A “serious incident” at the Paks 2 nuclear power plant during an outage in April has been traced to fuel cooling tank design faults and ‘problems associated with organisation and safety culture’ according to an incident report by the Hungarian Atomic Energy Authority.
The incident, which caused severe damage to fuel assemblies and the leakage of radioactive gases, but did not, according to official sources, put workers or the general population at risk, occurred during the unit’s annual refuelling outage. The fuel assemblies had been successfully removed and cleaned, and were being held in the cooling tank beacause of a delay elsewhere. But owing to ‘insufficient cooling’ they overheated, according to a Paks statement, leading, ultimately, to irreparable damage to a batch of 30 fuel assemblies. The cleaning tank was designed, manufactured and is operated by Framatome ANP.
Five batches of fuel had been successfully cleaned before the incident occurred. Inspections have confirmed that the leakage that resulted was the source of a discharge of radioactive gases, detected at an early stage in cooling tank ventilation air and alerting engineers to the problem. Framatome’s own preliminary report found the same cause. ‘As a result of the insufficient cooling, some of the fuel assemblies overheated. After lifting the lid of the cleaning tank, the overheated fuel assemblies were destroyed due to the backflow of cold water into the tank followed by sudden condensation.’ (sic). The term ‘condensation’ was later supplanted by ‘thermal shock’. HAEA additionally cited violent flash steam production as a probable cause.
It will take some time to repair the damage, thus extending the plant’s outage at a cost of around $220 000 per day plus around $450 000 to replace the written-off fuel.
The incident was provisionally registered as level 2 on the International Nuclear Event Scale (INES) but later up-rated to INES 3 (“serious incident”) when the extent of the damage was discovered. The HAEA and the governmental co-ordination committee found that there had been no increase in atmospheric emissions, and a co-ordinated programme for environmental measurements had started. They concluded that there was no need for countermeasures.
The Paks report confirmed the immediate cause of the incident, exonerating the Framatome staff at the plant, who were operating from incorrect data. The HAEA report confirmed this view of the matter. However it also pointed out that the Paks report was limited by being insufficiently self critical; more importantly, HAEA found errors and lack of vision in the design modelling, and that there were various ‘insufficiencies of planning and preparation’. They picked out design flaws including the omission of any detailed thermohydraulic analysis, inadequate consideration of the consequences of insufficient cooling and the use of a false algorithm to estimate the number of damaged fuel assemblies, resulting in an underestimate. The operators were working from an incorrect model, which led to an incorrect estimate of the time taken to boiling of the coolant if flow stopped. The installation of a much larger coolant tank (30 assemblies instead of 7) should have given rise to studies other than those concerned with cleaning efficiency, but didn’t. The incident management procedure described in the operating manual was ‘not achievable’. Only one (coolant) pump was installed, which violated the principle of tolerance to a single failure.
Further investigations are to be carried out by IAEA and Framatome.
