152.0 bar), which was the reset setpoint for the PORV. Electrical power to the PORV's solenoid was automatically cut, but the relief valve remained stuck open, allowing continuous coolant discharge [37].

Post-accident analyses identified the PORV's indication system as one of several design flaws within the operators' controls, instruments and alarms. Crucially, there was no direct indication of the valve's actual position. A control panel light, installed after the PORV had previously stuck open during startup testing [38], illuminated when the PORV opened [39]. When this light—labeled "Light on – RC-RV2 open" [40]—extinguished, operators erroneously concluded the valve had closed. In reality, the light merely indicated that power was supplied to the PORV pilot valve's solenoid, not the valve's physical state [41]. While the main relief valve remained stuck open, the operators, misinterpreting the extinguished light, failed to correctly diagnose the problem for several hours [42].

The operators had not received adequate training on the ambiguous nature of the PORV indicator or on alternative methods to confirm the main relief valve's closure. A downstream temperature indicator, positioned in the tailpipe between the PORV and the pressurizer relief tank, could have signaled a stuck valve had the operators paid attention to its elevated reading. However, this instrument was not considered part of the "safety grade" suite of indicators intended for use during emergencies, and personnel had not been trained in its utilization. Its location, obscured behind the seven-foot-high instrument panel, rendered it effectively invisible [43].

Depressurization of primary reactor cooling system

Less than a minute into the incident, the water level within the pressurizer began to rise despite a falling RCS pressure. With the PORV stuck open, coolant was being lost from the RCS, constituting a loss-of-coolant accident (LOCA). The expected symptoms of a LOCA are simultaneous drops in both RCS pressure and pressurizer level. The operators' training and plant procedures did not account for a scenario where these two parameters diverged. The rising pressurizer water level was a consequence of steam venting through the stuck-open PORV, which reduced pressure in the pressurizer due to the lost coolant inventory. This pressure reduction caused water from the coolant loop to surge into the pressurizer, creating a steam bubble in the reactor pressure vessel head, further fueled by the decay heat from the nuclear fuel [44].

This steam bubble was not visible to the operators, and this phenomenon had not been part of their training. Indications of high pressurizer water levels contributed to confusion, as operators were concerned about the primary loop "going solid" (i.e., the absence of a steam pocket buffer in the pressurizer), a condition they were explicitly trained to avoid at all costs. This confusion was a primary factor in the initial failure to recognize the event as a LOCA [45] and led to the operators shutting down the emergency core cooling pumps, which had automatically activated due to the stuck-open PORV and subsequent coolant loss, based on the erroneous fear of overfilling the system [46][47][48].

As the PORV continued to discharge coolant, the pressurizer relief tank, designed to collect this discharge, became overfilled. This overflow caused the containment building sump to fill, triggering an alarm at 4:11 a.m. This alarm, coupled with unusually high temperatures on the PORV discharge line and elevated containment building temperatures and pressures, were clear indicators of an ongoing LOCA. However, these signals were initially disregarded by the operators [49][50]. At 4:15 a.m., the relief diaphragm of the pressurizer relief tank ruptured, allowing radioactive coolant to leak into the general containment building. This contaminated coolant was then pumped from the containment building sump to an auxiliary building, situated outside the main containment, until the sump pumps were deactivated at 4:39 a.m. [50].

Partial meltdown and further release of radioactive substances

Around 5:20 a.m., after nearly 80 minutes during which a growing steam bubble had formed in the reactor pressure vessel head, the primary loop's four main reactor coolant pumps began to cavitate as they circulated a mixture of steam and water, rather than pure water. The pumps were consequently shut down, with the expectation that natural circulation would suffice to maintain water flow. However, the steam within the system impeded flow through the core, and as water circulation ceased, it increasingly converted to steam. Shortly after 6:00 a.m., the upper portion of the reactor core became exposed. The intense heat generated a reaction between the steam and the zircaloy cladding of the nuclear fuel rod, producing zirconium dioxide, hydrogen, and additional heat. This reaction compromised the cladding of the fuel rods and damaged the fuel pellets, leading to the release of radioactive isotopes into the reactor coolant and the generation of hydrogen gas, which is believed to have caused a minor explosion within the containment building later that afternoon [51].

A graphic illustrating the end-state configuration of the TMI-2 core:

• 2B inlet
• 1A inlet
• cavity
• loose core debris
• crust
• previously molten material
• lower plenum debris
• possible region depleted in uranium
• ablated incore instrument guide
• hole in baffle plate
• coating of previously molten material on bypass region interior surfaces
• upper grid damage

At 6:00 a.m., a shift change occurred in the control room. A newly arrived operator noticed excessive temperatures in the PORV tailpipe and holding tanks. He utilized a backup valve, known as a block valve, to halt the coolant discharge through the PORV. By this time, approximately 32,000 US gallons (120,000 L) of coolant had already leaked from the primary loop [52][53]. It wasn't until 6:45 a.m., a full 165 minutes after the initiation of the problem, that radiation alarms sounded as contaminated water reached detectors. At this point, radiation levels in the primary coolant water were approximately 300 times the expected levels, and the general containment building was heavily contaminated, with radiation levels reaching 800 Rem (unit)/h.

Emergency declaration and immediate aftermath

At 6:56 a.m., a plant supervisor declared a site area emergency. Less than 30 minutes later, station manager Gary Miller escalated this to a general emergency. Metropolitan Edison (Met Ed) subsequently notified the Pennsylvania Emergency Management Agency, which in turn alerted state and local authorities. Pennsylvania Governor Richard L. Thornburgh was informed, and he delegated responsibility for information gathering and dissemination regarding the accident to Lieutenant Governor William Scranton III [55]. The uncertainty prevailing among operators at the plant was mirrored in the fragmented, ambiguous, or contradictory statements issued by Met Ed to government agencies and the press, particularly concerning the potential for and severity of off-site radioactivity releases [56].

Scranton, at a press conference, attempted to be reassuring yet conveyed confusion about the possibility of off-site releases, stating that while there had been a "small release of radiation... no increase in normal radiation levels" had been detected. These assertions were contradicted by another official and by Met Ed's own statements, both of which claimed no radioactivity had been released. Readings from instruments both at the plant and from off-site detectors had indeed registered radioactivity releases, although at levels deemed unlikely to pose a public health threat if they were temporary and if containment of the highly contaminated reactor was maintained [58].

Angered by Met Ed's failure to inform them before venting steam from the plant and convinced that the company was minimizing the accident's severity, state officials turned to the NRC. Upon receiving word of the incident from Met Ed, the NRC activated its emergency response headquarters in Bethesda, Maryland, and dispatched personnel to Three Mile Island. NRC chairman Joseph Hendrie and commissioner Victor Gilinsky initially characterized the situation as "cause for concern but not alarm" [61].

Gilinsky briefed reporters and members of Congress, and informed White House staff. By 10:00 a.m., he met with two other commissioners. However, the NRC encountered the same difficulties in obtaining accurate information as the state, and was further hampered by its organizational unpreparedness for emergencies. It lacked a clear command structure and possessed no authority to direct the utility's actions or to order a local evacuation [62].

In a 2009 article, Gilinsky recounted that it took five weeks to ascertain that "the reactor operators had measured fuel temperatures near the melting point" [63]. He further stated, "We didn't learn for years—until the reactor vessel was physically opened—that by the time the plant operator called the NRC at about 8:00 a.m., roughly half of the uranium fuel had already melted" [63].

It remained unclear to the control room staff that the primary loop water levels were critically low and that over half of the core was exposed. A team of workers manually took readings from the thermocouples and obtained a sample of primary loop water. Seven hours into the emergency, new water was pumped into the primary loop, and the backup relief valve was opened to reduce pressure, allowing the loop to be refilled. After 16 hours, the primary loop pumps were reactivated, and the core temperature began to decrease. A significant portion of the core had melted, and the system was dangerously radioactive [citation needed].

The day after the accident, March 29, control room operators needed to confirm the integrity of the reactor vessel. To achieve this, a sample of the boron concentration in the primary system was required to ensure sufficient levels for complete reactor shutdown. Edward "Ed" Houser, the Unit 2 chemistry supervisor, volunteered to draw the sample, as his colleagues were hesitant. Richard Dubiel, the shift supervisor, requested that Pete Velez, the radiation protection foreman for Unit 2, accompany Houser. Velez was tasked with monitoring airborne radiation levels and ensuring neither individual would be overexposed [b].

