Leak from a CGH2 stationary storage

A leak developed on a main connector to a pallet consisting of 12 hydrogen bottles, which had not been re-tightened for a few weeks.
The alarm, at at 5% of the LEL, went off , with automatic call directly to the local fire brigade, which came onsite. An employee with knowledge of the gas system got an automatic sms notification and also came onsite
The first responders waited until the gas concentration went down, dispersed by the wind, and then closed the connection.

The INITIATING CAUSE was the release of hydrogen from a connection not properly tightened.
The ROOT CAUSE was probably the lack of effective regular inspection. It it unclear if the fact that the leack occurred during the hottest day of the year could have plaid a role.

connection, cylinders

No damages, just unknown quantity of lost gas (it can be evlauted at less of 8 kg, corresponding to a scenario with all bottles full, relasing almost everything) .

Thanks to an interlock, the triggering of the alarm sent automatically a all to the fire brigade and to lab responsible.
In absence of a fire, the intevention strategy consisted in leaving the hydrogen to disperse and to wait for the conrentrtion to go down below the safety threshold.

The safety systems worked as designed, and the event is at the border between a small (release) incident and a near miss.

The leaking connection was a bottle regulator connection with a PTFE sealing ring. Since the PTFE may become loose elasticity during the time it is compressed, the loss in thickness was compensated for by re-tightening the connection, with a specific torque value.
The unintended release showed that 3-4 weeks or more as interval between re-tightening was a too long period.

The hydrogen storage procedure was modified, foreseeing re-tightening of the connections every week.