Reactor accidents

The Windscale accident    

The Windscale reactors were graphite moderated and air cooled and were used for plutonium production. During operation the following process occurs with the graphite, whose temperature is about 200 to 300 Celsius: the neutrons collide with the carbon nuclei and with a certain likelihood the neutrons can knock them out from their location in the lattice. In this event, the displaced atom will reach a higher energy level and therefore the graphite will store some amount of energy. If we start to heat up a piece of graphite that had acquired such energy, the atoms can jump back to their original, lower energy locations and the energy difference in the form of heat will further warm up the graphite. This self-exciting process might as well lead to the ignition of the graphite.

The process is called the Wigner-effect after its discoverer. Eugene Wigner pointed out the possibility of such effects in the design phase of the Hanford plutonium production reactors and he also found the remedy: before the graphite would be over-excited, it must be regularly heated up in order that the stored heat can be emitted. The operators of the Windscale reactor were aware of this, however in 1957 they did the heating up too late and without sufficient care. The Wigner-effect manifested itself, a part of the reactor overheated and finally some of the graphite started to burn.

The reactor was flooded with carbon-dioxide, but they were not sure if it was sufficient. Eventually they decided to put out the fire with the aid of water. The filters built in the chimneys, which were 125 m high, prevented most of the radioactivity from being released and thus there was no serious environmental damage, nor any direct casualty. In a 500 km² area in the vicinity of the reactor milk was qualified as unsuitable for human consumption, and the authorities took it away because the concentration of the isotope ¹³¹I exceeded the authorized limit. One member of the personnel obtained a dose of 46 mSv, which is 20 times that of natural background radiation. The exposure dose of the inhabitants remained under the permitted value, due to the authority’s measures.

Accident at the Three Mile Island nuclear power plant     

The Three Mile Island nuclear power plant is located a few miles from Harrisburg, Pennsylvania. The TMI-2 was equipped with a pressurized water reactor, which had two cooling loops and the electrical power of the unit was 907 MW. The reactor went critical just a year before the day of the accident and the official commissioning was performed three months earlier.     

This most serious accident of purely peaceful, commercial-purpose nuclear energetics happened on 28 March 1979. It is essential to emphasize that, although the accident was technically a very serious case, practically a very low amount of radioactivity was released to the environment. The radioactivity that got into the environment was only 1/40,000 part of that of the Chernobyl accident and 1/400 of that of the Windscale fire.     

The serious of events that lead to the accident started at dawn on 28 March.

· During maintenance work, water got into a pipe which was not shown in the project plans. This caused one of the valves of the feedwater system of the heat exchanger to close down. Within a few minutes, this invoked turbine loss and started the emergency feedwater pumps.

. The emergency feedwater pumps did not carry enough water because two days earlier workers forgot that the required valves were shut.     

Translating the above to everyday language we can say that the removal of heat stopped in one of the reactor's loops due to an error of construction and a human fault. As a consequence, the pressure and temperature was rising rapidly in the reactor vessel. The pressure increase called forth two events:

· The emergency scram system started to work, thus the control and safety rods were inserted into the reactor core. The reactor stopped, however, the residual heat of the fission products is still significant after stopping. Therefore the core must be cooled further.

·  Due to the high pressure the so-called pressure control valve opened. The purpose of this valve is to prevent the reactor vessel from being under dangerously high pressure.     

The occurrence of the latter two events was right and necessary, but the situation got worse again because of another human fault and error of construction:

· After stopping of the reactor the pressure control valves should have closed upon lowering pressure. This did not happen, however, because the boric acid that had settled out on the valve hindered it. This fact in itself would not have been a problem, since the valve could be closed by hand. However, in the control room the display did not show the physical status of the valve but the valve actuator voltage on/off status. To be more understandable: the display showed whether the system was last commanded to open or to close the valve.

· The operators, who did not realize the reason for the falling pressure - i.e. the primary circuit was not closed - stopped the high pressure emergency core cooling system, which automatically and properly started in the meantime.     

The pressure further decreased in the reactor vessel and the coolant began to boil, while about two hours after the first event of the accident the upper part of the core became dry. The temperature of the fuel cladding reached 1,100 ^oC; the cladding opened up and a hydrogen-zirconium reaction started. Part of the produced hydrogen got out to the containment and later a smaller explosion also occurred. The explosion was audible in the control room but the operators did not notice it.     

The computer program that processed the signals from the thermal sensors located in the core did not take into account the possibility of a core drying out, therefore the machine was writing only question marks now. Accordingly the personnel did not have proper knowledge about what was going on in the core. Finally, an expert, who was contacted by phone asked the decisive question: Have you closed the pressure control valve manually? The answer was: "Yes, right now."    

