The TMI 2 Accident: Its Impact, Its Lessons

Key Facts

The accident at Three Mile Island Unit 2 in 1979 was caused by a combination of equipment failure and the inability of plant operators to understand the reactor’s condition. A gradual loss of cooling water to the reactor’s heat-producing core led to partial melting of the fuel rod cladding and the uranium fuel, and the release of a small amount of radioactive material.

The TMI 2 accident caused no injuries.  In addition, experts concluded that the amount of radiation released into the atmosphere was too small to result in discernible direct health effects to the population in the vicinity of the plant. At least a dozen epidemiological studies conducted since 1981 have borne this out.

Both the industry and the government responded swiftly and decisively to the TMI accident. Among other actions, the industry formed a center to study the accident and established the Institute of Nuclear Power Operations to promote excellence in operator training, plant management and operation. President Jimmy Carter appointed the Kemeny Commission to investigate the accident, and the Nuclear Regulatory Commission undertook its own study, which led to a post-TMI action plan for utilities as well as the issuance of many additional regulations.

Since the accident, the industry has not only complied with the recommendations of the Kemeny Commission but also transformed the way it manages and operates plants. As a result, U.S. nuclear plant performance has steadily improved, with plants operating more safely, more efficiently and more reliably from year to year.

After being defueled, TMI 2 was put into monitored storage in December 1993.  It is expected that TMI-2 will be decommissioned with when TMI 1 reaches the end of its service life.

The Accident: What Happened

On March 28, 1979, a combination of equipment failure and the inability of plant operators to understand the reactor’s condition culminated in the accident at Three Mile Island Unit 2, near Harrisburg, Pa.

Unit 2 was operating at about 100 percent power when the plant automatically shut down after a pump that provided cooling water stopped operating. Pressure and temperature increased in the reactor, causing a pressure relief valve to open. The valve opened as designed, and water and steam began flowing out of the reactor to a tank in the basement of the reactor building. As pressure returned to normal, the valve should have closed. But, unknown to the operators, the valve stuck open. It remained open for more than two hours, allowing water that covered and cooled the fuel core to escape from the reactor system.

Instrumentation in the TMI control room, however, indicated to the operators that the relief valve was closed and that too much water was being injected into the reactor vessel. Consequently, operators did not replace the water that was being lost as a result of the open relief valve. As pressure continued to drop, and more and more coolant turned to steam, the main coolant pumps began to shake violently because they were not designed to pump a mixture of water and steam.

Concerned about damaging the pumps and not realizing what was actually happening in the plant, the operators shut down the pumps. The loss of pressure and water caused a large steam bubble to form inside the reactor vessel, further preventing the flow of coolant through the core. Without coolant, core temperatures rose above the melting point of the fuel cladding and the uranium fuel. About half of the fuel melted before the bubble was dispersed and the coolant flow was restored. The colder cooling water also shattered some of the hot fuel rods. All the fuel was damaged.

As a result of the TMI 2 accident, 700,000 gallons of radioactive cooling water ended up in the basement of the reactor building and in tanks in the auxiliary building, contaminating them. In addition, a small amount of radioactive material was released into the atmosphere through the plant’s ventilation stack.

Public Health and Safety: No Injuries, No Measurable Effect

The TMI accident caused no injuries, and at least a dozen epidemiological studies conducted since 1981 have found no discernible direct health effects to the population in the vicinity of the plant.

Several studies were conducted by the Pennsylvania Department of Health. The conclusions are summarized below:

• A 1981 study reported that if the accident had had any effect on infant death rates, there would have been a significant increase in the six months after the accident. Instead, the infant death rate was lower than normal.
• A 1982 study found that the incidence of congenital hypothyroidism within a 10-mile radius of the plant was well within a normal range in the year after the accident.
• A 1982 study found no measurable impact on infant mortality within a 10-mile radius of the plant, compared with infant mortality rates for all Pennsylvania for 1977-1979.
• A 1985 study found no significant difference in cancer mortality within a 20-mile radius of the plant during the five years preceding the accident and the five years following it. In a more detailed analysis of four communities downwind
  of the plant, the study found no significant abnormalities in either cancer mortality or cancer incidence among residents considered to be at potentially higher risk.
• A 1988 study found no connection between radiation or psychological stress and such adverse pregnancy outcomes as fetal and neonatal mortalities and other problems.
• A 1989 study found no significant abnormalities in cancer mortality or incidence among residents of selected communities near the plant.
• Two 1991 studies showed no increased cancer incidence among people who lived near the plant in 1979. One study involved the general population living within a five-mile radius of the plant; the other involved women of child-bearing age who lived within a 10-mile radius.

