info on Three Mile Island:

http://www.enviroweb.org/tmia/
For two years before the accident, our concerned citizens group warned the Nuclear Regulatory Commission and Pennsylvanians that Three Mile Island Reactor Unit 2 was dangerously faulty. Our critics said we were "fear mongers" and that our warnings were "tantamount to yelling fire in a crowded theater."
On March 28, 1979 Pennsylvania and nearby states were threatened by a "nuclear nightmare" as described later that evening by Walter Cronkite on the CBS Evening News. TMI spokesmen were claiming that only a minor problem had occurred and that the reactor would be back on line soon. But in actuality, The reactor was out of control, the core was uncovered by the leaking coolant, fuel was melting and collapsing to the bottom of the reactor, the control room operators were bewildered, evacuation plans were pushed aside, and... We came within 30 minutes of an irreversible meltdown  

http://www.uilondon.org/safety/tmi.htm

IN THE EARLY HOURS of the morning of Wednesday 28 March 1979, the two water feed pumps in the secondary cooling circuit of unit 2 at the Three Mile Island (TMI) nuclear power station stopped operating.
While the reactor responded appropriately to this occurrence, shortcomings in training and in the man/machine interface meant that the operators did not.
Had the reactor operators been in a position to respond appropriately, the incident would have been little more than a notation in the reactor operations log book. Instead TMI-2 was destroyed and clean up costs around one billion dollars were incurred.
Most importantly, no one was killed and there were no serious injuries as a result of the accident. Since 1979 there have been more than 12 independent studies. None of them have found any radiation induced health effects in the population around TMI.
PRE-ACCIDENT CONDITIONS
Three Mile Island was a nuclear power station with two pressurised water reactors. Unit 1 is an 800 MWe unit which entered service in 1974 and remains in service today, unit 2 was a 900 MWe reactor operating since 1978. The station, located near Harrisburg, Pennsylvania, USA, was operated by Metropolitan Edison (Met Ed).
During the night of 27-28 March 1979 unit 2 was operating at 97% power. Maintenance work was being carried out on one of the machines, called a polisher, located in the non-nuclear part of the plant, that was used to remove dissolved minerals from the feedwater. A crew was using a mixture of air and water to dislodge a build up of resin from a transfer line.
Causes of the accident
The TMI-2 accident involved a small leak of water from the reactor system that wasn’t correctly diagnosed. Inadequate control room instrumentation and emergency response training proved to be root causes of the operators’ inability to respond properly to an unplanned automatic shutdown of the reactor at 4am on 28 March 1979.
This factsheet presents a simplified sequence of the events on 28 March 1979 at the Three Mile Island nuclear power station. Not all events are recorded. There are some discrepancies in the indicated times of events in different reports.
website also gives: sequence of events; subsequent events; cleanup; public reaction...  

http://www.nrc.gov/OPA/gmo/tip/tmi.htm
The accident at the Three Mile Island Unit 2 (TMI-2) nuclear power plant near Middletown, Pennsylvania, on March 28, 1979, was the most serious in U.S. commercial nuclear power plant operating history(1), even though it led to no deaths or injuries to plant workers or members of the nearby community. But it brought about sweeping changes involving emergency response planning, reactor operator training, human factors engineering, radiation protection, and many other areas of nuclear power plant operations. It also caused the U.S. Nuclear Regulatory Commission to tighten and heighten its regulatory oversight. Resultant changes in the nuclear power industry and at the NRC had the effect of enhancing safety.
The sequence of certain events - - equipment malfunctions, design related problems and worker errors - - led to significant damage to the TMI-2 reactor core but only very small off-site releases of radioactivity.
Summary of Events
The accident began about 4:00 a.m. on March 28, 1979, when the plant experienced a failure in the secondary, non-nuclear section of the plant. The main feedwater pumps stopped running, caused by either a mechanical or electrical failure, which prevented the steam generators from removing heat. First the turbine, then the reactor automatically shut down. Immediately, the pressure in the primary system (the nuclear portion of the plant) began to increase. In order to prevent that pressure from becoming excessive, the pressurizer relief valve (a valve located at the top of the pressurizer) opened. The valve should have closed when the pressure decreased by a certain amount, but it did not. Signals available to the operator failed to show that the valve was still open. As a result, the stuck-open valve caused the pressure to continue to decrease in the system.
