cnd logo    Scottish CND     Safety of Trident

4. REACTOR ACCIDENTS

There are a range of accidents which could occur involving a PWR 2 reactor - three types are looked at: core damage, loss of coolant and containment failure. A containment failure accident, is considered in detail, and compared with a Soviet submarine accident and with Chernobyl. The effect that such an accident could have on the central belt of Scotland is illustrated. The problems that arise when a nuclear powered and armed submarine is lost at sea are also discussed.

4.1 CORE DAMAGE ACCIDENT

A core damage accident is defined by the Royal Navy as an accident which is contained within the cooling circuit of the reactor. In a core damage accident, failure of the zirconium cladding on the fuel rods will lead to a release of fission products into the primary coolant circuit. Naval manuals suggest that this will not pose a major problem outwith the base, however there will be a long term problem of how to cope with such an incident. Removal of the fuel core would be hazardous and the reactor would be heavily contaminated. The volumes of nuclear waste and hazards to workers would complicate defuelling.

The Russian Navy has major problems with reactors on submarines. It is unclear how many incidents may be identified as core damage accidents, loss of coolant accidents, or other reactor problems. In the Northern fleet, 6 reactors on 4 submarines have been dumped at sea - "it was impossible to offload the fuel from all six submarine reactors dumped with spent nuclear fuel because of the emergency condition of the cores". In addition 85 % of the submarines in the Northern Fleet which have been decommissioned have not been defuelled, in some cases this is because the reactor cores are damaged. There are similar problems in the Pacific fleet - "damaged reactors of three submarines are stored with nuclear fuel, it is also impossible to offload their spent fuel assemblies". The Russian experience shows that there is a significant chance of reactors reaching a condition in which defuelling is very difficult.

4.2 LOSS OF COOLANT ACCIDENT

This is a more serious accident. The Royal Navy's Loss of Coolant Accident scenario assumes that there is a breach in the primary coolant circuit which releases fission products into the reactor compartment, however it is assumed that the reactor containment remains intact. The scenario does take into account the seepage of a small proportion of radioactive material through weakpoints in the bulkheads. This would then spread through the submarine, with some becoming dispersed into the atmosphere, presenting a hazard to the general public. The MoD describe this as a Benchmark 3 accident and claim that the chance of this type of accident is 1 in 10,000 reactor years. Some of the Russian accidents may fall into this category but details are unclear.

A Benchmark 3 accident on a Trafalgar class submarine could result in the release of 4 x 1013 Bq of iodine 131 as well as other radionuclides. The maximum figures for a Trident submarine could be higher, depending on the history of the reactor core. The whole body and single organ doses received have been calculated by the MoD for a Trafalgar class submarine accident. Risks of fatal cancer and thyroid cancer have been deduced on the basis of ICRP risk factors:

Distance 500 m - risk of fatal cancer 1 in 2,500
Distance 1 km - risk of fatal cancer 1 in 6,700
Distance 2 kms - risk of fatal cancer 1 in 20,000

Distance 500 m - risk of thyroid cancer 1 in 8,000
Distance 1 km - risk of thyroid cancer 1 in 40,000
Distance 2 km - risk of thyroid cancer 1 in 100,000

At 1 km from the accident in the short term 2 mSv of the whole body dose will be from cloudshine and 0.7 mSv from inhalation. The remaining 0.3 mSv is the dose from groundshine over a 50 year period, with 0.06 mSv of this received in the first week.

4.3 CONTAINMENT FAILURE ACCIDENT

If there is a loss of coolant accident and the containment fails then this more serious accident can occur. A much larger proportion of the fission products could be released into the atmosphere creating a greater threat to the general public.

Probability of a containment failure accident

The total experience of the US Navy in operating nuclear powered submarines up until 1988 was 3,100 reactor years. The Russian Navy may have 60 % of the submarine reactors in the world. This suggests that the total worldwide experience of operating nuclear reactors to date is around 10,000 reactor years. The MoD claim that the probability of a containment failure accident is less than one in a million years of reactor operations. However in 10,000 reactor years there has already been one such incident which involved a Soviet submarine at Chazhma Bay in 1985. This is described below.

