Hydrogen explosion in a unit of a nuclear power plant.

The explosion occurred while maintenance personnel were grinding a high-pressure injection (HPI) pipe, which had been recently cut at approximately 18 inches from the nozzle connecting the HPI pipe to the reactor coolant system (RCS) piping.
The craftsmen, who were grinding on the HPI pipe in preparation for welding, observed a bright flash at the outlet of the HPI line and heard a loud "bang". The craftsman actually performing the grinding was physically blown away from the HPI pipe a distance of about three feet. Personnel in other areas of the containment building heard the explosion and felt the resulting concussion and mechanical vibration. Additionally, some personnel outside of the containment building, including operators in the control room reported that they heard the explosion and felt varying degrees of vibration. Although there were no physical injuries as a result of this event, it should be mentioned that the craftsman's life was endangered as he was working on a scaffold that was over 30 feet high.

The reason for the presence of an explosive concentration of hydrogen is unknown. It could have been caused by (a) inadequate degassing, (b) failure to purge the HPI pipe with nitrogen, or (c) failure to temporarily plug the open HPI pipe.

Subsequent inspection of the affected HPI line, the first upstream check valve in the HPI line, and the corresponding nozzle and safe-end on the RCS cold leg indicated no signs of damage as a result of the explosion.

reactor cooling system, pipeline

The accident occurred during maintenance.
The most recent Reactor Coolant System measurement of dissolved gas in reactor coolant had been taken on 15 days before, when commencing the plant cool down and shutdown. It indicated 39 standard cc of total gas/litre of coolant. The hydrogen concentration was 43% of this total. Because the total gas concentration exceeded the maximum allowed by plant procedures (30 cc/litre) for RCS depressurization, the RCS was degassed for approximately 14 hours as cool down progressed. The total gas concentration was not measured after degassing had been terminated because the RCS had been depressurized and the sampling method is effective only when the RCS pressure is greater than several hundred psig. Atmospheric samples had not been taken to measure hydrogen and oxygen concentrations in the vicinity of the open HPI pipe.