This article describes several significant radiation accidents, their causes and the responses to them. A review of the events leading up to, during and following these accidents can provide planners with information to preclude future occurrences of such accidents and to enhance an appropriate, rapid response in the event a similar accident occurs again.
Acute Radiation Death Resulting from an Accidental Nuclear Critical Excursion on 30 December 1958
This report is noteworthy because it involved the largest accidental dose of radiation received by humans (to date) and because of the extremely professional and thorough work-up of the case. This represents one of the best, if not the best, documented acute radiation syndrome descriptions that exists (JOM 1961).
At 4:35 p.m. on 30 December 1958, an accidental critical excursion resulting in fatal radiation injury to an employee (K) took place in the plutonium recovery plant at the Los Alamos National Laboratory (New Mexico, United States).
The time of the accident is important because six other workers had been in the same room with K thirty minutes earlier. The date of the accident is important because the normal flow of fissionable material into the system was interrupted for year-end physical inventory. This interruption caused a routine procedure to become non-routine and led to an accidental “criticality” of the plutonium-rich solids that were accidentally introduced into the system.
Summary of estimates of K’s radiation exposure
The best estimate of K’s average total-body exposure was between 39 and 49 Gy, of which about 9 Gy was due to fission neutrons. A considerably greater portion of the dose was delivered to the upper half of the body than to the lower half. Table 1 shows an estimate of K’s radiation exposure.
Table 1. Estimates of K’s radiation exposure
Region and conditions
Total body (average)
Clinical course of patient
In retrospect, the clinical course of patient K can be divided into four distinct periods. These periods differed in duration, symptoms and response to supportive therapy.
The first period, lasting from 20 to 30 minutes, was characterized by his immediate physical collapse and mental incapacitation. His condition progressed to semi-consciousness and severe prostration.
The second period lasted about 1.5 hours and began with his arrival by stretcher at the emergency room of the hospital and ended with his transfer from the emergency room to the ward for further supportive therapy. This interval was characterized by such severe cardiovascular shock that death seemed imminent during the whole time. He seemed to be suffering severe abdominal pain.
The third period was about 28 hours long and was characterized by enough subjective improvement to encourage continued attempts to alleviate his anoxia, hypotension and circulatory failure.
The fourth period began with the unheralded onset of rapidly increasing irritability and antagonism, bordering on mania, followed by coma and death in approximately 2 hours. The entire clinical course lasted 35 hours from the time of radiation exposure to death.
The most dramatic clinicopathological changes were observed in the haemopoietic and urinary systems. Lymphocytes were not found in the circulating blood after the eighth hour, and there was virtually complete urinary shutdown despite administration of large amount of fluids.
K’s rectal temperature varied between 39.4 and 39.7°C for the first 6 hours and then fell precipitously to normal, where it remained for the duration of his life. This high initial temperature and its maintenance for 6 hours were considered in keeping with his suspected massive dose of radiation. His prognosis was grave.
Of all the various determinations made during the course of the illness, changes in white cell count were found to be the simplest and best prognostic indicator of severe irradiation. The virtual disappearance of lymphocytes from the peripheral circulation within 6 hours of exposure was considered a grave sign.
Sixteen different therapeutic agents were employed in the symptomatic treatment of K over about a 30-hour period. In spite of this and continued oxygen administration, his heart tones became very distant, slow and irregular about 32 hours after irradiation. His heart then became progressively weaker and suddenly stopped 34 hours 45 minutes after irradiation.
Windscale Reactor No. 1 Accident of 9-12 October 1957
Windscale reactor No. 1 was an air-cooled, graphite-moderated natural uranium-fuelled plutonium production reactor. The core was partially ruined by fire on 15 October 1957. This fire resulted in a release of approximately 0.74 PBq (10+15 Bq) of iodine-131 (131I) to the downwind environment.
According to a US Atomic Energy Commission accident information report about the Windscale incident, the accident was caused by operator judgement errors concerning thermocouple data and was made worse by faulty handling of the reactor that permitted the graphite temperature to rise too rapidly. Also contributory was the fact that fuel temperature thermocouples were located in the hottest part of the reactor (that is, where the highest dose rates occurred) during normal operations rather than in parts of the reactor which were hottest during an abnormal release. A second equipment deficiency was the reactor power meter, which was calibrated for normal operations and read low during the annealing. As a result of the second heating cycle, the graphite temperature rose on 9 October, especially in the lower front part of the reactor where some cladding had failed because of the earlier rapid temperature rise. Although there were a number of small iodine releases on 9 October, the releases were not recognized until 10 October when the stack activity meter showed a significant increase (which was not regarded as highly significant). Finally, on the afternoon of 10 October, other monitoring (Calder site) indicated the release of radioactivity. Efforts to cool the reactor by forcing air through it not only failed but actually increased the magnitude of the radioactivity released.
The estimated releases from the Windscale accident were 0.74 PBq of 131I, 0.22 PBq of caesium-137 (137Cs), 3.0 TBq (1012Bq) of strontium-89 (89Sr), and 0.33 TBq of strontium-90
(90Sr). The highest offsite gamma absorbed dose rate was about 35 μGy/h due to airborne activity. Air activity readings around the Windscale and Calder plants often were 5 to 10 times the maximum permissible levels, with occasional peaks of 150 times permissible levels. A milk ban extended over a radius of approximately 420 km.