Underestimating the probability of a major nuclear accident can in itself make such an event more likely. This was a factor in the Chernobyl disaster. A reactor operator working there on the day before the accident, AJ Uskov, said that they had been taught that such a chain of events was impossible - "..and already in the classrooms of their institutions they had beaten into their heads: a reactor cannot explode".

Accident at Chazhma Bay

On 10th August 1985 a containment failure accident occurred on a Soviet Victor class submarine at Chazhma Bay, near Vladivostock. The submarine had two reactors. The port reactor was being refuelled when the accident occurred. A crane was being used to reposition the reactor lid when it became warped, triggering a nuclear reaction This caused a thermal explosion which ruptured both the aft bulkhead and the pressure hull. The freshly loaded core was thrown out of the reactor and the roof of the refuelling hut was thrown 70 - 80 metres away. The official casualty figures were 10 killed, 10 cases of acute radiation sickness and 39 other cases of radiation sickness . An eye witness described the devastation - "The submarine looked as if it had been trampled by a giant beast".

There was a fire in the reactor compartment which was localised after 4 hours. Fission products were dispersed by the explosion and fire. Because the hull was ruptured fission products leached into the sea. The total amount of material released has been estimated as 1.85 x 1017 Bq not including 8.1 x 1016 Bq of noble gases. If these figures are correct the radioactivity released was around one seventh of the total released in the Chernobyl disaster.

In 1992, the chairman of the Soviet in the neighbouring town of Shkotovo-22 said "I consider the radiation conditions here remain unsuitable. Every Spring when the ground thaws a high level of radiation is released".

Causes of a containment failure accident

The reactor containment may fail because it is damaged from the outside as a result of a collision or explosion. It may also fail if an explosion occurs within the reactor compartment. During a loss of coolant accident there could be a series of explosions, one of which might rupture the containment. The sequence can be described as follows:

If the reactor is operating at full power and there is a catastrophic break in the primary coolant circuit, water in the circuit will flash into steam causing the first pressure peak, 1 minute after the break. The heat of the reactor core will rise and at 900o C there will be an explosive reaction involving the zirconium cladding of the fuel rods, 3 minutes after the break. The heat of the core will rise until the core melts at 2000o C . There will be a build up of hydrogen which could result in a third explosion, 20 minutes after the break. A fourth explosion may also occur 3 hours after the break.

Effect of a containment failure accident on a Trident submarine in the Clyde

A Royal Navy manual estimates the levels of radiation in a Trafalgar class submarine. If the reactor core is at the end of its life and has operated at full power for the preceding 100 hours then it is likely to contain 4 x 1018 Bq of mixed fission products of which 4 x 1016 Bq is iodine 131. Assuming that the plant is operating at normal temperature and pressure when there is a major accident and the containment is breached then there could be a release of 100% of the iodine, caesium and noble gases in the reactor and of smaller proportions of other radionuclides. Most of this would be released within one hour of the accident.

A second manual describes the possible effects of a Benchmark 6 accident, when the reactor containment fails. It is assumed that this relates to the scenario just described. The wind speed is 4 m/s, approximately 9 mph, and there is average air stability, category D. Whole body dose and thyroid dose figures at varying distances from the scene of the accident are indicated in a graph in the manual.

Example of the effects of a containment failure accident

The scenario is a containment failure accident on a Trident submarine which takes place in the Clyde estuary at the point where two ferry routes and the two shipping lanes to Loch Long and the Clyde intersect. It is assumed that the radiation released is the same as in the Navy model; higher levels of fission products would be present in the reactor of a Trident submarine at the end of the core life than in a Trafalgar class submarine. Average wind speed and air stability conditions are assumed as in the Navy model. The wind is blowing from the West, which is the case around 5% of the time. This example was chosen because radiation levels at varying distances can be associated with specific towns across Scotland; this does not represent the worst case scenario, either in terms of accident site or weather conditions.