Meanwhile, another problem appeared elsewhere in the plant. The emergency feedwater system (backup to main feedwater) was tested 42 hours prior to the accident. As part of the test, a valve is closed and then reopened at the end of the test. But this time, through either an administrative or human error, the valve was not reopened - - preventing the emergency feedwater system from functioning. The valve was discovered closed about eight minutes into the accident. Once it was reopened, the emergency feedwater system began to work correctly, allowing cooling water to flow into the steam generators.
As the system pressure in the primary system continued to decrease, voids (areas where no water is present) began to form in portions of the system other than the pressurizer. Because of these voids, the water in the system was redistributed and the pressurizer became full of water. The level indicator, which tells the operator the amount of coolant capable of heat removal, incorrectly indicated the system was full of water. Thus, the operator stopped adding water. He was unaware that, because of the stuck valve, the indicator can, and in this instance did, provide false readings.
Because adequate cooling was not available, the nuclear fuel overheated to the point where some of the zirconium cladding (the long metal tubes or jackets which hold the nuclear fuel pellets) reacted with the water and generated hydrogen. This hydrogen was released into the reactor containment building. By March 30, two days after the start of the chain of events, some hydrogen remained within the primary coolant system in the vessel surrounding the reactor, forming a "hydrogen bubble" above the reactor core.
The concern was that if reactor pressure decreased, the hydrogen bubble would expand and thus interfere with the flow of cooling water through the core. Over the next few days, the bubble was reduced by "degassing" the pressurizer -- adjusting air and water pressure.
Without water to cool it, and with the top of the reactor core uncovered, the primary damage to the reactor occurred two to three hours into the accident. Although no "meltdown" occurred in the classic sense of the word, in that fuel did not "melt" through the floor beneath the containment or through the steel reactor vessel, a significant amount of fuel did in fact melt. Radioactivity in the reactor coolant increased dramatically, and there were small leaks in the reactor coolant system which caused high radiation levels in other parts of the plant and small releases into the environment. Shortly after the accident began, some of the water, carrying fuel debris and fission products, escaped from the reactor coolant system and flowed into the reactor building basement. By the time the accident had ended, the water in the basement had been heated by residual heat from the reactor vessel, evaporated, condensed on the walls, and drained down onto the floors and back into the basement. The radionuclides then permeated into the porous surfaces of concrete and layers of iron which later became corroded (this area of the plant became a major focus of the subsequent clean-up and decontamination).
Response to the accident was swift. The NRC's regional office in King of Prussia, Pennsylvania, was notified at 7:45 a.m. on March 28. By 8:00, the NRC headquarters in Washington, D.C. was alerted and the NRC Operations Center in Bethesda, Maryland, was activated. The regional office promptly dispatched the first team of inspectors to the site and other agencies, such as the Department of Energy, and the Environmental Protection Agency, also mobilized their response teams. Helicopters hired by TMI's owner, General Public Utilities Nuclear, and the Department of Energy were sampling radioactivity in the atmosphere above the plant by midday. A team from the Brookhaven National Laboratory was also sent to assist in radiation monitoring. At 9:15 a.m., the White House was notified and at 11:00 a.m., all non-essential personnel were ordered off the plant's premises.
From the early stages of the accident, low levels of radioactive gas, mostly in the form of xenon, continued to be released to the environment. At the time, efforts to halt the releases were unsuccessful and there was some fear of an explosion from the buildup of hydrogen - -fortunately, this did not occur. However, on Friday, March 30, Governor Thornburgh of Pennsylvania ordered a precautionary evacuation of preschool children and pregnant women from within the 5-mile zone nearest the plant, and suggested that people living within 10 miles of the plant stay inside and keep their windows closed. Most evacuees had returned to their homes by April 4 -- by that time, the situation at the reactor had been brought under control.
The American Nuclear Insurers, an organization made up of nuclear insurance firms, had already begun distributing checks to evacuees to cover hotel and meal expenses, and was beginning to handle claims for property and liability losses.
Health Effects
Detailed studies of the radiological consequences of the accident have been conducted by the NRC, the Environmental Protection Agency, the Department of Health, Education and Welfare (now Health and Human Services) , the Department of Energy, and the State of Pennsylvania. Several independent studies have also been conducted. Estimates are that the average dose to about 2 million people in the area was about only about 1 millirem. To put this into context, exposure from a full set of chest x-rays is about 6 millirem. Compared to the natural radioactive background dose of about 100-125 millirem per year for the area, the collective dose to the community from the accident was very small. The maximum dose to a person at the site boundary would have been less than 100 millirem.