During operations to bring the reactor under control, 14 workers received dose equivalents greater than 30 mSv per calendar quarter, with the maximum dose equivalent at 46 mSv per calendar quarter.
There were many lessons learned concerning natural uranium reactor design and operation. The inadequacies concerning reactor instrumentation and reactor operator training also bring up points analogous to the Three Mile Island accident (see below).
No guidelines existed for short-term permissible exposure to radioiodine in food. The British Medical Research Council performed a prompt and thorough investigation and analysis. Much ingenuity was used in promptly deriving maximum permissible concentrations for 131I in food. The study Emergency Reference Levels that resulted from this accident serves as a basis for emergency planning guides now used worldwide (Bryant 1969).
A useful correlation was derived for predicting significant radioiodine contamination in milk. It was found that gamma radiation levels in pastures which exceeded 0.3 μGy/h yielded milk which exceeded 3.7 MBq/m3.
Absorbed dose from inhalation of external exposure to radioiodines is negligible compared to that from drinking milk or eating dairy products. In an emergency, rapid gamma spectroscopy is preferable to slower laboratory procedures.
Fifteen two-person teams performed radiation surveys and obtained samples. Twenty persons were used for sample coordination and data reporting. About 150 radiochemists were involved in sampling analysis.
Glass wool stack filters are not satisfactory under accident conditions.
Gulf Oil Accelerator Accident of 4 October 1967
Gulf Oil Company technicians were using a 3 MeV Van de Graaff accelerator for the activation of soil samples on 4 October 1967. The combination of an interlock failure on the power key of the accelerator console and the taping of several of the interlocks on the safety tunnel door and the target room inside door produced serious accidental exposures to three individuals. One individual received approximately 1 Gy whole-body dose equivalent, the second received close to 3 Gy whole-body dose equivalent and the third received approximately 6 Gy whole-body dose equivalent, in addition to approximately 60 Gy to the hands and 30 Gy to the feet.
One of the accident victims reported to the medical department, complaining of nausea, vomiting and generalized muscular aches. His symptoms initially were misdiagnosed as flu symptoms. When the second patient came in with approximately the same symptoms, it was decided that they may possibly have received significant radiation exposures. Film badges verified this. Dr. Niel Wald, University of Pittsburgh Radiological Health Division, supervised the dosimetry tests and also acted as coordinating physician in the work-up and treatment of the patients.
Dr. Wald very quickly had absolute filter units flown in to the western Pennsylvania hospital in Pittsburgh where the three patients had been admitted. He set up these absolute filter/laminar flow filters to clean the patients’ environment of all biological contaminants. These “reverse isolation” units were used on the 1 Gy exposure patient for about 16 days, and on the 3 and 6 Gy exposure patients for about a month and half.
Dr. E. Donnal Thomas from the University of Washington arrived to perform a bone marrow transplant on the 6 Gy patient on the eighth day after exposure. The patient’s twin brother served as the bone marrow donor. Although this heroic medical treatment saved the 6 Gy patient’s life, nothing could be done to save his arms and legs, each of which received tens-of-gray absorbed dose.
If the simple operating procedure of always using a survey meter when entering the exposure room had been followed, this tragic accident would have been avoided.
At least two interlocks had been taped closed for long periods of time prior to this accident. Defeating of protective interlocks is intolerable.
Regular maintenance checks should have been made on the key-operated power interlocks for the accelerator.
Timely medical attention saved the life of the person with the highest exposure. The heroic procedure of a complete bone marrow transplant together with the use of reverse isolation and quality medical care were all major factors in saving this person’s life.
Reverse isolation filters can be obtained in a matter of hours to be set up in any hospital to care for highly exposed patients.
In retrospect, medical authorities involved with these patients would have recommended amputation earlier and at a definitive level within two or three months after the exposure. Earlier amputation decreases the likelihood of infection, gives a shorter period of severe pain, reduces pain medication required for the patient, possibly reduces the patient’s hospital stay, and possibly contributes to earlier rehabilitation. Earlier amputation should, of course, be done while correlating dosimetry information with clinical observations.
The SL–1 Prototype Reactor Accident (Idaho, USA, 3 January 1961)
This is the first (and to date the only) fatal accident in the history of US reactor operations. The SL-1 is a prototype of a small Army Package Power Reactor (APPR) designed for air transportation to remote areas for production of electrical power. This reactor was used for fuel testing, and for reactor crew training. It was operated in the remote desert location of the National Reactor Testing Station in Idaho Falls, Idaho, by Combustion Engineering for the US Army. The SL-1 was not a commercial power reactor (AEC 1961; American Nuclear Society 1961).
At the time of the accident, the SL-1 was loaded with 40 fuel elements and 5 control rod blades. It could produce a power level of 3 MW (thermal) and was a boiling water–cooled and –moderated reactor.
The accident resulted in the deaths of three military personnel. The accident was caused by the withdrawal of a single control rod for a distance of more than 1 m. This caused the reactor to go into prompt criticality. The reason why a skilled, licensed reactor operator with much refuelling operation experience withdrew the control rod past its normal stop point is unknown.
One of the three accident victims was still alive when initial response personnel first reached the scene of the accident. High activity fission products covered his body and were embedded in his skin. Portions of the victim’s skin registered in excess of 4.4 Gy/h at 15 cm and hampered rescue and medical treatment.
