Sizewell B Charcoal Fire 2-3 July 2010

A fire broke out in a charcoal absorber in a ventilation duct just before 2100 BST on 2 July 2010. The duct was located in the spent fuel pond building, but which may extract air from the containment or just from the spent fuel pond enclosure. It was brought under control by 0330 BST after the charcoal absorber was flooded with water.

Charcoal absorbers are used in association with a demoisturiser and a HEPA filter in a section of a ventilation duct which is automatically opened in the event of a loss of coolant accident (LOCA). The HEPA (high efficiency particulate air) filter is to remove any radioactive particles in the extracted air, while the demoisturiser is to ensure its efficiency, which can be damaged by moisture ingress. The purpose of the charcoal is to absorb iodine-131 and iodine-129 which would have been present in the coolant. It is unlikely to absorb tritium, because activated carbon is used as a pre-filter in tritium detectors to eliminate other radioactive matter before measuring the radioactivity from the tritium. Reports mentioned that a “heater” may have caused the ignition of the charcoal and presumably the “demoisturiser” might be in the form of a heater.

There was a loss of coolant from a leaking pressuriser in March 2010, since when the reactor has been shut down pending repairs to the electric heaters in the base of the pressuriser. Presumably some of the “moisture” detected in the containment air entered the LOCA section of the duct and any radioactive particulates and iodine would have been deposited in the HEPA filter and the charcoal.

 

Carbon, iodine, caesium and rubidium isotopes would have been present in the coolant in the event of a leaking fuel can in the reactor core or present in the air above the spent fuel pond had a fuel module with a leaking can been placed in the pond. Also particles of zirconium from the outside of a fuel can could also be present in the coolant or spent fuel pond water. 

If for some reason the HEPA filter failed locally, it may be that radioactive particles accumulated at a small spot in the charcoal and initiated a fire. A fire in the charcoal could have been catalytically ignited by a “heavy metal” particle or particles. Actinides produced by the fission in the core would most likely have remained in it, but in the circumstances of a loss of coolant, the smoke from the charcoal should have been treated as if it were radioactive, especially as its purpose is to absorb radioactive iodine. Also all the components in the ventilation duct, especially the HEPA filter, should be treated as contaminated and handled appropriately. The statement by the station manager, that no radioactivity was involved therefore needs examination. 

The pressuriser base is a solid support skirt, sitting on a concrete floor. Access to the electric heaters is from its underside and this presumably has a cover. The operation of the pressuriser is to maintain the pressure and temperature of the coolant by the electric heaters and water sprays, so as the base of the pressuriser is not visually accessible, it could be that the leak on its underside had been present for some time and undiscovered because the controls “masked” its occurrence. The support skirt of a replacement pressuriser manufactured by MHI Japan is furnished with inspection holes, presumably because a leak in the pressuriser it replaced at Fort Calhoun in the US was undiscovered for some time. 

Leaks in the coolant circuit are normally indicated by a white trail of boron from the boric acid neutron absorber added to the coolant. The approach to the “near-miss” incident at Davis-Besse in Ohio in 2002 went unnoticed for some time, possibly because the pressuriser control water sprays compensated for the loss of coolant and because the control rod “penetrations” were hidden by a cover. Even if the current repairs prove to be satisfactory and the pressuriser does not have to replaced, it may be worthwhile cutting an inspection hole in the support skirt, whereby tell-tale white trails can be seen.  

It does seem that the incidents of the loss of coolant and the fire are linked. 

John Busby 20 July 2010 (Revised 6 September 2010)