Radiation levels directly downwind from the accident have been taken from Navy figures. Plume width was then calculated from estimates of radiation level to the North and South of the direct line, calculated for the given weather conditions. These calculations do not take account of the effects of topographic features or of rain.

300 mSv is the National Radiological Protection Board (NRPB) upper Emergency Reference Level (ERL) at which evacuation is recommended. 30 mSv is the lower ERL at which evacuation should be considered and the upper ERL at which sheltering is recommended. 3 mSv is the lower ERL at which sheltering should be considered. The area within which the thyroid dose is estimated to be above 50 mSv is also shown - this is the MoD Emergency Action Guidance Level (EAGL) at which potassium iodate tablets should be issued.

The results of this example showed that: sheltering should be considered at more than 100 km from the site, evacuation should be considered at 30 km and iodate tablets should be issued at 80 km. The probability of developing fatal cancer as a result of the accident is indicated for specific sites, these were calculated from the whole body dose estimates on the basis of the ICRP risk factor of 0.05 per Sv. The probability of developing thyroid cancer if iodate tablets are not taken immediately is estimated as 1 in 400 in Greenock, 1 in 2,700 in Milngavie and 1 in 10,000 in Edinburgh based on ICRP risk factor of 0.0025 per Sv.

Type of radiation exposure

The Navy manuals provide some information on how the doses were calculated. At 1 km from the scene 61 % of the whole body dose would be from inhalation, 4 % from cloudshine and 34 % from groundshine. The dose from inhalation and cloudshine would be received during the passage of the radioactive plume. The groundshine dose has been calculated over a 50 year period, with 20 % received in the first week and 50 % in the first year.

There would also be a risk from gamma shine in the immediate area of the accident. This would extend beyond the 550 m distance previously claimed by the MoD. The MoD argue that if the accident was at Faslane, buildings at the submarine base would limit the effect to this distance, but this would not apply at other sites, or even to the other side of Gareloch. Significant radiation doses from gamma shine could be received by those 1 km or more from the reactor. In the scenario described members of the public on the shore could be exposed to gamma shine.

Comparison with Chernobyl

Royal Navy training manuals refer to Chernobyl as a Benchmark 6 accident. The most serious type of submarine reactor accident detailed in the manual is also described as Benchmark 6. The amount of iodine 131 which was contained in the Chernobyl reactor has been estimated to be 7.5 x 1020 Bq. The total amount of iodine 131 in a PWR1 reactor in the model reactor scenario is estimated to be 4 x 1016 Bq. The equivalent for a PWR 2 reactor with the same core history could be twice as much. This suggests that the Chernobyl reactor had around 10,000 times as much iodine 131 as this PWR 2 reactor. However only a proportion of the iodine 131 was released into the atmosphere in the Chernobyl accident, the official estimate is 2.6 x 1017 Bq. The Benchmark 6 scenario described in the Greenwich training manual assumes that 100 % of the iodine in the submarine reactor is dispersed. This would be equivalent to one third of the official figure of the amount released at Chernobyl.

Lessons from Chernobyl

A wide range of figures from less than 50 to many thousands have been produced as estimates of the long term fatalities which might arise from the Chernobyl accident. There were significant rises in radiation levels across Europe and beyond. Estimated doses in some parts of Scotland, Wales and North West England were higher than 1 mSv. The incident was extended over 10 days; as wind directions changed, so radioactive material was dispersed in different directions. Particles with a diameter of 1 - 20 microns were found in Sweden. It is estimated that 60 % of all radioactive isotopes from the Chernobyl accident were deposited within 20 kms of the site and 40 % dispersed more than 20 kms away. In the Ukraine 5 million km2 were contaminated.