No reactor designed since the SL-1 accident can be brought to “prompt-critical” state with a single control rod.
All reactors must have portable survey meters onsite that have ranges greater than 20 mGy/h. Survey meters of 10 Gy/h maximum range are recommended.
Note: The Three Mile Island accident showed that 100 Gy/h is the required range for both gamma and beta measurements.
Treatment facilities are required where a highly contaminated patient can receive definitive medical treatment with reasonable safeguards for attendant personnel. Since most of these facilities will be in clinics with other ongoing missions, control of airborne and waterborne radioactive contaminants may require special provisions.
X-ray Machines, Industrial and Analytical
Accidental exposures from x-ray systems are numerous and often involve extremely high exposures to small portions of the body. It is not unusual for x-ray diffraction systems to produce absorbed dose rates of 5 Gy/s at 10 cm from the tube focus. At shorter distances, 100 Gy/s rates have often been measured. The beam is usually narrow, but even a few seconds’ exposure can result in severe local injury (Lubenau et al. 1967; Lindell 1968; Haynie and Olsher 1981; ANSI 1977).
Because these systems are often used in “non-routine” circumstances, they lend themselves to the production of accidental exposures. X-ray systems commonly used in normal operations appear to be reasonably safe. Equipment failure has not caused severe exposures.
Lessons learned from accidental x-ray exposures
Most accidental exposures occurred during non-routine uses when equipment was partially disassembled or shield covers had been removed.
In most serious exposures, adequate instruction for the staff and maintenance personnel had been lacking.
If simple and fail-safe methods had been used to ensure that x-ray tubes were turned off during repairs and maintenance, many accidental exposures would have been avoided.
Finger or wrist personnel dosimeters should be used for operators and maintenance personnel working with these machines.
If interlocks had been required, many accidental exposures would have been avoided.
Operator error was a contributing cause in most of the accidents. Lack of adequate enclosures or poor shielding design often worsened the situation.
Industrial radiography accidents
From the 1950s through the 1970s, the highest radiation accident rate for a single activity has consistently been for industrial radiographic operations (IAEA 1969, 1977). National regulatory bodies continue to struggle to reduce the rate by a combination of improved regulations, strict training requirements and ever tougher inspection and enforcement policies (USCFR 1990). These regulatory efforts have generally succeeded, but many accidents associated with industrial radiography still occur. Legislation allowing huge monetary fines may be the most effective tool in keeping radiation safety focused in the minds of industrial radiography management (and also, therefore, in workers’ minds).
Causes of industrial radiography accidents
Worker training. Industrial radiography probably has lower education and training requirements than any other type of radiation employment. Therefore, existing training requirements must be strictly enforced.
Worker production incentive. For years, major emphasis for industrial radiographers was placed on the amount of successful radiographs produced per day. This practice can lead to unsafe acts as well as to occasional non-use of personnel dosimetry so that exceeding dose equivalent limits would not be detected.
Lack of proper surveys. Thorough surveying of source pigs (storage containers) (figure 1) after every exposure is most important. Not performing these surveys is the single most probable cause of unnecessary exposures, many of which are unrecorded, since industrial radiographers rarely use hand or finger dosimeters (figure 1).
Figure 1. Industrial radiography camera
Equipment problems. Because of heavy use of industrial radiographic cameras, source winding mechanisms can loosen and cause the source to not completely retract into its safe storage position (point A in figure 1). There are also many instances of closet-source interlock failures that cause accidental exposures of personnel.
Design of Emergency Plans
Many excellent guidelines, both general and specific, exist for the design of emergency plans. Some references are particularly helpful. These are given in the suggested readings at the end of this chapter.
Initial drafting of emergency plan and procedures
First, one must assess the entire radioactive material inventory for the subject facility. Then credible accidents must be analysed so that one can determine the probable maximum source release terms. Next, the plan and its procedures must enable the facility operators to:
- recognize an accident situation
- classify the accident according to severity
- take steps to mitigate the accident
- make timely notifications
- call for help efficiently and quickly
- quantify releases
- keep track of exposures both on- and offsite, as well as keep emergency exposures ALARA
- recover the facility as quickly as practical
- keep accurate and detailed records.
Types of accidents associated with nuclear reactors
A list, from most likely to least likely, of types of accidents associated with nuclear reactors follows. (The non-nuclear reactor, general-industrial type accident is by far the most likely.)
- Low level unexpected release of radioactive material with little or no external radiation exposure to personnel. Usually occurs during major overhauls or in shipment of spent resin or spent fuel. Coolant system leakage and coolant-sample sink spills are often causes of spread of radioactive contamination.
- Unexpected external exposure of personnel. This usually occurs during major overhauls or routine maintenance.
- A combination of contamination spread, contamination of personnel, and low-level personnel external radiation exposure is the next most likely accident. These accidents occur under the same conditions as 1 and 2 above.
- Gross surface contamination due to a major reactor coolant system leak or a leak of spent fuel coolant.
- Chips or large particles of activated CRUD (see definition below) in or on skin, ears or eyes.
- High-level radiation exposure of plant personnel. This is usually caused by carelessness.
- Release of small but greater than permissible quantities of radioactive wastes to outside the plant boundary. This is usually associated with human failures.