From this moment on, the state of the reactor was not dangerous, but it still took ten hours to start one of the main circulating pumps before the condition of the reactor could stabilize.     

The TMI-2 accident gave a vast push to American reactor safety. John Kemény (note that BASIC language was worked out by him) led the panel of experts investigating the circumstances of the accident. His most important conclusions and recommendations on reactor safety development, which have already been implemented, were the following:

· "The machines worked quite well. If there had been no human fault, the whole thing would have been a minor failure. (...) (The operators) were trained int he manner of: 'You have to push this button.' (...) We need operators who thoroughly understand the operation of the whole nuclear power plant and consequently can react to the smallest disturbances in the appropriate manner."

· "A major failure needs a quick response and this has to be done by the machine itself. The smaller disturbances usually evolve more slowly and the clearance of these can be entrusted to human decision. Since these kinds of failures are more frequent, more attention must be paid to them." That is, contrary to the earlier practice, not only the potentially most serious accident has to be prepared for, but also the smaller and more frequent ones. The TMI-2 accident drew attention to the fact that a major accident can evolve at a higher likelihood from these disturbances.

· During the accident it occurred that an operator pushed 80 buttons in 1 minute. These and similar experiences led to the recognition that one of the main safety enhancement objectives should be the simplification of the monitoring and control desk, as well as the elements that belong to a reactor.     

It is tragi-comical that the most serious damage to the inhabitants of the vicinity was caused by a failure outside the power plant. Not even this could have caused such damage if the media had not generated the atmosphere for it. The reporters were informed at 9 o'clock Tuesday that there was a failure at the power plant, but no radioactivity was released to the environment. The next day the CBS channel made the following announcement (which was totally false): "This is the first step towards a nuclear nightmare. The radioactivity is so intense inside the power plant that it can penetrate through the 1 m thick shielding wall and can be measured at a distance of a mile." On Thursday they kept the schools closed. The last failure happened on Friday: the Harrisburgh air-raid alarms sounded accidentally. The state governor suggested that all pregnant women and all children be transported away from the area. The freeways were flooded by thousands of those seeking shelter and a great number of accidents happened.

Reactor accidents - Chernobyl     

A high-power, boiling water type reactor (RBMK), in which the coolant flows in pressure tubes, was developed at the Soviet Technical-Energetical Research Institute at the beginning of the 1970s. The first two reactors of this type were built in St. Petersburg during the seventies. Due to the energy needs of developing industry, further RBMK reactors were built in the western parts of Russia, e.g. in Kursk, furthermore in Ignalina, Lithuania, and Chernobyl, Ukraine.     

Minor problems with RBMK reactors had arisen before the Chernobyl accident. At the Ignalina NPP an interesting phenomenon was observed: when the operators pushed the control rods into the reactor, they experienced a temporary increase in reactivity instead of its expected decrease. The same phenomenon occurred during the test run of the fourth Chernobyl reactor, but it was not considered important enough to draw the operators' attention to it, or to write it down in the operating manual of the reactor.     

An engineer of the Chernobyl plant, Anatolij Diatlov, was concerned about another problem. In the event that there was a sudden loss of electric supply, the pumps that circulate the coolant of the reactor core would stop. In such cases the automatic control systems shut the reactor down; however, the remnant heat must be carried away further on. The electricity, which is necessary to operate the main circulating pumps is generated using Diesel motors. Start-up of these motors requires a few minutes, however. Diatlov found the following solution to the problem. Although after a loss of electric supply the reactor stops and thus the steam supply of turbines terminates, the turbine rotors will remain spinning due to their inertia for some minutes. Electric current can be generated if the slowing down turbine rotors drive generators. As the turbine slows, the current will go down, of course. This problem can be remedied with the aid of complicated switches in electric circuits. The method cannot be tried in practice during operation. They had to wait until they could shut down one of the reactors for the spring refuelling and overhaul period. Diatlov obtained permission from Fomin, the chief engineer of the nuclear power plant, to perform the "electrical engineering" experiment on unit No. 4 in the spring of 1986. 

· At 1 o'clock dawn on Friday, 25 April 1986, they started to reduce the 3.2 GW thermal power.

· By 13:00, the power had gone down to 1.6 GW. One of the turbines was disconnected from the reactor.

· At 14:00, the electric distribution centre informed the Chernobyl Lenin NPP that in spite of the approaching weekend the energy need of consumers was greater than expected. Therefore, they did not decrease the power further.