The highest doses of radiation were received by the reactor personnel and by the large number of people who were called in to deal with the accident. Fire crews who tackled the initial fires were particularly at risk. In an accident at a submarine berth, base emergency services, local fire and ambulance crews and personnel in the base assisting on the scene could be exposed to high levels of radiation.

A total of 600,000 people were exposed to radiation during the clean up operation. The Soviet army had many units trained in decontamination and most of them were deployed at some stage during the follow up operation. Clean up operations after a major submarine accident would involve radiation exposure to a large number of people.

Chernobyl provides some evidence of the effect of a nuclear accident on wildlife. Visitors to the nearby town of Pripyat commented on the dreadful state of animals and birds which they found 6 weeks after the accident. The local residents thought that they were only leaving for 3 days and so they left their pets, in other villages many of them were shot. " .. in Pripyat they (the animals) weren't shot. Here (they) crawled, half alive, along the road, in terrible pain. Birds looked as if they had crawled out of water ... unable to fly or walk ... cats with dirty fur, as if it had been burnt in places ..". The visitor also reported that half of the animals were blind. It is estimated that after 6 weeks they may have accumulated doses of 7 - 10 Sv. Further from the site there were signs of the effects on animals. Deformed calves 100 kms from the site were believed to be due to the accident.

Radiation concentrated in lichen and in the bodies of reindeer in Northern Scandinavia. In 1993 the carcasses of 2,700 reindeer in Sweden were destroyed because of levels of radioactivity from Chernobyl. In the same year there were still some restrictions on sheep sales in Scotland because of the fallout.

Pine trees in a large area around Chernobyl died as a result of the accident. There were abnormalities observed in other species of trees, including oak, around the area.

Local Authority response

The effects of a major nuclear accident may be reduced if appropriate action is taken by local authorities. In order to do so they have to be given detailed information at an early stage. There are many examples of how the MoD has failed to inform Emergency Planners of incidents. One case was on 30th April 1992 when the MoD failed to inform Plymouth City Council immediately of a serious fire on a nuclear powered submarine at Devonport. Captain David Hall, the Chief Staff Officer (Nuclear) at Devonport later admitted - "It was a bureaucratic mess up. Unfortunately the message that we passed to our central authority for permission to speak to the public got lost in the system and we didn't get permission to speak out until quite late on. .... It was a Ministry of Defence cock up, as simple as that".

4.4 RADIATION HAZARD FROM A SUNKEN SUBMARINE

There are already more than 5 nuclear powered submarines on the seabed which have sunk during accidents. This suggests that the chance of a nuclear powered submarine being sunk in peacetime is greater than 1 in 2,000 reactor years.

In April 1989 the Soviet submarine Komsomolets sank to a depth of 1700 m near Bear Island. Sea bed examination of the vessel in May 1992 showed that there were cracks along the entire length of the titanium hull. Some cracks were of 30 and 40 cms width. The examination also found that the primary coolant circuit was not hermetic, ie fission products could leach out into sea water. If a Trident submarine sank it is likely that there might be similar cracks in the hull and bulkheads, which form the containment of the reactor. Royal Navy submarines are designed with a maximum diving depth of around 400 m. The actual strength of each hull varies, depending on the history of the submarine.

If a Trident submarine is lost at sea, in addition to fatalities amongst the crew, the accident would create a long term radiation risk, even if there was no short term nuclear hazard. Radioactive products from the reactor would leach into the sea. There would be a particular problem from long lived isotopes which would eventually be released even if the primary circuit and the containment remained intact for several years. The plutonium in the warheads would present a long term hazard.

Radioactive products in the deep sea will be absorbed by organisms feeding on or near the sea bed; these then become food for whales, dolphins and other species. Over time some radioactive material will be moved from deep to surface waters by biological activity and also by sea currents. A small proportion of what was released into the sea would end up in the human food chain.

Scottish CND     Safety of Trident