Although the studies have found no increased incidence of cancer as a result of the accident, they did find evidence of psychological stress, lasting in some cases for five to six years. According to Dr. Jane Bratz, chief of the Pennsylvania Department of Health’s Three Mile Island Health Research Program, the people suffering from stress perceived their health as being poorer than it actually was when the Health Department checked the medical records.

In addition to the Pennsylvania Health Department studies, at least three other studies have examined the health impact of the TMI accident on the population:

A study presented by K. Ramaswamy at the 1988 annual meeting of the American Public Health Association compared post-accident cancer deaths over a six-year period for residents within a five-mile radius of the plant with cancer deaths of a large control population. The study concluded that the normal death rate and life expectancy for people around TMI were not affected by the accident. A second study, presented by E. Digon at the 1988 annual meeting of the American Public Health Association, concluded that—based on a comprehensive analysis of statistical data by health researchers—fetal and infant mortality in the vicinity of the plant was neither significantly higher than expected nor significantly different from that in the years before the accident.

Dr. Maureen Hatch (associate professor of epidemiology at Columbia University’s School of Public Health), Dr. Mervyn Susser (Gertrude H. Sergievsky Professor of Epidemiology, also at Columbia) and Dr. Jan Beyea (a nuclear physicist in the New York office of the National Audubon Society) studied cancers among the nearly 160,000 residents within a 10-mile radius of the TMI plant. The principal cancers considered were leukemia and childhood malignancies. The study, issued in September 1990, concluded: “Overall, the pattern of results does not provide convincing evidence that radiation releases from the Three Mile Island nuclear facility influenced cancer risk during the limited period of follow-up.”

 In 1990, the National Cancer Institute of the National Institutes of Health released the results of a two-year study of cancer data in 107 U.S. counties that contained, or were adjacent to, major nuclear facilities that had started up before 1982. Among the counties were York, Lancaster and Dauphin near the TMI plant in Pennsylvania. The study, which compared cancer mortality rates in the 107 counties with rates in counties with no nuclear facilities, found no increased cancer mortality for people living near the nuclear installations. The study also found no evidence that leukemia for any age was linked to routine operations at TMI or to the accident at TMI 2.

Swift, Decisive Action by Industry, Government

In the weeks and months after the TMI 2 accident, the nuclear industry and the federal government explored the causes of the accident and took steps to ensure it would not happen again. In the years that followed, actions by the industry and the Nuclear Regulatory Commission (NRC) changed the face of nuclear energy in the United States.

Kemeny Commission’s Report Lauded

Two weeks after the accident, President Jimmy Carter appointed a 12-member commission, headed by the late John Kemeny, then president of Dartmouth College, to investigate what had happened and its possible impact on the health and safety of the public and plant personnel.

The commission’s report on its investigation, issued in October 1979, contained several recommendations. Among them:

• The industry must set and police its own standards of
  excellence.
• Within each utility, a single organization with the requisite expertise must be accountable for nuclear operations.
• Every nuclear plant must systematically gather, review and analyze operating experience.
• Agency-accredited training institutions must be established for nuclear plant operators and immediate supervisors of operations.
• The nuclear industry must dramatically change its attitudes toward safety and regulations.

The industry, the NRC and the White House all praised the report, endorsing the recommendations.

NRC Requirements Proliferate

The NRC also moved quickly, setting up a group to study the accident. The group, headed by attorney Mitchell Rogovin, reached many of the same conclusions as the Kemeny Commission. In addition, the NRC placed a moratorium on issuing plant construction permits or operating licenses.

About nine months after the accident, the NRC proposed a post-TMI 2 action plan for utilities. The plan was criticized within the NRC for failing to make a distinction between important and unimportant tasks, and for duplication of tasks. It was also criticized for imposing costly requirements on utilities. While changes were clearly needed in the way the industry managed and operated its nuclear plants, the NRC plan imposed a regulatory burden on utilities that drove up costs and did little to enhance safety.

Industry Changes Increase Safety

Within weeks of the accident, industry executives asked the Electric Power Research Institute to study the accident. EPRI responded by forming the Nuclear Safety Analysis Center (NSAC), which provided information to the Kemeny Commission, the NRC’s investigative group and congressional staff studying the accident.

NSAC acted as a clearing-house for information on plant operations and as a source of recommendations on improving reactor safety. It set up two programs to develop suggestions on safety-related improvements, which it then communicated to utilities. In addition, NSAC was responsible for responding to NRC actions of a generic nature, commenting on new regulations, criteria and standards.

Within nine months of the accident, the industry had formed the Institute of Nuclear Power Operations (INPO), whose mission was to promote the highest levels of safety and reliability in the operation of nuclear power plants.