Equipped with extensive protective gear—three pairs of gloves, rubber boots, and a respirator—the two navigated the reactor auxiliary building to collect the sample. However, Houser misplaced his pocket dosimeter while taking measurements. Houser observed that the sample he drew resembled "Alka-Seltzer" and registered extremely high radioactivity, with readings reaching 1,000 Rem/h. They spent five minutes in the building before withdrawing. Houser had exceeded the NRC's quarterly dose limit for radiation exposure (3 Rem/quarter in 1979) by one Rem and was only permitted to return to work the following quarter [c].

On the third day following the accident, a hydrogen bubble was detected in the dome [clarification needed] of the pressure vessel, becoming a primary focus of concern. A hydrogen explosion could breach the pressure vessel and, depending on its magnitude, potentially compromise the containment building's integrity, leading to a large-scale release of radioactive material. However, it was determined that no oxygen was present within the pressure vessel, a necessary condition for hydrogen combustion or explosion. Immediate measures were implemented to reduce the hydrogen bubble, and by the following day, it had significantly diminished in size. Over the subsequent week, steam and hydrogen were systematically removed from the reactor using a catalytic recombiner and by venting directly into the atmosphere [72].

Identification of released radioactive material

The release of radioactive material occurred when the fuel rod cladding was damaged while the PORV remained stuck open. Nuclear fission product isotopes were released into the reactor coolant. Given that the PORV was open and the loss-of-coolant accident was ongoing, the primary coolant, containing fission products and potentially fuel fragments, was discharged and eventually entered the auxiliary building, which was situated outside the primary containment boundary.

This pathway was confirmed by the radiation alarms that eventually activated. However, due to the fact that very little of the released fission products were solids at ambient temperature, minimal radiological contamination was detected in the environment. No significant radiation levels attributable to the TMI-2 accident were recorded beyond the TMI-2 facility itself. According to the Rogovin report, the overwhelming majority of the released radioisotopes consisted of noble gases, xenon and krypton, resulting in an average dose of 1.4 mrem (14 μSv) to the approximately two million people residing near the plant. For perspective, a chest X-ray administers a dose of 3.2 mrem (32 μSv), more than double the average dose received by individuals near the plant. On average, a U.S. resident is exposed to approximately 310 mrem (3,100 μSv) annually from natural background radiation [75].

Within hours of the accident, the United States Environmental Protection Agency (EPA) commenced daily environmental sampling at the three stations nearest to the plant. Continuous monitoring was established at 11 stations on April 1, expanding to 31 stations by April 3. An inter-agency analysis concluded that the accident did not elevate radioactivity levels sufficiently above background to cause even a single additional cancer death among the local population. However, measurements of beta radiation were not included, as the EPA found no contamination in water, soil, sediment, or plant samples [76].

Researchers at the nearby Dickinson College, equipped with radiation monitoring equipment sensitive enough to detect fallout from Chinese atmospheric atomic weapons testing, collected soil samples from the area over the subsequent two weeks. They detected no elevated radioactivity levels, except after rainfall, which was likely attributable to natural radon plate-out rather than the accident [77]. Furthermore, the tongues of white-tailed deer harvested more than 50 miles (80 km) from the reactor after the accident exhibited significantly higher levels of cesium-137 compared to deer in the counties immediately surrounding the power plant. Even these elevated levels remained below those observed in deer in other parts of the country during the peak of atmospheric nuclear weapons testing [78]. Had there been significant releases of radioactivity, increased levels of iodine-131 and cesium-137 would have been expected in milk samples from cattle and goats. Such elevated levels were not detected [79]. A subsequent study indicated that the officially reported emission figures were consistent with available dosimeter data [80], although other analyses have pointed to the incompleteness of this data, particularly concerning early releases [81].

Kemeny Commission

Numerous state and federal government agencies launched investigations into the crisis. The most prominent among these was the President's Commission on the Accident at Three Mile Island, established by U.S. President Jimmy Carter in April 1979. This commission comprised 12 individuals carefully selected for their neutrality regarding nuclear power, with John G. Kemeny, president of Dartmouth College, serving as chairman. Tasked with producing a final report within six months, the commission conducted public hearings, took depositions, and collected extensive documentation, releasing its comprehensive study on October 31, 1979.

According to official figures compiled by the Kemeny Commission from Met Ed and NRC data, the accident released a maximum of 480 PBq (13 MCi) of radioactive noble gases, predominantly xenon. These noble gases were considered relatively benign, with only 481–629 GBq (13.0–17.0 Ci) of thyroid cancer-causing iodine-131 released. These releases represented a small fraction of the estimated 370 EBq (10 GCi) contained within the reactor. It was later discovered that approximately half of the core had melted, and the cladding around 90% of the fuel rods had failed [22][85]. A substantial portion of the core, approximately 5 feet (1.5 m), was lost, with around 20 short tons (18 t) of uranium migrating to the bottom head of the pressure vessel, forming a mass of corium. Despite this severe damage, the reactor vessel—the second containment barrier—maintained its integrity, effectively containing the damaged fuel and nearly all the radioactive isotopes within the core [87].

Anti-nuclear political organizations challenged the Kemeny Commission's findings, presenting independent measurements that indicated radiation levels up to seven times higher than normal in locations hundreds of miles downwind from TMI [88]. Arnie Gundersen, a former nuclear industry executive and vocal critic of nuclear power, asserted, "I think the numbers on the NRC's website are off by a factor of 100 to 1,000" [90][verification needed][91]. Gundersen provided evidence, derived from pressure monitoring data, suggesting a hydrogen explosion occurred shortly before 2:00 p.m. on March 28, 1979, which could account for a significant release of radiation. Gundersen cited affidavits from four reactor operators indicating that the plant manager was aware of a dramatic pressure spike, followed by a drop to outside pressure. Gundersen also claimed the control room experienced shaking and doors were blown off their hinges. However, official NRC reports categorize this event merely as a "hydrogen burn" [90][verification needed].

The Kemeny Commission noted "a burn or an explosion that caused pressure to increase by 28 pounds per square inch (190 kPa) in the containment building" [92], while The Washington Post reported, "At about 2:00 p.m., with pressure almost down to the point where the huge cooling pumps could be brought into play, a small hydrogen explosion jolted the reactor" [93]. Subsequent research conducted for the Department of Energy in the 1980s determined that the hydrogen burn (deflagration), which went largely unnoticed for the first few days, occurred 9 hours and 50 minutes after the accident's initiation, lasted between 12 and 15 seconds, and did not involve a detonation [94][95].

The investigation heavily criticized Babcock & Wilcox, Met Ed, General Public Utilities, and the NRC for deficiencies in quality assurance and maintenance, inadequate operator training, failure to communicate critical safety information, poor management practices, and overall complacency. However, the commission refrained from drawing definitive conclusions regarding the future of the nuclear industry. The most pointed criticism from the Kemeny Commission stated that "... fundamental changes will be necessary in the organization, procedures, and practices—and above all—in the attitudes" of the NRC and the nuclear industry [97]. Kemeny himself characterized the operators' actions as "inappropriate" but acknowledged that they "were operating under procedures that they were required to follow, and our review and study of those indicates that the procedures were inadequate" and that the control room "was greatly inadequate for managing an accident" [98].

The Kemeny Commission also highlighted that Babcock & Wilcox's PORV had a history of malfunction, having failed on 11 previous occasions, nine of which involved sticking in the open position and allowing coolant to escape. The initial sequence of events at TMI had been replicated 18 months earlier at another Babcock & Wilcox reactor, the Davis–Besse Nuclear Power Station. The critical differences were that operators at Davis–Besse identified the valve failure within 20 minutes, whereas at TMI it took 80 minutes, and the Davis–Besse facility was operating at 9% power compared to TMI's 97%. Despite recognizing the problem, Babcock engineers failed to adequately inform their customers about the valve issue [99].

The Pennsylvania House of Representatives conducted its own investigation, focusing primarily on the necessity of improving evacuation procedures [100].

In 1985, a television camera was employed to survey the interior of the damaged reactor. In 1986, core samples and debris samples were retrieved from the corium layers at the bottom of the reactor vessel and subsequently analyzed [101].