- Meltdown of reactor. Gross contamination offsite plus high personnel exposure would probably occur.
- Reactor excursion (SL–1 type of accident).
Radionuclides expected from water-cooled reactor accidents:
- activated corrosion and erosion products (commonly known as CRUD) in the coolant; for example, cobalt-60 or -58 (60Co, 58Co), iron-59 (59Fe), manganese-58 (58Mn) and tantalum-183 (183Ta)
- low level fission products usually present in the coolant; for example, iodine-131 (131I) and caesium-137 (137Cs)
- in boiling water reactors, 1 and 2 above plus continuous off-gassing of low levels of tritium
- (3H) and noble radioactive gases such as xenon-133 and -135 (133Xe, 135Xe), argon-41 (41Ar), and krypton-85 (85Kr)
- tritium (3H) manufactured inside the core at the rate of 1.3 × 10–4 atoms of 3H per fission (only a fraction of this leaves the fuel).
Figure 2. Example of a nuclear power plant emergency plan, table of contents
Typical Nuclear Power Plant Emergency Plan, Table of Contents
Figure 2 is an example of a table of contents for a nuclear power plant emergency plan. Such a plan should include each chapter shown and be tailored to meet local requirements. A list of typical power reactor implementation procedures is given in figure 3.
Figure 3. Typical power reactor implementation procedures
Radiological Environmental Monitoring during Accidents
This task is often called EREMP (Emergency Radiological Environmental Monitoring Programme) at large facilities.
One of the most important lessons learned for the US Nuclear Regulatory Commission and other government agencies from the Three Mile Island accident was that one cannot successfully implement EREMP in one or two days without extensive prior planning. Although the US government spent many millions of dollars monitoring the environment around the Three Mile Island nuclear station during the accident, less then 5% of the total releases were measured. This was due to poor and inadequate prior planning.
Designing Emergency Radiological Environmental Monitoring Programmes
Experience has shown that the only successful EREMP is one that is designed into the routine radiological environmental monitoring programme. During the early days of the Three Mile Island accident, it was learned that an effective EREMP cannot be established successfully in a day or two, no matter how much manpower and money are applied to the programme.
All routine radiological environmental monitoring programme locations will be used during long-term accident monitoring. In addition, a number of new locations must be set up so that motorized survey teams have pre-determined locations in each portion of each 22½° sector (see figure 3). Generally, sampling locations will be in areas with roads. However, exceptions must be made for normally inaccessible but potentially occupied sites such as camp grounds and hiking trails within about 16 km downwind of the accident.
Figure 3. Sector and zone designations for radiological sampling and monitoring points within emergency planning zones
Figure 3 shows the sector and zone designation for radiation and environmental monitoring points. One may designate 22½° sectors by cardinal directions (for example, N, NNE, and NE) or by simple letters (for example, A through R). However, use of letters is not recommended because they are easily confused with directional notation. For example, it is less confusing to use the directional W for west rather than the letter N.
Each designated sample location should be visited during a practice drill so that people responsible for monitoring and sampling will be familiar with the location of each point and will be aware of radio “dead spaces,” poor roads, problems with finding the locations in the dark and so on. Since no drill will cover all the pre-designated locations within the 16 km emergency protection zone, drills must be designed so that all sample points will be visited eventually. It is often worthwhile to predetermine the ability of survey team vehicles to communicate with each pre-designated point. The actual locations of the sample points are chosen utilizing the same criteria as in the REMP (NRC 1980); for example, line of site, minimum exclusion area, closest individual, closest community, closest school, hospital, nursing home, milch animal herd, garden, farm and so on.
Radiological monitoring survey team
During an accident involving significant releases of radioactive materials, radiological monitoring teams should be continuously monitoring in the field. They also should continuously monitor onsite if conditions allow. Normally, these teams will monitor for ambient gamma and beta radiation and sample air for the presence of radioactive particulates and halogens.
These teams must be well trained in all monitoring procedures, including monitoring their own exposures, and be able to accurately relay these data to the base station. Details such as survey-meter type, serial number, and open-or closed-window status must be carefully reported on well-designed log sheets.
At the beginning of an emergency, an emergency monitoring team may have to monitor for 12 hours without a break. After the initial period, however, field time for the survey team should be decreased to eight hours with at least one 30 minute break.
Since continuous surveillance may be needed, procedures must be in place to supply the survey teams with food and drink, replacement instruments and batteries, and for back-and-forth transfer of air filters.
Even though survey teams will probably be working 12 hours per shift, three shifts a day are needed to provide continuous surveillance. During the Three Mile Island accident, a minimum of five monitoring teams was deployed at any one time for the first two weeks. The logistics for supporting such an effort must be carefully planned in advance.
Radiological environmental sampling team
The types of environmental samples taken during an accident depend on the type of releases (airborne versus water), direction of the wind and time of year. Soil and drinking water samples must be taken even in winter. Although radio-halogen releases may not be detected, milk samples should be taken because of the large bioaccumulation factor.
Many food and environmental samples must be taken to reassure the public even though technical reasons may not justify the effort. In addition, these data may be invaluable during any subsequent legal proceedings.
Pre-planned log sheets using carefully thought out offsite data procedures are essential for environmental samples. All persons taking environmental samples should have demonstrated a clear understanding of procedures and have documented field training.