· At 23:10 the distribution centre announced that finally the energy need of consumers had decreased to a level that made it possible to disconnect the fourth unit from the network. So the innovators, who were a little tired of the delay, could start to implement their idea, which was actually intended to enhance safety.     

The young electrical engineers mainly kept an eye on the electric supply of the pumps. They did not take into account that the xenon-poisoning at low power operation makes the reactor unstable, as had been discovered by John Archibald Wheeler and Eugene Wigner as early as the 1940s in Hanford. Saturday came, which was Easter-eve according to the orthodox calendar. The specialists as well as the decision-makers travelled to their weekend houses for Easter. (Most failures happen at dawn at weekends.) Due to the accumulated reactor poison most control rods were pulled out far more than allowed by the regulations.  

Control rods of the Chernobyl reactor    

Due to a design error of RBMK reactors, the upper and lower parts of the control rods contain graphite. According to the regulations, in a shut-down reactor the control rod should be at position D. During operation it should be at position C, in which case graphite is located in the reactor core instead of neutron absorbing borated steel. Before the accident, however, due to the accumulated reactor poisons, the automatic control system pulled the rods out to level A, which is not allowed. Therefore, the space of the control rods was occupied by water instead of graphite. If one inserts a control rod into the reactor in order to decrease power, graphite takes the place of water. Since graphite practically does not absorb neutrons, while water does, there will be a temporary increase in power, as had been observed earlier in Ignalina. However, the operators were not informed about this phenomenon and thus they decided not to take into account the regulations limiting the extent to which a control rod can be pulled out. The relevant Soviet leaders said later in vain: "Under such circumstances, even the prime minister does not have the right to give permission to operate the reactor." Concerning the dynamic behaviour, in those minutes the reactor was different from what the operators thought it was like. The fact that the design of the control rod moving equipment made the excessive control rod pulling out possible is considered as a further construction fault.     

Diatlov gave permission to start the experiment. The operators themselves wanted to control the reactor instead of the "unimaginative" automations. The emergency core cooling system was switched off - of course against the regulations - at 2 pm on Friday. At dawn on the 26th, with the permission of Diatlov, the automatic equipment that was responsible for controlling the evenness of the power density of the huge reactor was switched off, as well.

· 0:28 am, 26 April 1986. To make sure, the operators increased the flow rate of cooling water above the authorized value. Therefore, the water cooled down and the amount of steam produced in the reactor dropped. When they started to decrease the power from 1.6 GW to the planned 0.7 GW, it went down more than expected due to the positive void coefficient: it dropped to 0.03 GW. They should have waited a day for the decay of the accumulated ¹³⁵I and ¹³⁵Xe and thus the instability caused by xenon-poisoning could have disappeared.

· 1:07 am. Alexej Akinov and Leonid Toptunov, the two operators started to hesitate referring to the regulations, but Diatlov commanded them to pull the control rods even further out. In this way they managed to stabilize the power at 0.2 GW. (The regulations prohibit operation under 0.7 GW.) Thinking of the low thermal power they decreased the flow rate of the cooling water.

· 1:22 am. The last data printed by the computer: 0.2 GW.

· 1:23 am. Eventually, the real experiment started. The operators disabled the SCRAM automatics too, which would have stopped the reactor if the number of neutrons was rising too quickly. (This action was very much against regulations. In the case of a modern plant, it is physically impossible.) Next they switched off the generator of the second turbine because the goal of the experiment was to ensure cooling to the reactor in case of electricity loss.

· 1:23:20 am. Hardly 20 seconds elapsed when, due to the loss of steam consumption of the turbine, the coolant temperature started to rise and consequently the control rods began to move downwards. However, this resulted in situation B, when the place of water was occupied by graphite and so the power increased by several per cents.

· 1:23:40 am. The power of the reactor with positive feed-back jumped to 0.32 GW from 0.2 GW. As soon as operator Akimov observed this, he pushed the scram (emergency shutdown) button.

· 1:23:43 am. The thermal power reached 1.4 GW. At some positions the reactor became supercritical to prompt neutrons, too, and thus uncontrollable. Thermal expansion due to the sudden superheating distorted the metal channels of the control rods and the sinking rods got stuck halfway.

· 1:23:45 am. The thermal power was 3 GW now. More and more of the cooling water boiled away. What Teller and colleagues foresaw in the 1950s happened here: because of the positive void coefficient the chain reaction ran away in the whole reactor.

· 1:23:47 am. Due to the uneven thermal expansion the cladding of fuel rods opened up.