Mitigation policies

Voluntary evacuation

On Wednesday, March 28, mere hours after the accident began, Lieutenant Governor Scranton informed the public at a press briefing that Met Ed had assured the state that "everything is under control" [102]. Later that day, Scranton revised his statement, acknowledging that the situation was "more complex than the company first led us to believe" [102]. Conflicting reports regarding radioactivity releases fueled public anxiety. Consequently, schools were closed, and residents were advised to remain indoors. Farmers were instructed to keep their livestock sheltered and to use stored feed [102][103].

Based on the advice of the Chairman of the NRC and in the interest of taking every precaution, I am advising those who may be particularly susceptible to the effects of any radiation, that is, pregnant women and pre-school aged children, to leave the area within a five-mile radius of the Three Mile Island facility until further notice. We've also ordered the closings of any schools within this area.

— Dick Thornburgh

Governor Thornburgh, acting on the recommendation of NRC chairman Joseph Hendrie, advised the evacuation of "pregnant women and pre-school age children... within a five-mile radius of the Three Mile Island facility." This evacuation zone was subsequently expanded to a 20-mile radius on March 30 [104]. Within days, approximately 140,000 individuals had departed the area [22][102][105]. Despite the evacuation advisories, over half of the 663,500 residents living within the 20-mile radius remained in their homes [104]. A survey conducted in April 1979 indicated that 98% of those who evacuated had returned to their residences within three weeks [104].

Post-TMI surveys revealed that less than 50% of the American public expressed satisfaction with the handling of the accident by Pennsylvania state officials and the NRC. Public approval was even lower concerning the utility (General Public Utilities) and the plant designer [106].

Effect on nuclear power industry

The global trajectory of nuclear power development was significantly impacted by the Three Mile Island accident, as noted by the IAEA [107]. Between 1963 and 1979, the number of reactors under construction worldwide saw consistent annual increases, with the sole exceptions of 1971 and 1978. However, in the aftermath of the TMI incident, the number of reactors under construction in the U.S. declined from 1980 to 1998, accompanied by escalating construction costs and extended completion timelines for some projects [108]. Consequently, numerous Babcock & Wilcox reactors that had been ordered were canceled. In total, 52 U.S. nuclear reactors were canceled between 1980 and 1984 [109].

The accident, while not the sole cause of the U.S. nuclear power industry's decline, did effectively halt its established growth. Furthermore, preceding the TMI event, the 1973 oil crisis and subsequent analyses indicating potential overcapacity in base load generation had already led to the cancellation of 40 planned nuclear power plants. At the time of the TMI incident, 129 nuclear power plants had received approval, but only 53 of those not already operational were ultimately completed. The prolonged regulatory review process, further complicated by the Chernobyl disaster seven years later, imposed increasingly stringent federal safety requirements and design standards. Local opposition intensified, construction times were significantly lengthened, and costs escalated dramatically [110]. Until 2012, no new U.S. nuclear power plant had been authorized for construction since 1978. Globally, the trend of increasing nuclear power plant construction reversed more catastrophically with the Chernobyl disaster in 1986 (see graph).

Cleanup

Initially, GPU harbored plans to repair the reactor and return it to service [112]. However, TMI-2 was deemed too severely damaged and contaminated for further operation; the reactor was progressively deactivated and permanently shut down. Although TMI-2 had been operational for a mere three months, it now housed a ruined reactor vessel and a containment building rendered unsafe for entry. The cleanup operation commenced in August 1979 and officially concluded in December 1993, with total cleanup costs amounting to approximately $1 billion [19]. Benjamin K. Sovacool, in his 2007 preliminary assessment of major energy accidents, estimated the total property damages resulting from the TMI accident at$2.4 billion [113].

Efforts were concentrated on the cleanup and decontamination of the site, with a particular focus on the defueling of the damaged reactor. Beginning in 1985, nearly 100 short tons (91 t) of radioactive fuel were removed from the site. The planning and execution of this process were partially hindered by overly optimistic assessments of the extent of the damage [114].

In 1988, the NRC announced that while further decontamination of the Unit 2 site was feasible, the residual radioactivity had been sufficiently contained to pose no threat to public health and safety. The initial major phase of the cleanup was completed in 1990, with workers finishing the shipment of 150 short tons (140 t) of radioactive wreckage to Idaho for storage at the Department of Energy's National Engineering Laboratory. However, the contaminated cooling water that had leaked into the containment building had permeated the concrete structure, leaving behind radioactive residue that was deemed too impractical to remove. Consequently, further cleanup efforts were postponed to allow for the natural decay of radiation levels and to capitalize on the potential economic advantages of retiring both Unit 1 and Unit 2 concurrently [19].

Health effects and epidemiology

Main article: Three Mile Island accident health effects

In the wake of the accident, investigations focused on quantifying the amount of radioactivity released. In total, approximately 2.5 megacuries (93 PBq) of radioactive gases and approximately 15 curies (560 GBq) of iodine-131 were released into the environment [115]. According to the American Nuclear Society, based on the official radioactivity emission figures, "The average radiation dose to people living within 10 miles of the plant was eight millirem (0.08 mSv), and no more than 100 millirem (1 mSv) to any single individual. Eight millirem is about equal to a chest X-ray, and 100 millirem is about a third of the average background level of radiation received by US residents in a year" [116].

Health researcher Joseph Mangano noted that early scientific publications estimated no additional cancer deaths in the 10-mile (16 km) area around TMI, based on these figures. However, disease rates in areas beyond 10 miles from the plant were not examined [88]. Local activism in the 1980s, spurred by anecdotal reports of adverse health effects, led to the commissioning of scientific studies. A variety of epidemiological studies have concluded that the accident had no observable long-term health effects [d].

A peer-reviewed research article by Dr. Steven Wing identified a significant increase in cancers between 1979 and 1985 among individuals residing within ten miles of TMI [119]. In 2009, Dr. Wing stated that radiation releases during the accident were likely "thousands of times greater" than the NRC's estimates. A retrospective analysis of the Pennsylvania Cancer Registry revealed an increased incidence of thyroid cancer in certain counties south of TMI (though notably not in Dauphin County, where the reactor was located) and within high-risk age groups. However, the study did not establish a causal link between these occurrences and the accident [13][14]. The Talbott lab at the University of Pittsburgh reported identifying a few small increases in cancer risks within the TMI population [15]. A more recent study reported "findings consistent with observations from other radiation-exposed populations," suggesting "the possibility that radiation released from [Three Mile Island] may have altered the molecular profile of [thyroid cancer] in the population surrounding TMI," thereby establishing a potential causal mechanism, though not definitively proving causation [120].

The Radiation and Public Health Project, an organization with limited credibility among epidemiologists [121], cited calculations by Mangano indicating a spike in infant mortality in downwind communities two years after the accident [88][122]. Anecdotal accounts also document effects on the region's wildlife [88]. John Gofman utilized his own non-peer reviewed low-level radiation health model to predict 333 excess cancer or leukemia deaths resulting from the 1979 Three Mile Island accident [11]. The ongoing epidemiological research concerning TMI has been accompanied by discussions regarding methodological challenges in dose estimation due to a lack of precise data and variations in illness classifications [123].

Activism and legal action

Anti-nuclear protests erupted in Harrisburg, Pennsylvania, the state capital, following the Three Mile Island accident on April 9, 1979. The accident significantly bolstered the perceived credibility of anti-nuclear groups and triggered protests globally [124][125]. President Carter, who had specialized in nuclear power during his service in the United States Navy, informed his cabinet after visiting the plant that the accident was minor. However, he reportedly refrained from making this assessment public to avoid alienating Democrats who opposed nuclear power [126].

In the ensuing months, concerned members of the American public staged numerous anti-nuclear demonstrations across the nation in response to the release of radioactive gas from the accident. The largest demonstration, held in New York City in September 1979, drew 200,000 participants, with speeches delivered by prominent figures like Jane Fonda and Ralph Nader [127][128][129]. This rally coincided with a series of nightly "No Nukes" concerts held at Madison Square Garden from September 19 to 23, featuring performances by Musicians United for Safe Energy. In May of that year, an estimated 65,000 people, including California Governor Jerry Brown, participated in a march and rally against nuclear power in Washington, D.C. [128].

