Hydrogen explosion on the hydrogen storage facility of a pharmaceutical plant

The event early on a Saturday morning, when the plant was thus practically empty. Nevertheless, the chronology of the accident can be reconstructed from the testimonies of the few operators who were present at the time:
Phase 1: Gas leak - Loud whistling from the hydrogen storage facility could be heard , lasting 20 to 40 seconds.
Phase 2: Explosion - The released hydrogen cloud ignites. A very bright, red-orange fireball measuring 15 to 20 m in diameter was visible for 3 minutes. Witnesses in a common room, located 100 m away from the pit, felt the shock wave and the heat from the blast. The hydrogen in the associated tanks also burned completely.

The INITIATING cause of the event was the sudden mechanical failure of the storage cylinder due to hydrogen embrittlement.

The root cause is probably to be found in operation and organization dimensions, but the company had recycled the cylinder without any documentation of its previous history. Possibility of embrittlement and the fatigue stress due to pressure cycling had not been taken into account. Thus also lack of a properly performed risk assessment.

gaseous hydrogen storage, steel cylinder

The affected tank had been recycled, but its previousoperative history was unknown.

The blast effect of the explosion resulted in property damage:
1) Broken windows and store fronts up to 515 m away;
2) Damage to building cladding, walls or ceilings up to 350 m away;
3) Displaced roofing tiles up to 280 m away.

Thermal effects of the fireball were virtually non-existent:
4) Burn marks on supports, cylinders and the metal uprights of the storage tank roof;
5) Temperature rise of the dry foam lining the wall of the nearest workshop

Ten or so cylinder fragments were found after the accident: 6 in the pit and 2 outside the plant (up to 150 m away). One of these fragments severed the compressed air line supplying the storage facility's isolation valve.

Propelled to the side opposite its rupture, the exploding cylinder collided the pit's east wall, and then flew back in the opposite direction.

Several factors contributed to the hydrogen embrittlement of the steel in this accident:
A) The pressure contributing to the hydrogen sorption phenomenon on the surface of the steel;
B) The purity of the hydrogen gases used could even be a determining factor that triggers the hydrogen embrittlement phenomenon. The gas used on the site had a very low oxygen content (0.5 ppm) which can aggravate the hydrogen embrittlement phenomenon. Oxygen helps inhibit this phenomenon;
C) The level of stress of the cylinder's casing against its support structure, the cylinder being used as an accumulator;
D) The mechanical performance characteristics of the cylinder's steel;
E) The lack of information related to the previous operative history (age, cycles numbers, pressure ranges) of the tank.
F) The operating conditions for a pressure vessel, particularly when used as an buffer tank, should be designed and assessed considering fatigue phenomena.