· 1:23:49 am. Thermal deformation of the fuel rods broke the coolant pipes. The suddenly generated steam caused a steam explosion and burst the reactor cover open.

· 1:24:00 am. Above 1100 °C water reacts with the zirconium alloy of the rod cladding. The product of the reaction is hydrogen. Because of the cracks, steam contacted graphite as well and this reaction led to the production of carbon monoxide and hydrogen:

Zr + 2 H₂O = ZrO₂ + 2 H₂,

C + H₂O = CO + H₂.    

The flammable hydrogen and carbon monoxide mixed with the oxygen of air and exploded. This second, chemical explosion brushed off the roof of the building. Graphite started to burn in air and the smoke contaminated the building and its vicinity with radioactivity. Two persons, a technician and an electrical engineer immediately died.    

The temperature inside the reactor reached 3,000 °C. The fission products diffused from the fuel to the burning graphite and to the air from there. All the radioactive noble gases (⁸⁵Kr, ¹³⁵Xe), furthermore about 20 % of the mobile alkaline metal ions (¹³⁷Cs) and volatile iodine (¹³¹I) got out. Only 4 % of the other, less diffusible radioactive metals (⁸⁹Sr, ⁹⁰Sr, ²³⁹Pu) got to the environment. (Unfortunately, human and animal organisms cannot distinguish between Cs and K. However, Ca and Sr are distinguished: the ratio Sr/Ca accumulated in the body was only 20 % of that measurable in the food.)     

The graphite was burning for ten days. Only then could the fire be put out using borated sand and lead. The purpose of boron was absorption of neutrons, while lead sealed off the reactor from air when it melted. During this period, 4 EBq (4·10¹⁸ Bq) activity was released to the air, which was 400 times the amount of radioactivity released during the Hiroshima atomic bomb explosion.     

In 1986 Gorbachov led the Soviet Union. Prime minister Rizkov was informed about the accident on Saturday and he set up a government committee, headed by himself, to start the investigation of the accident and mitigate the damage. On Sunday, 27 April, Legasov flew to the scene of the accident as professional chairman. On Monday morning, upon entering the Sweden’s Forsmark nuclear power plant, which is 1,600 km away from Chernobyl, the clothes of the personnel were measured as being contaminated by radioactivity. By opserving the direction of the wind, the Swedes realized that the radioactivity did not come from a Swedish NPP,  but from the south. They suspected the nearby Lithuanian Ignalina NPP and asked for information from Moscow in a diplomatic way. The information agency TASS released the first announcement at 9 am on Monday. The press caught on to the sensation. Some stated that if the Swedes measured such high radioactivity 1,600 km from Chernobyl, then in the territories of Ukraine and Belarus millions of people might be exposed to life-threatening radiation. They did not account for the fact that the wind was blowing towards Sweden and that was why the Swedes observed significant radiation, while in between there was mostly deserted marshland.     

Legasov informed the International Atomic Energy Agency in Vienna in the spirit of openness about what had happened in Chernobyl. Two basic structural deficiencies were disclosed: the positive void coefficient of the graphite-water system and the faulty construction of the control rod. Later Legasov himself told the democratic newspaper Novij Mir (New World): due to their instability, a similar accident could occur at any of the RBMK reactors. He started a campaign to set up an independent committee to supervise the safety of reactors.     

The trial of the persons accused of being responsible for the Chernobyl accident began in July 1987 in Kiev. Two operators who were initially accused, turned out to be innocent and their steadfastness was appreciated. (Due to the large dose exposure both of them have since died.) A verdict was delivered three weeks later. Brukhanov, director of the power plant and Fomin, chief engineer (for tolerating the constructional deficiencies during the building phase of the plant) and Diatlov deputy chief engineer (for the experiment performed irresponsibly) were sentenced for ten years of imprisonment, while three others were sentenced to between five and two years. Some of them had to released because their health had been damaged by the radiation dose.    

The fourth unit has been entombed. The third unit, which is located in the same building, is still operating. There is increasing pressure to shut down the fifteen still operating RBMK plants in Ukraine, Lithuania and Russia. However, this has not been done yet because of the energy needs of these countries. Anyway, as a safety enhancement measure, the enrichment of fuel was increased to 2.4 % (from 1.8 %). The number of control rods has been raised and the drop time shortened. The arbitrary switch off of the automatic scram system has been made impossible. It is also hoped that the operators have drawn the conclusions from the accident. However, this does not change the fact that the uranium-graphite-water reactor is inherently unstable.

Based on György Marx: Atommag-közelben.