If possible, offsite environmental sample data collection should be done by an independent offsite group. It is also preferable that routine environmental samples be taken by the same offsite group, so that the valuable onsite group may be used for other data collection during an accident.
It is notable that during the Three Mile Island accident every single environmental sample that should have been taken was collected, and not one environmental sample was lost. This occurred even though the sampling rate increased by a factor of more than ten over pre-accident sampling rates.
Emergency monitoring equipment
The inventory of emergency monitoring equipment should be at least double that needed at any given time. Lockers should be placed around nuclear complexes in various places so that no one accident will deny access to all of these lockers. To ensure readiness, equipment should be inventoried and its calibration checked at least twice a year and after each drill. Vans and trucks at large nuclear facilities should be completely outfitted for both on and offsite emergency surveillance.
Onsite counting laboratories may be unusable during an emergency. Therefore, prior arrangements must be made for an alternate or a mobile counting laboratory. This is now a requirement for US nuclear power plants (USNRC 1983).
The type and sophistication of environmental monitoring equipment should meet the requirements of attending the nuclear facility’s worst credible accident. Following is a list of typical environmental monitoring equipment required for nuclear power plants:
- Air sampling equipment should include units which are battery operated for short-term sampling and AC operable with strip chart recorders and alarm capabilities for longer-term surveillance.
- Liquid sampling equipment should contain continuous samplers. The samplers must be operable in the local environment, no matter how harsh it is.
- Portable gamma survey meters for implant work should have a maximum range of 100 Gy/h, and separate survey equipment should be able to measure beta radiation up to 100 Gy/h.
- Onsite, personnel dosimetry must include beta measurement capability, as well as finger thermoluminescent dosimeters (TLDs) (figure 4). Other extremity dosimetry also may be needed. Extra sets of control dosimeters are always needed in emergencies. A portable TLD reader may be needed to link with the station computer via telephone modem in emergency locations. In-house survey teams, such as rescue and repair teams, should have low- and high-range pocket dosimeters as well as pre-set alarm dosimeters. Careful thought must be given to pre-established dose levels for teams that may be in high radiation areas.
- Supplies of protective clothing should be supplied in emergency locations and in emergency vehicles. Extra back-up protective clothing should be available in case of accidents lasting for an extended period of time.
- Respiratory protection equipment should be in all emergency lockers and vehicles. Up-to-date lists of respiratory trained personnel should be kept in each of the major emergency equipment storage areas.
- Mobile vehicles equipped with radios are essential for emergency radiation monitoring survey teams. The location and availability of back-up vehicles must be known.
- Environmental survey team equipment should be stored in a convenient place, preferably offsite, so that it is always available.
- Emergency kits should be placed in the Technical Support Center and the Emergency Offsite Facility so that replacement survey teams need not go onsite in order to receive equipment and be deployed.
- For a severe accident involving the release of radioactive materials into the air, preparations must be in place for the use of helicopters and single-engine airplanes for airborne surveillance.
Figure 4. An industrial radiographer wearing a TLD badge and a ring thermoluminescent dosimeter (optional in the US)
Environmental data analysis during a serious accident should be shifted as soon as possible to an offsite location such as the Emergency Offsite Facility.
Pre-set guidelines about when environmental sample data are to be reported to management must be established. The method and frequency for transfer of environmental sample data to governmental agencies should be agreed upon early in the accident.
Health Physics and Radiochemistry Lessons Learned from the Three Mile Island Accident
Outside consultants were needed to perform the following activities because plant health physicists were fully occupied by other duties during the early hours of the 28 March 1979 Three Mile Island accident:
- radioactive effluent release assessment (gaseous and liquid), including sample collection, coordination of laboratories for sample counting, quality control of laboratories, data collection, data analysis, report generation, distribution of data to government agencies and power plant owner
- dose assessment, including suspected and actual overexposure investigations, skin contamination and internal deposition investigations, significant exposure mock-ups, and dose calculations
- radiological environmental monitoring programme, including complete coordination of sample taking, data analysis, report generation and distribution, action-point notifications, expansion of programme for the accident situation and then contraction of the programme for up to one year after the accident
- special beta dosimetry studies, including studies of the state of the art in beta personnel monitoring, modelling of the beta dose to skin from radioactive contaminants, inter-comparisons of all commercially available beta-gamma TLD personnel dosimetry systems.
The above list includes examples of activities that the typical utility health physics staff cannot adequately accomplish during a serious accident. The Three Mile Island health physics staff was very experienced, knowledgeable and competent. They worked 15 to 20 hours per day for the first two weeks of the accident without a break. Yet, additional requirements caused by the accident were so numerous that they were unable to perform many important routine tasks that ordinarily would be performed easily.
Lessons learned from the Three Mile Island accident include:
Auxiliary building entry during accident
- All entries must be on a new radiation work permit reviewed by the senior health physicist onsite and signed by the unit superintendent or designated alternate.
- The appropriate control room should have absolute control over all Auxiliary and Fuel Handling Building entries. No entries must be allowed unless a health physicist is at the control point during the entry.
- No entries without a properly operating survey meter of appropriate range should be allowed. A spot check of meter response should be performed immediately prior to entry.
- Exposure history for all persons prior to their entry into a high radiation area must be obtained.