In 1981, citizen groups achieved success in a class-action lawsuit against TMI, securing a $25 million out-of-court settlement. A portion of these funds was allocated to establish the TMI Public Health Fund [130]. In 1983, a federal grand jury indicted Metropolitan Edison on criminal charges for falsifying safety test results prior to the accident [131]. Through a plea-bargaining agreement, Met Ed pleaded guilty to one count of falsifying records and no contest to six other charges, with four counts dismissed. The company agreed to pay a$45,000 fine and establish a $1 million fund to support emergency planning in the area surrounding the plant [132].

According to Eric Epstein, chairman of Three Mile Island Alert, the TMI plant operator and its insurers disbursed at least $82 million in publicly documented compensation to residents for "loss of business revenue, evacuation expenses and health claims" [133]. However, a class action lawsuit alleging detrimental health effects caused by the accident was dismissed by Harrisburg United States district court Judge Sylvia Rambo, and the final appeal of this decision was unsuccessful in 2002 [134][135].

Normal accident theory

The Three Mile Island accident served as a primary catalyst for Charles Perrow's influential "normal accident theory." This theory posits that in highly complex systems, unanticipated interactions among multiple failures are not merely possible but inevitable, leading to accidents that are "unexpected, incomprehensible, uncontrollable and unavoidable" [136].

Perrow concluded that the TMI failure stemmed from the inherent complexity of the system. He argued that such modern, high-risk technologies are prone to failures regardless of management efficacy, making what he termed a 'normal accident' an eventual certainty. Consequently, he proposed that a radical redesign of such systems or, failing that, their complete abandonment might be the more prudent course of action [137].

The concept of "normal" accidents, or system accidents, arises from the premise that within extremely complex systems, multiple failures that interact in unpredictable ways are bound to occur, despite all efforts to prevent them [138]. Events that initially appear trivial can cascade and amplify unpredictably, ultimately culminating in catastrophic outcomes [139].

The theory of Normal Accidents significantly contributed to a paradigm shift in the understanding of safety and risk during the 1980s. It underscored the importance of examining technological failures not as isolated incidents of equipment malfunction, operator error, or acts of nature, but as products of highly interacting systems, with organizational and managerial factors playing a central role in their causation [137].

Comparison to U.S. Navy operations

Following the TMI incident, President Carter commissioned a study, the "Report of the President's Commission on the Accident at Three Mile Island" (1979) [92].

Admiral Hyman G. Rickover was later called upon to explain to Congress why naval nuclear propulsion (as utilized in submarines) had not experienced any reactor accidents, defined as the uncontrolled release of fission products resulting from core damage. In his testimony, Rickover stated:

Over the years, many people have asked me how I run the Naval Reactors Program, so that they might find some benefit for their own work. I am always chagrined at the tendency of people to expect that I have a simple, easy gimmick that makes my program function. Any successful program functions as an integrated whole of many factors. Trying to select one aspect as the key one will not work. Each element depends on all the others [140].

21st-century status

Following the TMI-2 incident, the NRC suspended the operating license for TMI-1, which was owned and operated by Met Ed, a subsidiary of General Public Utilities Corporation. In 1982, residents of the three surrounding counties overwhelmingly voted in a non-binding resolution to permanently retire Unit 1 [141][142]. In 1985, by a 4–1 vote, the Nuclear Regulatory Commission authorized the resumption of operations at TMI-1 [141][142].

GPU established General Public Utilities Nuclear Corporation as a subsidiary to manage and operate the company's nuclear facilities, including Three Mile Island. In 1996, General Public Utilities shortened its name to GPU Inc. By 1998, GPU had sold TMI-1 to AmerGen Energy Corporation, a joint venture between Philadelphia Electric Company and British Energy. GPU was contractually obligated to continue maintaining and monitoring TMI-2. In 2001, GPU was acquired by FirstEnergy Corporation and subsequently dissolved, with the maintenance and administration of Unit 2 being contracted to AmerGen.

In 2000, Philadelphia Electric merged with Unicom Corporation to form Exelon. By 2003, Exelon had acquired the remaining shares of AmerGen from British Energy. In 2009, Exelon Nuclear absorbed and dissolved AmerGen. In addition to TMI Unit 1, Exelon Nuclear operates the Clinton Power Station and several other nuclear facilities [e].

Unit 2 remains under license and regulation by the Nuclear Regulatory Commission in a status known as Post Defueling Monitored Storage [148]. The TMI-2 reactor has been permanently shut down. Its coolant system has been drained, the radioactive water decontaminated and evaporated, radioactive waste shipped off-site, and the reactor fuel and most core debris transported to a Department of Energy facility. The remainder of the site is under ongoing monitoring. The owner had intended to maintain the facility in long-term monitored storage until the expiration of TMI-1's operating license, at which point both plants would undergo decommissioning [22].

In 2009, the NRC granted a license extension, permitting the TMI-1 reactor to operate until April 19, 2034 [149][150]. By 2017, it was announced that operations would cease by 2019 due to financial pressures from inexpensive natural gas, unless legislative intervention occurred to sustain its operation [151]. When it became apparent that subsidy legislation would not pass, Exelon made the decision to retire the plant [152]. TMI Unit 1 ceased operations on September 20, 2019, entering the decommissioning phase and moving to SAFSTOR status [153][154].

In 2020, the site was acquired by TMI-2 Solutions, a subsidiary of EnergySolutions, with the objective of completing the site cleanup at a lower cost than the allocated dedicated fund [155]. On May 8, 2023, TMI-2 Solutions announced that 99% of the nuclear fuel had been removed and that the site had entered the next phase of cleanup, projected to last until 2029 [156][155]. TMI-2 Solutions plans to finalize the cleanup and demolish the plant by 2052 [155].

In September 2024, Constellation Energy revealed plans to restart the Three Mile Island nuclear plant, intending to supply power to Microsoft for its data centers, highlighting the substantial energy demands of the technology sector, particularly in supporting artificial intelligence infrastructure [157]. Constellation anticipates the Unit 1 reactor at Three Mile Island to recommence operations in 2028, pending approval from the Nuclear Regulatory Commission. Constellation also plans to seek an extension of the plant's operational lifespan to at least 2054 [158].

Timeline

| Date | Event | | :---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

• Date: March 28, 1979
• Time: 04:00 Eastern Time Zone UTC−5
• Location: Londonderry Township, Dauphin County, Pennsylvania, U.S.
• Outcome: INES Level 5 (accident with wider consequences)

The Three Mile Island accident was a partial nuclear meltdown of the Unit 2 reactor (TMI-2) of the Three Mile Island Nuclear Generating Station, situated on the Susquehanna River in Londonderry Township, Dauphin County, near Harrisburg, Pennsylvania. The accident commenced at 4:00 a.m. on March 28, 1979, and resulted in the release of radioactive gases and radioactive iodine into the environment. It is recognized as the most severe accident in U.S. commercial nuclear power plant history, although its limited radioactive releases had no detectable health impacts on plant workers or the public. Rated Level 5, an "Accident with Wider Consequences," on the seven-point logarithmic International Nuclear Event Scale, the TMI-2 reactor accident was a pivotal event.

The incident began with failures in the non-nuclear secondary system [7], followed by a stuck-open pilot-operated relief valve (PORV) in the primary system [8]. This valve allowed substantial amounts of water to escape from the pressurized, isolated coolant loop. The mechanical failures were compounded by the initial inability of plant operators to identify the situation as a loss-of-coolant accident (LOCA). Inadequate training and operational procedures at TMI left operators and management unprepared for the escalating crisis. These shortcomings were further exacerbated by design flaws, including poor control design, the use of multiple similar alarms, and a failure of equipment to indicate crucial information such as coolant inventory level or the position of the stuck-open PORV [9].

The accident significantly heightened public concerns about nuclear safety and led to the implementation of new regulations for the nuclear industry, accelerating the decline in new reactor construction efforts [10]. Activists within the anti-nuclear movement raised alarms about potential regional health effects stemming from the accident [11]. While some epidemiological studies examining cancer rates in and around the affected area since the accident reported a statistically significant increase, others did not. Due to the nature of these studies, establishing a definitive causal link between the accident and cancer remains difficult [a]. The cleanup of TMI-2, which began in August 1979, officially concluded in December 1993, with a total expenditure of approximately $1 billion (equivalent to$2 billion in 2024). TMI-1 was restarted in 1985 but was retired in 2019 due to operating losses. It is anticipated to resume service in either 2027 or 2028 as part of a deal with Microsoft to power its data centers [20][21].