In the months following the accident, although questions were raised about possible adverse effects from radiation on human, animal, and plant life in the TMI area, none could be directly correlated to the accident. Thousands of environmental samples of air, water, milk, vegetation, soil, and foodstuffs were collected by various groups monitoring the area. Very low levels of radionuclides could be attributed to releases from the accident. However, comprehensive investigations and assessments by several well-respected organizations have concluded that in spite of serious damage to the reactor, most of the radiation was contained and that the actual release had negligible effects on the physical health of individuals or the environment.
Impact of the Accident
Today, the TMI-2 reactor is permanently shut down and defueled, with the reactor coolant system decontaminated, the radioactive liquids treated, most components shipped to a licensed low-level waste disposal site, with the remainder of the site being monitored. The owner, General Public Utilities Nuclear Corporation, says it will keep the facility in long-term storage until the operating license for the TMI-1 plant expires in 2014, at which time both plants will be decommissioned.
The causes of the accident continue to be debated to this day. However, based on a series of investigations, the main factors appear to have been a combination of personnel error, design deficiencies, and component failures. There is no doubt that the accident at Three Mile Island permanently changed both the nuclear industry and the NRC. Public fear and distrust increased, NRC's regulations and oversight became broader and more robust, and management of the plants was scrutinized more carefully. The problems identified from careful analysis of the events during those days have led to permanent and sweeping changes in how NRC regulates its licensees - - which, in turn, has strengthened public health and safety.
Here are some of the major changes which have occurred since the accident:
Expansion of NRC's resident inspector program - first authorized in 1977 - whereby at least two inspectors live nearby and work exclusively at each plant in the U.S to provide daily surveillance of licensee adherence to NRC regulations; Establishment of the Systematic Assessment of Licensee Performance (SALP) program to integrate NRC observations, findings, and conclusions about licensee performance and management effectiveness into a periodic, public report; Regular analysis of plant performance by senior NRC managers who identify those plants needing additional regulatory attention; Expansion of performance-oriented as well as safety-oriented inspections, and the use of risk assessment to identify vulnerabilities of any plant to severe accidents; Strengthening and reorganization of enforcement as a separate office within the NRC; Upgrading and strengthening of plant design and equipment requirements. This includes fire protection, piping systems, auxiliary feedwater systems, containment building isolation, reliability of individual components (pressure relief valves and electrical circuit breakers), and the ability of plants to shut down automatically; Identifying human performance as a critical part of plant safety, revamping operator training and staffing requirements, followed by improved instrumentation and controls for operating the plant, and establishment of fitness for duty programs for plant workers to guard against alcohol or drug abuse; Enhancement of emergency preparedness to include immediate NRC notification requirements for plant events and an NRC operations center which is now staffed 24 hours a day. Drills and response plans are now tested by licensees several times a year, and state and local agencies participate in drills with the Federal Emergency Management Agency and NRC; The installing of additional equipment by licensees to mitigate accident conditions, and monitor radiation levels and plant status; The establishment of the Institute of Nuclear Power Operations (INPO), the industry's own "policing" group, and formation of what is now the Nuclear Energy Institute to provide a unified industry approach to generic nuclear regulatory issues, and interaction with NRC and other government agencies; Employment of major initiatives by licensees in early identification of important safety-related problems, and in collecting and assessing relevant data so lessons of experience can be shared and quickly acted upon; Expansion of NRC's international activities to share enhanced knowledge of nuclear safety with other countries in a number of important technical areas.
also has glossary of terms and diagrams  

http://www.cannon.net/~gonyeau/nuclear/tmi.htm In March 1979, an event occurred at the Three Mile Island Unit 2 that resulted in the first case of melted fuel in a full scale commerical nuclear power plant. There had been prior cases of small scale fuel melting, e.g. the Fermi 1 reactor near Monroe, Michigan. TMI-2 was a Babcock & Wilcox unit with a vertical once-through steam generator. In the event a valve in the secondary system closed and initiated the sequence of events. Two diagrams are being used to illustrate the events that occurred. The first shows the overall cycle for the pressurized water reactor design. The second focuses on the containment.
The sequence of events was:
1. A valve in the condensate system (between the condenser and the pump on the secondary side) failed closed, which reduced the amount of water being supplied to the steam generator; the main feedwater pumps and the turbine tripped within seconds.
