An explosion occurred on a new boiler set up inside a power plant (10 tonnes of steam/hr). This auxiliary boiler had been intended to complement the provision of steam required to reheat the heavy fuel oil used in the Unit 3 burner storage and cooling areas. This was a corrugated firebox tube boiler with 3 exhaust paths. Combustion gases were directed to the back of the boiler and then channelled towards the front via the lower smoke tubes before being conveyed to the chimney located in back via the upper tubes. This boiler was designed to function in a buffer capacity on the network, in parallel with another boiler of the same type (which had been shut down on the day of the accident) and with a set of steam transformers producing the tapping steam on the turbo alternators. The accident happened at the end of a series of boiler start-up tests that were being overseen by 1 technician employed by the product manufacturer assisted by 2 boiler room technicians. At the time of the accident, a firebox tube end broke apart from the tubular plate in creating an opening on the back surface of the boiler. The water contained inside the boiler, exposed to the action of instantaneous vaporisation of the pressurised steam (approx. 13 bar), escaped through this opening. The boiler was projected by the ensuing reaction some ten metres backwards, causing it to become entangled in the scraper of a 250-MW boiler. The steam escaping from the boiler crossed the material handling bay, blasted the mechanical workshop wall and then, by vaporising partially under atmospheric pressure, filled a much larger volume, causing burns to workshop personnel. The human toll of this explosion amounted to 1 death and 17 injured; all victims were working in the same mechanical shop. Even though for some design computation codes the characteristics of this boiler were not acceptable, nonetheless it was still compliant with ISO code rules and French Standard NFE 32.104.

Hydrocarbons heavier than water at the boiler operating temperature were present in the feed water; these hydrocarbons were then deposited onto the firebox tube. This would have caused the transition to vaporisation into film and thus a temperature rise in the metal, which had become hotter than the maximum warranty temperature relative to the type of steel used to build the boiler. The potential for polluting the vapour circuit by fuel oil during its reheating cannot be overlooked: during the recovery of vapour condensates, the fuel oil could have entered into the sumps used to supply the boiler. Since the actual operating conditions at the time of the explosion were not known with certainty, the presence of fuel oil in the feed water combined with extreme design characteristics triggered this accident.