Accident

Background

A simplified schematic diagram of the TMI-2 plant [22].

In the late hours preceding the incident, the TMI-2 reactor was operating at 97% power, while the companion TMI-1 reactor was offline for refueling [23]. The chain of events leading to the partial core meltdown on March 28, 1979, began at 4:00:36 a.m. EST within TMI-2's secondary loop, one of the three main water/steam circuits in a pressurized water reactor [24].

The accident's initial cause stemmed from an attempt to clear a blockage in one of the eight condensate polishers, sophisticated filters designed to remove impurities from the secondary loop water, thereby preventing buildup in the steam generators and reducing corrosion.

Blockages in these resin filters are common and typically resolved easily. However, in this instance, the usual method of using compressed air failed. Operators then employed a technique of forcing compressed air into the water, relying on water pressure to dislodge the resin. This process inadvertently allowed a small volume of water to bypass a stuck-open check valve and enter an instrument air line. This infiltration eventually led to the shutdown of the feedwater pumps, condensate booster pumps, and condensate pumps around 4:00 a.m., triggering a turbine trip.

Reactor overheating and malfunction of relief valve

With feedwater flow to the steam generators halted, heat transfer from the reactor coolant system (RCS) [25] was significantly diminished, causing RCS temperature to rise. The expanding, rapidly heating coolant surged into the pressurizer [26][27][28], compressing the steam bubble at its top. When RCS pressure reached 2,255 psi (155.5 bar), the pilot-operated relief valve (PORV) opened, releasing steam through piping to the reactor coolant drain tank [29] in the containment building basement. RCS pressure continued to climb, reaching the reactor protection system high-pressure trip setpoint of 2,355 psi (162.4 bar) eight seconds after the turbine trip. The reactor automatically tripped, with its control rods falling into the core under gravity, halting the nuclear chain reaction and subsequent fission heat generation. However, the reactor continued to produce decay heat, initially equivalent to approximately 6% of the pre-trip power level. With steam no longer being utilized by the turbine and no feedwater being supplied to the steam generators, heat removal from the reactor's primary water loop was limited to the steam generated by the residual water in the secondary side of the steam generators, vented to the condenser via turbine bypass valves [30][31][32].

When the feedwater pumps tripped, three emergency feedwater pumps automatically activated. An operator observed the pumps running but failed to notice that a block valve in each of the two emergency feedwater lines was closed, preventing emergency feed flow to both steam generators. The valve position lights for one block valve were obscured by a yellow maintenance tag. The reason for the operator's failure to note the status of the second valve is unknown, though one theory suggests his physical size may have blocked his view [33]. The valves may have been left closed during a surveillance test conducted two days prior [34][35]. The closure of these valves violated a key Nuclear Regulatory Commission (NRC) rule requiring reactor shutdown if all auxiliary feed pumps are closed for maintenance. This violation was later identified by NRC officials as a critical failure [36].

After the reactor trip, secondary system steam valves operated to reduce steam generator temperature and pressure, cooling the RCS and causing the primary coolant to contract [Thermal expansion]. This contraction, combined with the coolant loss through the open PORV, led to a drop in RCS pressure and, after peaking 15 seconds post-turbine trip, a decrease in pressurizer level. Concurrently, at 15 seconds after the turbine trip, coolant pressure had fallen to 2,205 psi (152.0 bar), the reset setpoint for the PORV. Electrical power to the PORV's solenoid was automatically cut, but the relief valve remained stuck open, allowing continuous coolant release [37].

Post-accident investigations revealed that the PORV's indication system was one of many design flaws identified in the operators' controls, instruments and alarms. There was no direct indication of the valve's actual position. A light on the control panel, installed after the PORV had previously stuck open during startup testing [38], illuminated when the PORV opened [39]. When this light, labeled "Light on – RC-RV2 open" [40], went out, operators mistakenly believed the valve had closed. In reality, the light only indicated that the PORV pilot valve's solenoid was energized, not the valve's physical state [41]. While the main relief valve remained stuck open, the operators, interpreting the unlit lamp as confirmation of closure, failed to correctly diagnose the problem for several hours [42].

Operators had not been adequately trained to understand the ambiguous nature of the PORV indicator or to seek alternative confirmation of the main relief valve's closure. A downstream temperature indicator, whose sensor was located in the tailpipe between the PORV and the pressurizer relief tank, could have signaled a stuck valve had operators noticed its higher-than-normal reading. However, it was not considered part of the "safety grade" suite of indicators designated for use after an incident, and personnel had not been trained to utilize it. Its placement behind the seven-foot-high instrument panel also rendered it effectively out of sight [43].

Depressurization of primary reactor cooling system

Less than a minute after the event began, the water level in the pressurizer started to rise, even as RCS pressure was falling. With the PORV stuck open, coolant was being lost from the RCS, constituting a loss-of-coolant accident (LOCA). Expected LOCA symptoms include drops in both RCS pressure and pressurizer level. The operators' training and plant procedures did not cover a scenario where these two parameters moved in opposite directions. The rising pressurizer water level was attributed to steam venting from the pressurizer space through the stuck-open PORV, which lowered the pressurizer pressure due to the lost coolant inventory. This pressure reduction caused water from the coolant loop to surge into the pressurizer, creating a steam bubble in the reactor pressure vessel head, facilitated by the decay heat from the fuel [44].

This steam bubble was invisible to the operators, and this mechanism had not been included in their training. Indications of high pressurizer water levels contributed to confusion, as operators were concerned about the primary loop "going solid" (i.e., the absence of a steam pocket buffer in the pressurizer), a condition they were instructed to avoid at all costs during training. This confusion was a key factor in the initial failure to recognize the accident as a LOCA [45] and led operators to shut down the emergency core cooling pumps, which had automatically activated following the PORV sticking and coolant loss, due to concerns about overfilling the system [46][47][48].

With the PORV still open, the pressurizer relief tank that collected the discharge overfilled, causing the containment building sump to fill and trigger an alarm at 4:11 a.m. This alarm, along with higher-than-normal temperatures on the PORV discharge line and elevated containment building temperatures and pressures, served as clear indications of an ongoing LOCA, but these were initially disregarded by operators [49][50]. At 4:15 a.m., the relief diaphragm of the pressurizer relief tank ruptured, and radioactive coolant began to leak into the general containment building. This radioactive coolant was pumped from the containment building sump to an auxiliary building, outside the main containment, until the sump pumps were shut off at 4:39 a.m. [50].

Partial meltdown and further release of radioactive substances

Around 5:20 a.m., after nearly 80 minutes of a growing steam bubble in the reactor pressure vessel head, the primary loop's four main reactor coolant pumps began to cavitate as a steam-water mixture, rather than water, flowed through them. The pumps were shut down, with the expectation that natural circulation would maintain water movement. However, steam in the system prevented flow through the core, and as water circulation ceased, it increasingly converted to steam. Shortly after 6:00 a.m., the top of the reactor core became exposed. The intense heat caused a reaction between the steam forming in the core and the zircaloy cladding of the nuclear fuel rod, producing zirconium dioxide, hydrogen, and additional heat. This reaction melted the fuel rod cladding and damaged the fuel pellets, releasing radioactive isotopes into the reactor coolant and generating hydrogen gas, which is believed to have caused a minor explosion in the containment building later that afternoon [51].

A graphic illustrating the end-state configuration of the TMI-2 core:

• 2B inlet
• 1A inlet
• cavity
• loose core debris
• crust
• previously molten material
• lower plenum debris
• possible region depleted in uranium
• ablated incore instrument guide
• hole in baffle plate
• coating of previously molten material on bypass region interior surfaces
• upper grid damage

At 6:00 a.m., a shift change occurred in the control room. A new operator noticed excessive temperatures in the PORV tailpipe and holding tanks and used a backup valve, known as a block valve, to stop the coolant discharge through the PORV. By this time, approximately 32,000 US gallons (120,000 L) of coolant had leaked from the primary loop [52][53]. It wasn't until 6:45 a.m., 165 minutes after the problem began, that radiation alarms activated as the contaminated water reached detectors. At this point, radiation levels in the primary coolant water were about 300 times expected levels, and the general containment building was heavily contaminated, with radiation levels reaching 800 Rem (unit)/h.