- Allowable exposures during entry, no matter how important the task should be designated.
Primary coolant sampling during accident
- All samples to be taken on a new radiation work permit should be reviewed by the senior health physicist onsite and signed by the unit superintendent or alternate.
- No coolant samples should be taken unless an extremity dosimeter is worn.
- No coolant samples should be taken without the availability of shielded gloves and tongs at least 60 cm long in case a sample is more radioactive than expected.
- No coolant samples should be taken without a leaded-glass personnel shield in place in case a sample is more radioactive than expected.
- Sample-taking should be discontinued if the exposure to an extremity or to the whole-body is likely to exceed pre-set levels stated on the radiation work permit.
- Significant exposures should be distributed among a number of workers if possible.
- All cases of skin contamination in excess of action levels within 24 hours should be reviewed.
Make-up valve room entry
- Beta and gamma area surveys using remote detectors with appropriate maximum range must be performed.
- Initial entry in an area with an absorbed dose rate of more than 20 mGy/h must have prior review to verify that exposure to radiation will be kept as low as reasonably achievable.
- When water leaks are suspected, possible floor contamination should be detected.
- A consistent programme for type and placement of personnel dosimetry must be put into operation.
- With persons entering an area with an absorbed dose rate of more than 20 mGy/h, TLDs must be assessed immediately after exit.
- It should be verified that all radiation work permit requirements are being carried out prior to entry into an area with an absorbed dose rate of more than 20 mGy/h.
- Controlled-time entries into hazardous areas must be timed by a health physicist.
Protective actions and offsite environmental surveillance from the local government’s perspective
- Before beginning a sampling protocol, criteria for stopping it should be established.
- Outside interference should not be allowed.
- Several confidential telephone lines should be in place. The numbers should be changed after each crisis.
- The capabilities of aerial measuring systems are better than most people think they are.
- A tape recorder should be in hand and data recorded regularly.
- While the acute episode is in progress, the reading of newspapers, watching television and listening to the radio should be abandoned as these activities only add to existing tensions.
- Food delivery and other comforts such as sleeping facilities should be planned for as it may be impossible to go home for a while.
- Alternate analytic capabilities should be planned for. Even a small accident can alter laboratory background radiation levels significantly.
- It should be noted that more energy will be expended in heading off unsound decisions than in dealing with real problems.
- It should be understood that emergencies cannot be managed from remote locations.
- It should be noted that protective action recommendations are not amenable to committee vote.
- All non-essential calls should be put on hold, time-wasters are to be hung up on.
The Goiânia Radiological Accident of 1985
A 51 TBq 137Cs teletherapy unit was stolen from an abandoned clinic in Goiânia, Brazil, on or around 13 September 1985. Two people looking for scrap metal took home the source assembly of the teletherapy unit and attempted to disassemble the parts. The absorbed dose rate from the source assembly was about 46 Gy/h at 1 m. They did not understand the meaning of the three-bladed radiation symbol on the source capsule.
The source capsule ruptured during disassembly. Highly soluble caesium-137 chloride (137CsCl) powder was disbursed throughout a part of this city of 1,000,000 people and caused one of the most serious sealed source accidents in history.
After the disassembly, remnants of the source assembly were sold to a junk dealer. He discovered that the 137CsCl powder glowed in the dark with a blue colour (presumably, this was Cerenkov radiation). He thought that the powder could be a gemstone or even supernatural. Many friends and relatives came to see the “wonderful” glow. Portions of the source were given to a number of families. This process continued for about five days. By this time a number of people had developed gastro-intestinal syndrome symptoms from radiation exposure.
Patients who went to the hospital with severe gastro-intestinal disorders were misdiagnosed as having allergic reactions to something they ate. A patient who had severe skin effects from handling the source was suspected of having some tropical skin disease and was sent to the Tropical Disease Hospital.
This tragic sequence of events continued undetected by knowledgeable personnel for about two weeks. Many people rubbed the 137CsCl powder on their skins so that they could glow blue. The sequence might have continued much longer except that one of the irradiated persons finally connected the illnesses with the source capsule. She took the remnants of the 137CsCl source on a bus to the Public Health Department in Goiânia where she left it. A visiting medical physicist surveyed the source the next day. He took actions on his own initiative to evacuate two junkyard areas and to inform authorities. The speed and overall size of response of the Brazilian government, once it became aware of the accident, were impressive.
About 249 people were contaminated. Fifty-four were hospitalized. Four people died, one of whom was a six-year-old girl who received an internal dose of about 4 Gy from ingesting about 1 GBq (109 Bq) of 137Cs.
Response to the accident
The objectives of the initial response phase were to:
- identify the main sites of contamination
- evacuate residences where levels of radioactivity exceeded the intervention levels adopted
- establish health physics controls around these areas, preventing access where necessary
- identify persons who had incurred significant doses or were contaminated.
The medical team initially:
- upon its arrival in Goiânia, took histories and triaged according to acute radiation syndrome symptoms
- sent all acute radiation patients to Goiânia Hospital (which was set up in advance for contamination and exposure control)
- transferred by air the next day the six most critical patients to the tertiary care center at a naval hospital in Rio de Janeiro (later eight more patients were transferred to this hospital)
- made arrangements for cytogenetic radiation dosimetry
- based medical management on each patient on that patient’s clinical course
- gave informal instruction to clinical laboratory staff to diminish their fears (the Goiânia medical community was reluctant to help).