INPO developed an assortment of tools to do the job:

• Performance objectives and criteria for overall nuclear plant operations and for corporate management support of nuclear programs. These documents set the benchmarks for excellence in safety and reliability and are the basis for INPO’s on-site evaluations of plant performance and corporate support.
• Guidelines and “Good Practice” documents that define one method of meeting INPO performance objectives in specific areas. The guidelines establish the basis for sound programs in selected areas of nuclear plant operation. The “Good Practice” documents identify techniques, programs or processes that have been proven particularly effective at one or more nuclear plants.
• Regular evaluations of nuclear plants and periodic evaluations of corporate support of operating plants. After each evaluation, INPO supplies the utility with a formal report, giving its findings and recommending improvements. Follow-up visits ensure the recommendations are implemented.
• Special assistance visits at the request of member utilities. These visits, conducted by qualified INPO personnel and industry peers, are designed to help utilities deal with specific technical issues.

INPO also developed several technical programs for gathering and analyzing information on day-to-day plant operations.

In 1985, INPO formed the National Academy for Nuclear Training, which reviews and accredits nuclear utilities’ training programs for all key positions at each plant.

A Transformed Nuclear Industry

INPO has had a profound impact on the way nuclear plants are managed and operated. The proof is the steady improvement in plant performance over the past 25 years.

INPO started monitoring performance indicators in 1981. Today, INPO compiles data on U.S. nuclear plants and generates industry averages for 10 World Association of Nuclear Operators performance indicators. It also generates industry goals, based on individual utility goals. Plants use the indicators to compare their performance with the industry as a whole. Some of the indicators are:

• Unit capability factor, which is the percentage of maximum energy generation
  a plant is capable of supplying. It has risen from 62.7 percent in 1980 to 88.7 percent in 1999.
• Unplanned automatic scrams, which are plant shutdowns caused by some imbalance in operations. The number has fallen from 7.3 per plant in 1980 to virtually zero in 1999 per 7,000 hours of operation (about one year).
• Industrial safety accident rates, which have been cut by more than 85 percent since 1980 and now number 0.34 industrial accidents per 200,000 work hours.

In March 1989, 10 years after the TMI accident, INPO issued a report on the industry’s responses to the Kemeny Commission recommendations. The report detailed what the industry had done to comply with the commission’s recommendations and noted that the industry “has demonstrably altered its posture toward nuclear safety and has embraced standards of excellence for nuclear plant operation.”

On the occasion of the 10th anniversary of the accident, John Kemeny, chairman of the commission, said of the INPO report: “[It] clearly shows that the Commission’s commendations have stood the test of time and have served as a catalyst for significant change.”

He added: “The [industry’s] improvements over the past decade have been impressive and are very reassuring.”

TMI 2 Today: In Monitored Storage

In October 1979, GPU Nuclear began cleaning up the damaged Unit 2. Between 1985 and 1990, GPU Nuclear removed some 100 metric tons of damaged uranium fuel and some 50 metric tons of damaged structural material from the reactor pressure vessel. To defuel the reactor vessel, workers used tools with handles up to 40 feet long, which they inserted through a slot in a shielded rotating work platform. They wore protective clothing and respirators to prevent contamination by radioactive particles. Radiation levels were about the same as those experienced during refueling at an operating plant.

While the fuel was being removed, the utility did much of the decontamination work. High-pressure water sprays were used to wash floors, walls and pipes. Air chisels and hydraulic pounders were used to break up the top layer of contaminated concrete. The evaporation of 2.23 million gallons of water generated by the accident was begun in 1991 and completed in 1993.

In its cleanup work, GPU Nuclear made extensive use of robots to measure radiation, carry out visual inspections, sample floors and walls, decontaminate walls and equipment, and remove debris.

In 1991, GPU Nuclear requested approval from the NRC to put the unit into monitored storage, which would have involved locking and monitoring it, but an individual opposed to this approach challenged the request.

In September 1992, the utility reached agreement with the individual, clearing the way for NRC approval of the request. As part of the agreement, GPU Nuclear provided nearly $900,000 over seven years to track and sponsor robotics research and use. It also has provided approximately $68,000 in funding over the 1992-1999 period to a local citizens’ group with a special interest in TMI 2 monitored storage and environmental monitoring.

              GPU Nuclear placed the unit in monitored storage December 28, 1993. In December 1999, GPU completed the sale of TMI 1 to AmerGen Energy Co., a joint venture of Exelon and British Energy Co. Under the terms of the sale, the ownership of TMI 2 will remain with GPU. In-plant and off-site monitoring of Unit 2 will continue until it is fully decommissioned, with regular reports made to the NRC, the Commonwealth of Pennsylvania and the public. AmerGen will assume full responsibility for decommissioning TMI 1. The two units will be decommissioned jointly when TMI 1 is taken out of service.  In 2001, First Energy merged with GPU and TMI-2 is now owned by First Energy Corp.