Emergency declaration and immediate aftermath

At 6:56 a.m., a plant supervisor declared a site area emergency. Less than 30 minutes later, station manager Gary Miller announced a general emergency. Metropolitan Edison (Met Ed) notified the Pennsylvania Emergency Management Agency, which subsequently contacted state and local agencies. Pennsylvania Governor Richard L. Thornburgh was informed, and he assigned responsibility for gathering and reporting information about the accident to Lieutenant Governor William Scranton III [55]. The uncertainty among operators at the plant was reflected in the fragmented, ambiguous, or contradictory statements made by Met Ed to government agencies and the press, particularly regarding the possibility and severity of off-site radioactivity releases [56].

At a press conference, Scranton attempted to be reassuring while conveying confusion about potential off-site releases, stating that although a "small release of radiation... had occurred, no increase in normal radiation levels" had been detected. These assertions were contradicted by another official and by Met Ed's own statements, both claiming no radioactivity had been released. Readings from instruments at the plant and off-site detectors had indeed registered radioactivity releases, albeit at levels unlikely to threaten public health if they were temporary and containment of the highly contaminated reactor was maintained [58].

Upset by Met Ed's failure to inform them before venting steam from the plant and convinced the company was downplaying the accident's severity, state officials turned to the NRC. After being notified of the accident by Met Ed, the NRC activated its emergency response headquarters in Bethesda, Maryland, and dispatched personnel to Three Mile Island. NRC chairman Joseph Hendrie and commissioner Victor Gilinsky initially viewed the accident as a "cause for concern but not alarm" [61].

Gilinsky briefed reporters and members of Congress and informed White House staff. By 10:00 a.m., he met with two other commissioners. However, the NRC faced the same challenges in obtaining accurate information as the state and was further hampered by its organizational unpreparedness for emergencies, lacking a clear command structure and the authority to direct the utility's actions or order a local evacuation [62].

In a 2009 article, Gilinsky wrote that it took five weeks to learn that "the reactor operators had measured fuel temperatures near the melting point" [63]. He further stated, "We didn't learn for years—until the reactor vessel was physically opened—that by the time the plant operator called the NRC at about 8:00 a.m., roughly half of the uranium fuel had already melted" [63].

It remained unclear to control room staff that primary loop water levels were low and that over half the core was exposed. A group of workers took manual readings from the thermocouples and obtained a sample of primary loop water. Seven hours into the emergency, new water was pumped into the primary loop, and the backup relief valve was opened to reduce pressure, enabling the loop to be filled. After 16 hours, the primary loop pumps were restarted, and the core temperature began to decrease. A significant portion of the core had melted, and the system was dangerously radioactive [citation needed].

On the day following the accident, March 29, control room operators needed to ensure the integrity of the reactor vessel. To do this, a sample of the boron concentration in the primary system was required to confirm sufficient levels for complete reactor shutdown. Edward "Ed" Houser, the Unit 2 chemistry supervisor, volunteered to draw the sample after his colleagues hesitated. Shift supervisor Richard Dubiel asked Pete Velez, the radiation protection foreman for Unit 2, to accompany Houser. Velez's role was to monitor airborne radiation levels and ensure neither individual incurred overexposure [b].

Wearing extensive protective gear—three pairs of gloves, rubber boots, and a respirator—the two navigated the reactor auxiliary building to draw the sample. However, Houser lost his pocket dosimeter while taking measurements. Houser noted the sample appeared "like Alka-Seltzer" and was highly radioactive, with readings as high as 1,000 Rem/h. The pair spent five minutes in the building before withdrawing. Houser had exceeded the NRC's quarterly dose limit for radiation exposure (3 Rem/quarter in 1979) by one Rem and was eligible to return to work only in the following quarter [c].

On the third day after the accident, a hydrogen bubble was detected in the dome [clarification needed] of the pressure vessel, becoming a major concern. A hydrogen explosion could breach the pressure vessel and potentially compromise the containment building, leading to a large release of radioactive material. However, it was determined that no oxygen was present in the pressure vessel, a prerequisite for hydrogen combustion or explosion. Immediate steps were taken to reduce the hydrogen bubble, and it was significantly smaller by the following day. Over the next week, steam and hydrogen were removed from the reactor using a catalytic recombiner and by venting directly to the atmosphere [72].

Identification of released radioactive material

The release occurred when the fuel rod cladding was damaged while the PORV remained stuck open. Nuclear fission product isotopes were released into the reactor coolant. As the PORV was open and the loss-of-coolant accident was ongoing, primary coolant containing fission products and possibly fuel was discharged and ultimately entered the auxiliary building, which was outside the containment boundary.

This was evidenced by the radiation alarms that eventually sounded. However, since very little of the released fission products were solids at room temperature, minimal radiological contamination was reported in the environment. No significant radiation levels attributable to the TMI-2 accident were detected outside the TMI-2 facility. According to the Rogovin report, the vast majority of the released radioisotopes were noble gases, xenon and krypton, resulting in an average dose of 1.4 mrem (14 μSv) to the approximately two million people near the plant. In comparison, a chest X-ray delivers a dose of 3.2 mrem (32 μSv), more than twice the average dose received by those near the plant. On average, a U.S. resident receives an annual radiation dose from natural sources of about 310 mrem (3,100 μSv) [75].

Within hours of the accident, the United States Environmental Protection Agency (EPA) began daily environmental sampling at the three stations closest to the plant. Continuous monitoring was established at 11 stations on April 1 and expanded to 31 stations on April 3. An inter-agency analysis concluded that the accident did not elevate radioactivity far enough above background levels to cause even one additional cancer death among the local population. However, measurements of beta radiation were not included, as the EPA found no contamination in water, soil, sediment, or plant samples [76].

Researchers at nearby Dickinson College, equipped with radiation monitoring equipment sensitive enough to detect fallout from Chinese atmospheric atomic weapons testing, collected soil samples from the area over the next two weeks. They detected no elevated radioactivity levels, except after rainfalls, which were likely due to natural radon plate-out, not the accident [77]. Furthermore, the tongues of white-tailed deer harvested over 50 miles (80 km) from the reactor after the accident showed significantly higher levels of cesium-137 than deer in the counties immediately surrounding the power plant. Even then, these elevated levels remained below those observed in deer elsewhere in the country during the peak of atmospheric nuclear weapons testing [78]. Had there been significant releases of radioactivity, increased levels of iodine-131 and cesium-137 would have been expected in milk samples from cattle and goats. Such elevated levels were not found [79]. A later study indicated that the official emission figures were consistent with available dosimeter data [80], although other analyses have pointed out the incompleteness of this data, particularly for early releases [81].

Kemeny Commission

Several state and federal government agencies initiated investigations into the crisis. The most prominent was the President's Commission on the Accident at Three Mile Island, established by U.S. President Jimmy Carter in April 1979 [82]. The commission consisted of 12 individuals carefully chosen for their lack of strong pro- or anti-nuclear views, chaired by John G. Kemeny, president of Dartmouth College. Tasked with delivering a final report within six months, the commission conducted public hearings, took depositions, and collected documents, releasing its findings on October 31, 1979 [83].

According to official figures compiled by the Kemeny Commission from Met Ed and NRC data, the accident released a maximum of 480 PBq (13 MCi) of radioactive noble gases, primarily xenon. These noble gases were considered relatively harmless, with only 481–629 GBq (13.0–17.0 Ci) of thyroid cancer-causing iodine-131 released. These releases represented a small fraction of the estimated 370 EBq (10 GCi) contained within the reactor. It was later determined that approximately half of the core had melted, and the cladding around 90% of the fuel rods had failed [22][85]. A significant portion of the core, approximately 5 feet (1.5 m), was lost, with about 20 short tons (18 t) of uranium migrating to the bottom head of the pressure vessel, forming a mass of corium. The reactor vessel—the second containment barrier after the cladding—maintained its integrity, containing the damaged fuel and nearly all radioactive isotopes within the core [87].