2. The design of the vertical one-through steam generator is such that there is not much water on the secondary (non- radioactive) outer side of the steam generator tubes that will boil to steam when the plant is at full power and the reactor continues to put out full power; thus all the water on the secondary side was rapidly converted to steam within minutes. The emergency feedwater pumps, which started as expected, were unable to inject water into the steam generators because several valves in the system were closed.
3. The reactor continued to heat the reactor coolant. The reactor coolant pumps continued circulating the water to the steam generators, however no heat could be removed by the secondary side since there was no water in the steam generators. The reactor coolant system started to heat up.
4. Pressure rose in the reactor cooling system until the reactor shutdown. A power operated relief valve opened in the line between the pressurizer and the quench tank.. This valve failed to reclose when it was supposed to - after pressure dropped below the setpoint for closure. This relief valve continued to discharge to the quench tank. The fact that the valve was open allowed steam to continue discharging to the quench tank. Pressure dropped in the reactor cooling system because the valve was still open (however, due to poor control board design and a failure to indicate the valve position properly, the operators did not know the valve was open). The quench tank has a rupture disc that opens at about 10-12 pounds per square inch. When this happened, the steam was released to the containment.
5. The pressurizer is normally at about 650F. As pressure dropped in the reactor cooling system, eventually water in the upper-most area of the reactor (about 10-15 feet above the fuel) flashed to steam. The indicated water level in the pressurizer stayed high (the relationship between the pressurizer and the reactor was like a manometer).
6. The operators turned off the emergency water injection pumps because they thought there was still water in the pressurizer.
7. The operators turned off the reactor cooling pumps because they were concerned about damage due to potential excessive vibration.. This resulted in a steam void forming in the reactor coolant loop. In addition, a steam bubble formed in the upper part of the reactor above the fuel. Eventually as the fuel heated, this void expanded. Eventually, the fuel cladding material overheated. It is likely that some hydrogen was produced by a chemical reaction between the zircaloy clad and the steam in the reactor. In addition, the hydrogen normally present in the reactor cooling system (used to reduce the presence of oxygen and subsequent corrosion in the system) was released to the containment through
8. At one point, containment pressure rapidly spiked to 28 pounds per square inch; then rapidly dropped. This was most likely due to the chemical reaction of hydrogen with the oxygen in the containment.
9. Water was added to the reactor cooling system and the level raised in the pressurizer until cooling of the reactor was assured.
What was the maximum radiation exposure offsite?
Studies conducted indicate that the maximum potential offsite radiation exposure likely was 83 millirem. An actual individual located on a nearby island is believed to have received at most 37 millirem. Extensive studies by federal agencies led to these conclusions and to an estimate that one excess cancer fatality due to the accident could be expected over a 30 year period. What Good came from the TMI Event?
Several thorough investigations occurred - the most important was that conducted by the Kemeny Commission which was appointed by President Carter and resulted in a number of recommendations. Improvements were needed in the follwing areas:
Operator training
Emergency planning
Dissemination of industry information Use of probabilistic safety assessment and analysis of more probable events. The electric utlities recognized their responsibilities. An industry self-assessment group was formed - the Institute of Nuclear Power Operations. This organization, based in Atlanta, serves several functions:
Evaluates events and practices within the US nuclear industry and disseminates recommendations Conducts periodic assessments of each utility in the United States, including operations, maintenance, engineering, training, radiation protection, chemistry, and corporate support; the results of these inspections factor into the insurance ratings of the utility. Provides highly specialized training programs for utility personnel, including plant managers. All electric utilities expanded significantly the training conducted for personnel who work at and support nuclear plant operations. This included establishing the National Nuclear Academy which accredits the plant training programs in 10 areas. Also, all utilities purchased simulators for training personnel who work in the main control room.
The NRC also took decisive action imposing a number of changes related to equipment, analysis, practices, and personnel. The NRC conducted their own self examination referred to as the Rogovin study. Two major documents were issued - NUREG-0696 and NUREG-0737. Equipment changes included monitoring instrumentation capable of withstanding severe accidents and hydrogen recombiners. Analysis involved small break loss of coolant events. Practices changed included upgrading of emergency operating procedures and development of emergency plans. Personnel-related changes involved upgrading of training and qualification requirements and a requirement to have a degreed shift technical advisor assigned to each shift to evaluate abnormal conditions.
The Federal Emergency Management Agency, in conjunction with the NRC, developed criteria for classifying emergencies, emergency planning, and evacuation plans.
also has diagrams