Small leak during CGH2 delivery

A leak occurred when transferring gaseous hydrogen from a tube trailer to the well of a hydrogen underground storage test facility. The leak was located at a threaded fitting which was part of the connection piece between the site discharge panel and the hydrogen hose of the tube trailer, and was detected after opening of the tube trailer valve. The leak was audible and was confirmed by the alarms of the operators' portable 4-gas detectors.
The valve on the tube trailer side was immediately closed by the trailer driver. The injection was stopped and all the piping was purged with gaseous nitrogen and then vented.
The system had been tested with nitrogen at a pressure of 10 MPa, which is below the service pressure. This value was the available N2 pressure in the N2 bottle rack on site. The leak was determined at the fitting of the connection piece at a pressure level around 350 bar.
No injury and no material damage were to be reported.

The INITIATING CAUSE was the failure of a fitting at the connection to the trailer hose when the pressure reached 350 bar.
The ROOT CAUSE was a material failure, but it could not be detected before, because the system had not been tested above 100 bar. It was not possible to test at higher pressure, the limiting factor being the pressure in the nitrogen bottle rack available on site at that time.

connection, hose

This was the first attempt to transfer hydrogen to a underground salt cavern via a well.

DESCRIPTION OF THE FACILITY
The panel where the leak was detected controlled the injection to the well. Among other equipment, it consisted in a pressure reducer, pressure relief valve, shutdown valve, check valve, and measuring instruments.
The design pressure of the hose was 38 MPa, that of the connection piece 38 MPa. The final pressure in the well will be 15 MPa

No injury and no damage other than the high-pressure connection.

The existenceof the leak was checked by operators' portable 4-gas detectors.
The valve on the tube trailer side was immediately closed by the trailer driver. The hydrogen transfer was stopped, and all the piping was purged with gaseous nitrogen and then vented.

A technical reassessment of the transfer process was executed.
A conclusive helium tightness test was carried out with a compressor at 380 barg prior to a new attempt at injecting hydrogen.
The design of the connection piece was simplified so as to reduce the number of threaded fittings to just one.