- assisted physicians in radiation dosimetry, bioassay and skin decontamination
- coordinated and interpreted analysis of 4,000 urine and faecal samples in a four-month period
- whole-body counted 600 individuals
- coordinated radio-contamination monitoring of 112,000 individuals (249 were contaminated)
- performed aerial survey of entire city and suburbs utilizing hastily assembled NaI detectors
- performed auto-mounted NaI detector surveys of over 2,000 km of roads
- set up action levels for decontamination of people, buildings, autos, soil and so on
- coordinated 550 workers employed in decontamination efforts
- coordinated demolition of seven houses and decontamination of 85 houses
- coordinated hauling of 275 truckloads of contaminated waste
- coordinated decontamination of 50 vehicles
- coordinated packaging of 3,500 cubic metres of contaminated waste
- utilized 55 survey meters, 23 contamination monitors and 450 self-reading dosimeters.
Acute radiation syndrome patients
Four patients died as a result of absorbed doses ranging from 4 to 6 Gy. Two patients exhibited severe bone marrow depression, but lived in spite of absorbed doses of 6.2 and 7.1 Gy (cytogenetic estimate). Four patients survived with estimated absorbed doses from 2.5 to 4 Gy.
Radiation-induced skin injury
Nineteen of twenty hospitalized patients had radiation-induced skin injuries, which started with swelling and blistering. These lesions later ruptured and secreted fluid. Ten of the nineteen skin injuries developed deep lesions about four to five weeks after irradiation. These deep lesions were indicative of significant gamma exposure of deeper tissues.
All skin lesions were contaminated with 137Cs, with absorbed dose rates up to 15 mGy/h.
The six-year-old girl who ingested 1 TBq of 137Cs (and who died one month later) had generalized skin contamination that averaged 3 mGy/h.
One patient required an amputation about a month after exposure. Blood-pool imaging was useful in determining the demarcation between injured and normal arterioles.
Internal contamination result
Statistical tests showed no significant differences between body burdens determined by whole body counting as opposed to those determined by urinary excretion data.
Models that related bioassay data with intakes and body burden were validated. These models were also applicable for different age groups.
Prussian Blue was useful in promoting the elimination of 137CsCl from the body (if dosage was greater than 3 Gy/d).
Seventeen patients received diuretics for the elimination of 137CsCl body burdens. These diuretics were ineffective in de-corporating 137Cs and their use was stopped.
Skin decontamination using soap and water, acetic acid, and titanium dioxide (TiO2) was performed on all patients. This decontamination was only partly successful. It was surmised that sweating resulted in recontaminating the skin from the 137Cs body burden.
Contaminated skin lesions are very difficult to decontaminate. Sloughing of necrotic skin significantly reduced contamination levels.
Follow-up study on cytogenetic analysis dose assessment
Frequency of aberrations in lymphocytes at different times after the accident followed three main patterns:
In two cases the frequencies of incidence of aberrations remained constant up to one month after the accident and declined to about 30% of the initial frequency three months later.
In two cases a gradual decrease of about 20% every three months was found.
In two of the cases of highest internal contamination there were increases in the frequency of incidence of aberrations (by about 50% and 100%) over a three-month period.
Follow-up studies on 137Cs body burdens
- Patients’ actual committed doses followed by bioassay.
- Effects of Prussian Blue administration followed.
- In vivo measurements for 20 people made on blood samples, wounds and organs to look for non-homogenous distribution of 137Cs and its retention in body tissues.
- A woman and her newborn baby studied to look for retention and transfer by nursing.
Action levels for intervention
House evacuation was recommended for absorbed dose rates greater than 10 μGy/h at 1 m height inside the house.
Remedial decontamination of property, clothing, soil and food was based on a person not exceeding 5 mGy in a year. Applying this criterion for different pathways resulted in decontaminating the inside of a house if the absorbed dose could exceed 1 mGy in a year and decontaminating soil if the absorbed dose rate could exceed 4 mGy in a year (3 mGy from external radiation and 1 mGy from internal radiation).
The Chernobyl Nuclear Power Reactor Unit 4 Accident of 1986
General description of the accident
The world’s worst nuclear power reactor accident occurred on 26 April 1986 during a very low-powered electrical engineering test. In order to perform this test, a number of safety systems were switched off or blocked.
This unit was a model RBMK-1000, the type of reactor that produced about 65% of all nuclear power generated in the USSR. It was a graphite-moderated, boiling-water reactor that generated 1,000 MW of electricity (MWe). The RBMK-1000 does not have a pressure-tested containment building and is not commonly built in most countries.
The reactor went prompt critical and produced a series of steam explosions. The explosions blew off the entire top of the reactor, destroyed the thin structure covering the reactor, and started a series of fires on the thick asphalt roofs of units 3 and 4. Radioactive releases lasted for ten days, and 31 people died. The USSR delegation to the International Atomic Energy Agency studied the accident. They stated that the Chernobyl Unit 4 RBMK experiments that caused the accident had not received required approval and that the written rules on reactor safety measures were inadequate. The delegation further stated, “The staff involved were not adequately prepared for the tests and were not aware of the possible dangers.” This series of tests created the conditions for the emergency situation and led to a reactor accident which most believed could never occur.