Anti-nuclear political groups contested the Kemeny Commission's findings, asserting that independent measurements indicated radiation levels up to seven times higher than normal in areas hundreds of miles downwind from TMI [88]. Arnie Gundersen, a former nuclear industry executive and anti-nuclear advocate, stated, "I think the numbers on the NRC's website are off by a factor of 100 to 1,000" [90][verification needed][91]. Gundersen presented evidence, based on pressure monitoring data, suggesting a hydrogen explosion occurred shortly before 2:00 p.m. on March 28, 1979, which could have facilitated a significant radiation release. Gundersen cited affidavits from four reactor operators indicating that the plant manager was aware of a dramatic pressure spike followed by a drop to ambient pressure. Gundersen also claimed the control room shook and doors were dislodged. Official NRC reports, however, refer only to a "hydrogen burn" [90][verification needed].

The Kemeny Commission mentioned "a burn or an explosion that caused pressure to increase by 28 pounds per square inch (190 kPa) in the containment building" [92], while The Washington Post reported, "At about 2:00 p.m., with pressure almost down to the point where the huge cooling pumps could be brought into play, a small hydrogen explosion jolted the reactor" [93]. Research conducted for the Department of Energy in the 1980s concluded that the hydrogen burn (deflagration), which went largely unnoticed for the first few days, occurred 9 hours and 50 minutes after the accident's initiation, lasted 12 to 15 seconds, and did not involve a detonation [94][95].

The investigation strongly criticized Babcock & Wilcox, Met Ed, General Public Utilities, and the NRC for lapses in quality assurance and maintenance, inadequate operator training, lack of communication of critical safety information, poor management, and complacency. However, the commission avoided drawing conclusions about the future of the nuclear industry. The Kemeny Commission's most severe criticism stated that "... fundamental changes will be necessary in the organization, procedures, and practices—and above all—in the attitudes" of the NRC and the nuclear industry [97]. Kemeny described the operators' actions as "inappropriate" but noted that they "were operating under procedures that they were required to follow, and our review and study of those indicates that the procedures were inadequate" and that the control room "was greatly inadequate for managing an accident" [98].

The Kemeny Commission observed that Babcock & Wilcox's PORV had previously failed on 11 occasions, nine of which involved sticking in the open position, allowing coolant to escape. The initial sequence of events at TMI had been duplicated 18 months earlier at another Babcock & Wilcox reactor, the Davis–Besse Nuclear Power Station. The key differences were that operators at Davis–Besse identified the valve failure within 20 minutes, whereas at TMI it took 80 minutes, and the Davis–Besse facility was operating at 9% power compared to TMI's 97%. Despite recognizing the problem, Babcock engineers failed to clearly notify their customers about the valve issue [99].

The Pennsylvania House of Representatives conducted its own investigation, focusing on the need to improve evacuation procedures [100].

In 1985, a television camera was used to view the interior of the damaged reactor. In 1986, core samples and debris samples were obtained from the corium layers at the bottom of the reactor vessel and analyzed [101].

Mitigation policies

Voluntary evacuation

On Wednesday, March 28, hours after the accident began, Lieutenant Governor Scranton stated at a press briefing that Met Ed had assured the state that "everything is under control" [102]. Later that day, Scranton revised his statement, acknowledging the situation was "more complex than the company first led us to believe" [102]. Conflicting reports about radioactivity releases heightened public concern. Consequently, schools were closed, and residents were advised to stay indoors. Farmers were instructed to keep their animals under cover and on stored feed [102][103].

Based on the advice of the Chairman of the NRC and in the interest of taking every precaution, I am advising those who may be particularly susceptible to the effects of any radiation, that is, pregnant women and pre-school aged children, to leave the area within a five-mile radius of the Three Mile Island facility until further notice. We've also ordered the closings of any schools within this area.

— Dick Thornburgh

Governor Thornburgh, following the advice of NRC chairman Joseph Hendrie, recommended the evacuation of "pregnant women and pre-school age children... within a five-mile radius of the Three Mile Island facility." The evacuation zone was expanded to a 20-mile radius on March 30 [104]. Within days, 140,000 people had left the area [22][102][105]. Over half of the 663,500 population within the 20-mile radius remained in their homes [104]. A survey conducted in April 1979 indicated that 98% of the evacuees had returned home within three weeks [104].

Post-TMI surveys revealed that less than 50% of the American public was satisfied with the handling of the accident by Pennsylvania state officials and the NRC. Public satisfaction was even lower regarding the utility (General Public Utilities) and the plant designer [106].

Effect on nuclear power industry

The Three Mile Island accident marked a significant turning point in the global development of nuclear power, as noted by the IAEA [107]. From 1963 to 1979, the number of reactors under construction worldwide increased annually, with the sole exceptions of 1971 and 1978. However, following the incident, the number of reactors under construction in the U.S. declined from 1980 to 1998, accompanied by rising construction costs and extended completion dates for some reactors [108]. Consequently, many similar Babcock & Wilcox reactors that had been ordered were canceled. In total, 52 U.S. nuclear reactors were canceled between 1980 and 1984 [109].

The accident did not initiate the decline of the U.S. nuclear power industry but did halt its historical growth. Furthermore, in response to the earlier 1973 oil crisis and subsequent analyses suggesting potential overcapacity in base load generation, 40 planned nuclear power plants had already been canceled before the accident. At the time of the incident, 129 nuclear power plants had been approved, but only 53 of those not yet operating were completed. The protracted review process, further complicated by the Chernobyl disaster seven years later, led to more stringent federal requirements for safety improvements and design modifications. Local opposition intensified, construction times significantly lengthened, and costs escalated dramatically [110]. Until 2012, no new U.S. nuclear power plant had been authorized for construction since 1978. Globally, the increase in nuclear power plant construction ended more dramatically with the Chernobyl disaster in 1986 (see graph).

Cleanup

Initially, GPU planned to repair the reactor and return it to service [112]. However, TMI-2 was too severely damaged and contaminated to resume operations; the reactor was gradually deactivated and permanently closed. TMI-2 had been operational for only three months but now contained a ruined reactor vessel and a containment building unsafe for occupancy. Cleanup began in August 1979 and officially concluded in December 1993, with a total cleanup cost of approximately $1 billion [19]. Benjamin K. Sovacool, in his 2007 preliminary assessment of major energy accidents, estimated the total property damages from the TMI accident at$2.4 billion [113].

Efforts focused on the cleanup and decontamination of the site, particularly the defueling of the damaged reactor. Starting in 1985, nearly 100 short tons (91 t) of radioactive fuel were removed from the site. Planning and work were partly hampered by overly optimistic perceptions of the damage [114].

In 1988, the NRC announced that while further decontamination of Unit 2 was possible, the remaining radioactivity had been sufficiently contained to pose no threat to public health and safety. The initial major phase of cleanup concluded in 1990, with workers shipping 150 short tons (140 t) of radioactive wreckage to Idaho for storage at the Department of Energy's National Engineering Laboratory. However, the contaminated cooling water that leaked into the containment building had seeped into the concrete, leaving residual radioactivity too impractical to remove. Consequently, further cleanup efforts were deferred to allow for radiation decay and to consider the potential economic benefits of retiring both Unit 1 and Unit 2 together [19].

Health effects and epidemiology

Main article: Three Mile Island accident health effects

In the aftermath of the accident, investigations focused on quantifying the radioactivity released. In total, approximately 2.5 megacuries (93 PBq) of radioactive gases and approximately 15 curies (560 GBq) of iodine-131 were released into the environment [115]. According to the American Nuclear Society, based on official radioactivity emission figures, "The average radiation dose to people living within 10 miles of the plant was eight millirem (0.08 mSv), and no more than 100 millirem (1 mSv) to any single individual. Eight millirem is about equal to a chest X-ray, and 100 millirem is about a third of the average background level of radiation received by US residents in a year" [116].

According to health researcher Joseph Mangano, early scientific publications estimated no additional cancer deaths in the 10-mile (16 km) area around TMI, based on these figures. Disease rates in areas farther than 10 miles from the plant were not examined [88]. Local activism in the 1980s, driven by anecdotal reports of negative health effects, prompted the commissioning of scientific studies. Several epidemiological studies have concluded that the accident had no observable long-term health effects [d].

