Explosion in the hydrogenation reactor of a food products plant

Explosion in a food products plant (Seveso classified)

The event occurred in the settling cyclone of a hydrogenation reactor of a starch production unit.
The process consisted in introducing the starch hydrolysate mixture into the reactor with a catalyst (Raney nickel), under agitation and 50 bar of hydrogen at 130°C. The reaction lasted three hours, after which hydrogenated mixture was racked and decanted before purification.
The explosion occurred during the discharge of the reactor into the settling cyclone. These operations took place 6 to 8 times a year.

The ARIA report (see references) does not state with certainty the cause(s), but mentions the opinion of the plant operator, that the explosion was possibly due to the presence of hydrogen in the settling tank. This would indicated the existence of a leak due to a welding defect in part of the reactor containing the gaseous phase.

The INITIATING cause, assuming correct the analysis of the plant operator, was the hydrogen leak from the reactor to the cyclone through a welding degradation.

starch hydrogenation reactor, settling cyclone

DESCRIPTION OF THE PROCESS
A settling cyclone is a type of dusk collector/removal.
(Hydrogenated) starch hydrolysates are mixtures of several sugar alcohols.

The ARIA report (see references) does not mention improvement actions undertaken after the explosion.

Assuming correct the proposed explosion cause, a hydrogen leak via a (growing) welding defect, the most logic lesson learnt would be to improve inspection procedures for a better control of welding and assessment of the overall tightness of the hydrogenation reactor.