Release of Chernobyl Unit 4 accident fission products
Total activity released
Roughly 1,900 PBq of fission products and fuel (which together were labelled corium by the Three Mile Island Accident Recovery Team) were released over the ten days that it took to put out all the fires and seal off Unit 4 with a neutron absorbing shielding material. Unit 4 is now a permanently sealed steel and concrete sarcophagus that properly contains the residual corium in and around the remains of the destroyed reactor core.
Twenty-five per cent of the 1,900 PBq was released on the first day of the accident. The rest was released during the next nine days.
The most radiologically significant releases were 270 PBq of 131I, 8.1 PBq of 90Sr and 37 PBq of 137Cs. This can be compared with the Three Mile Island accident, which released 7.4 TBq of 131I and no measurable 90Sr or 137Cs.
Environmental dispersion of radioactive materials
The first releases went in a generally northern direction, but subsequent releases went toward the westerly and southwesterly directions. The first plume arrived in Sweden and Finland on 27 April. Nuclear power plant radiological environmental monitoring programmes immediately discovered the release and alerted the world about the accident. Part of this first plume drifted into Poland and East Germany. Subsequent plumes swept into eastern and central Europe on 29 and 30 April. After this, the United Kingdom saw Chernobyl releases on 2 May, followed by Japan and China on 4 May, India on 5 May and Canada and the US on 5 and 6 May. The southern hemisphere did not report detecting this plume.
The deposition of the plume was governed mostly by precipitation. The fallout pattern of the major radionuclides (131I, 137Cs, 134Cs, and 90Sr) was highly variable, even within the USSR. The major risk came from external irradiation from surface deposition, as well as from ingestion of contaminated food.
Radiological consequences of the Chernobyl Unit 4 accident
General acute health consequences
Two persons died immediately, one during the building collapse and one 5.5 hours later from thermal burns. An additional 28 of the reactor’s staff and fire-fighting crew died from radiation injuries. Radiation doses to the offsite population were below levels that can cause immediate radiation effects.
The Chernobyl accident almost doubled the worldwide total of deaths due to radiation accidents through 1986 (from 32 to 61). (It is interesting to note that the three dead from the SL-1 reactor accident in the US are listed as due to a steam explosion and that the first two to die at Chernobyl are also not listed as radiation accident deaths.)
Factors which influenced onsite health consequences of the accident
Personnel dosimetry for the onsite persons at highest risk was not available. The absence of nausea or vomiting for the first six hours after exposure reliably indicated those patients who had received less than potentially fatal absorbed doses. This also was a good indication of patients who did not require immediate medical attention because of radiation exposure. This information together with blood data (decrease in lymphocyte count) was more useful than personnel dosimetry data.
Fire-fighters’ heavy protective garments (a porous canvas) allowed high specific activity fission products to contact bare skin. These beta doses caused severe skin burns and were a significant factor in many of the deaths. Fifty-six workers received severe skin burns. The burns were extremely difficult to treat and were a serious complicating element. They made it impossible to decontaminate the patients prior to transport to hospitals.
There were no clinically significant internal radioactive material body burdens at this time. Only two people had high (but not clinically significant) body burdens.
Of the about 1,000 people screened, 115 were hospitalized due to acute radiation syndrome. Eight medical attendants working onsite incurred the acute radiation syndrome.
As expected, there was no evidence of neutron exposure. (The test looks for sodium-24 (24Na) in blood.)
Factors which influenced offsite health consequences of the accident
Public protective actions can be divided into four distinct periods.
- The first 24 h: The downwind public remained indoors with doors and windows shut. Distribution of potassium iodide (KI) began in order to block thyroid uptake of 131I.
- One to seven days: Pripyat was evacuated after safe evacuation routes were established. Decontamination stations were established. The Kiev region was evacuated. The total number of people evacuated was more than 88,000.
- One to six weeks: The total number of evacuated people rose to 115,000. All these were medically examined and resettled. Potassium iodide was administered to 5.4 million Russians, including 1.7 million children. Thyroid doses were reduced by about 80 to 90%. Tens of thousands of cattle were removed from contaminated areas. Local milk and foodstuffs were banned over a large area (as dictated by derived intervention levels).
- After 6 weeks: The 30 km radius circle of evacuation was divided into three sub-zones: (a) a zone of 4 to 5 km where no public re-entry is expected in the foreseeable future, (b) a 5 to 10 km zone where limited public re-entry will be allowed after a specific time and (c) a 10 to 30 km zone where the public will eventually be allowed to return.
A great effort has been expended in decontaminating offsite areas.
The total radiological dose to the USSR population was reported by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) to be 226,000 person-Sv (72,000 person-Sv committed during the first year). The worldwide estimated collective dose equivalent is on the order of 600,000 person-Sv. Time and further study will refine this estimate (UNSCEAR 1988).
International Atomic Energy Agency
P.O. Box 100
International Commission on Radiation Units and Measurements
7910 Woodmont Avenue
Bethesda, Maryland 20814
International Commission on Radiological Protection
P.O. Box No. 35
International Radiation Protection Association
Eindhoven University of Technology
P.O. Box 662
5600 AR Eindhoven
United Nations Committee on the Effects of Atomic Radiation
4611-F Assembly Drive
Lanham, Maryland 20706-4391