A peer-reviewed research article by Dr. Steven Wing found a significant increase in cancers between 1979 and 1985 among individuals residing within ten miles of TMI [119]. In 2009, Dr. Wing stated that radiation releases during the accident were likely "thousands of times greater" than the NRC's estimates. A retrospective analysis of the Pennsylvania Cancer Registry identified an increased incidence of thyroid cancer in some counties south of TMI (though not in Dauphin County, where the reactor was located) and in high-risk age groups, but did not establish a causal link to the accident [13][14]. The Talbott lab at the University of Pittsburgh reported finding a few small increases in cancer risks within the TMI population [15]. A more recent study reported "findings consistent with observations from other radiation-exposed populations," raising "the possibility that radiation released from [Three Mile Island] may have altered the molecular profile of [thyroid cancer] in the population surrounding TMI," establishing a potential causal mechanism, though not definitively proving causation [120].

The Radiation and Public Health Project, an organization with limited credibility among epidemiologists [121], cited calculations by Mangano indicating a spike in infant mortality in downwind communities two years after the accident [88][122]. Anecdotal evidence also records effects on the region's wildlife [88]. John Gofman employed his own, non-peer reviewed low-level radiation health model to predict 333 excess cancer or leukemia deaths resulting from the 1979 Three Mile Island accident [11]. The ongoing TMI epidemiological research has been accompanied by discussions regarding challenges in dose estimation due to a lack of precise data and variations in illness classifications [123].

Activism and legal action

Anti-nuclear protests took place in Harrisburg, Pennsylvania, the state capital, on April 9, 1979, following the Three Mile Island accident. The accident significantly enhanced the perceived credibility of anti-nuclear groups and triggered protests worldwide [124][125]. President Carter, who had a background in nuclear power during his United States Navy service, informed his cabinet after visiting the plant that the accident was minor but reportedly avoided making this public to avoid alienating Democrats who opposed nuclear power [126].

In the months following the accident, members of the American public, concerned about the release of radioactive gas, organized numerous anti-nuclear demonstrations across the country. The largest demonstration occurred in New York City in September 1979, drawing 200,000 people, with speeches by Jane Fonda and Ralph Nader [127][128][129]. This rally was part of a series of nightly "No Nukes" concerts held at Madison Square Garden from September 19 to 23 by Musicians United for Safe Energy. In May of that year, an estimated 65,000 people, including California Governor Jerry Brown, attended a march and rally against nuclear power in Washington, D.C. [128].

In 1981, citizen groups successfully pursued a class-action lawsuit against TMI, winning a $25 million out-of-court settlement. A portion of these funds was used to establish the TMI Public Health Fund [130]. In 1983, a federal grand jury indicted Metropolitan Edison on criminal charges for falsifying safety test results prior to the accident [131]. Under a plea-bargaining agreement, Met Ed pleaded guilty to one count of falsifying records and no contest to six other charges, with four counts dropped. The company agreed to pay a$45,000 fine and establish a $1 million account to assist with emergency planning in the area surrounding the plant [132].

According to Eric Epstein, chair of Three Mile Island Alert, the TMI plant operator and its insurers paid at least $82 million in publicly documented compensation to residents for "loss of business revenue, evacuation expenses and health claims" [133]. However, a class action lawsuit alleging detrimental health effects caused by the accident was rejected by Harrisburg United States district court Judge Sylvia Rambo, with the final appeal of this decision failing in 2002 [134][135].

Normal accident theory

The Three Mile Island accident inspired Charles Perrow's normal accident theory, which seeks to explain "unanticipated interactions of multiple failures in a complex system." TMI exemplified this type of accident, being "unexpected, incomprehensible, uncontrollable and unavoidable" [136].

Perrow concluded that the Three Mile Island failure was a consequence of the system's profound complexity. He argued that such modern, high-risk systems are inherently prone to failures, irrespective of management practices, making what he termed a 'normal accident' inevitable. He suggested that a radical redesign or, if that proved unfeasible, abandonment of such technology might be more prudent [137].

"Normal" accidents, or system accidents, are so named by Perrow because they are considered inevitable in highly complex systems. Given the nature of these systems, multiple interacting failures are bound to occur despite efforts to prevent them [138]. Events that initially appear minor can cascade and multiply unpredictably, leading to catastrophic outcomes [139].

The theory of Normal Accidents contributed key concepts to a body of thought in the 1980s that revolutionized the understanding of safety and risk. It advocated for analyzing technological failures as products of highly interacting systems, highlighting organizational and managerial factors as primary causes. Technological disasters could no longer be attributed solely to isolated equipment malfunctions, operator errors, or unforeseen natural events [137].

Comparison to U.S. Navy operations

Following the TMI incident, President Carter commissioned the "Report of the President's Commission on the Accident at Three Mile Island" (1979) [92].

Admiral Hyman G. Rickover was later asked to explain to Congress why naval nuclear propulsion (as used in submarines) had not experienced any reactor accidents, defined as the uncontrolled release of fission products resulting from core damage. In his testimony, Rickover stated:

Over the years, many people have asked me how I run the Naval Reactors Program, so that they might find some benefit for their own work. I am always chagrined at the tendency of people to expect that I have a simple, easy gimmick that makes my program function. Any successful program functions as an integrated whole of many factors. Trying to select one aspect as the key one will not work. Each element depends on all the others [140].

21st-century status

Following the TMI-2 incident, the NRC suspended the operating license for TMI-1, owned and operated by Met Ed, a subsidiary of General Public Utilities Corporation. In 1982, residents of the three surrounding counties voted overwhelmingly in a non-binding resolution to permanently retire Unit 1 [141][142]. In 1985, the Nuclear Regulatory Commission, by a 4–1 vote, authorized TMI-1 to resume operations [141][142].

GPU formed General Public Utilities Nuclear Corporation as a subsidiary to own and operate the company's nuclear facilities, including Three Mile Island. In 1996, General Public Utilities was renamed GPU Inc. In 1998, GPU sold TMI-1 to AmerGen Energy Corporation, a joint venture between Philadelphia Electric Company and British Energy. (GPU remained legally obligated to maintain and monitor TMI-2.) In 2001, GPU was acquired by FirstEnergy Corporation and dissolved, with the maintenance and administration of Unit 2 contracted to AmerGen.

In 2000, Philadelphia Electric merged with Unicom Corporation to form Exelon. By 2003, Exelon had acquired the remaining AmerGen shares from British Energy. In 2009, Exelon Nuclear absorbed and dissolved AmerGen. In addition to TMI Unit 1, Exelon Nuclear operates the Clinton Power Station and several other nuclear facilities [e].

Unit 2 continues to be licensed and regulated by the Nuclear Regulatory Commission under a status known as Post Defueling Monitored Storage [148]. The TMI-2 reactor has been permanently shut down, with the reactor coolant system drained, the radioactive water decontaminated and evaporated, radioactive waste shipped off-site, and the reactor fuel and most core debris transported to a Department of Energy facility. The remainder of the site is under ongoing monitoring. The owner planned to maintain the facility in long-term monitored storage until the expiration of the TMI-1 operating license, at which point both plants would undergo decommissioning [22].

In 2009, the NRC granted a license extension, allowing the TMI-1 reactor to operate until April 19, 2034 [149][150]. By 2017, it was announced that operations would cease by 2019 due to financial pressure from inexpensive natural gas, unless lawmakers intervened to keep it operational [151]. When it became clear that subsidy legislation would not pass, Exelon decided to retire the plant [152]. TMI Unit 1 shut down on September 20, 2019, entering decommissioning and moving to SAFSTOR status [153][154].

In 2020, the site was purchased by TMI-2 Solutions, a subsidiary of EnergySolutions, with the goal of cleaning up the site at a lower cost than the available dedicated fund [155]. On May 8, 2023, TMI-2 Solutions announced that 99% of the nuclear fuel had been removed and that the site had entered the next phase of cleanup, scheduled to last until 2029 [156][155]. TMI-2 Solutions plans to complete the cleanup and demolish the plant by 2052 [155].

In September 2024, Constellation Energy announced plans to restart the Three Mile Island nuclear plant to sell power to Microsoft, highlighting the significant energy demands of the tech sector as they build data centers to support artificial intelligence [157]. Constellation expects the Unit 1 reactor at Three Mile Island to return to service in 2028, subject to approval by the Nuclear Regulatory Commission. Constellation also plans to seek an extension of the plant's operational lifespan to at least 2054 [158].

Timeline

| Date | Event