Leak in a dispenser at a mobile HRS

A leak occurred while refuelling a fuel cell vehicle with hydrogen at a mobile hydrogen refuelling station. The hydrogen gas leak detection alarm (at 1% LEL) on top of the dispenser body went off, triggering the automatically depressurisation of all the pipes and discharge of the hydrogen through the vent.

A portable hydrogen gas detector was used to locate the leak, unsuccessfully. The next day, a more detailed search was conducted in the presence of the equipment manufacturer. The leak was discovered at the shut-off valve inside the hydrogen dispenser. There were no injuries or property damage.

The incident investigation found that the shutoff valve in the dispenser was made of EPDM (ethylene-propylene diene rubber), which had a minimum operating temperature of -40°C, the same as the normal operating temperature of the dispensed hydrogen.
The lower part of the shutoff valve was maintained, therefore the ring was constantly working at the lower limit of its temperature range specifications. On the day of the leak, the first three vehicles were filled without any problems, but a leak occurred in the fourth vehicle. It is plausible to assume that the leak occurred because filling multiple vehicles created harsh operating conditions.

The incident investigation found that the shutoff valve in the dispenser was made of EPDM (ethylene-propylene diene rubber), which had a minimum operating temperature of -40°C, the same as the normal operating temperature of the dispensed hydrogen.
The lower part of the shutoff valve was maintained, therefore the ring was constantly working at the lower limit of its temperature range specifications. On the day of the leak, the first three vehicles were filled without any problems, but a leak occurred in the fourth vehicle. It is plausible to assume that the leak occurred because filling multiple vehicles created harsh operating conditions.

The INITIATING CAUSE of the leak was the deterioration of a rubber o-ring.
The ROOT CAUSE was a design shortcoming, according to which the O-ring was expected to work at a temperature at the limit of the manufacturer’s specification.

O-ring, shut-off valve

The HRS had started operation 8 days before. The leak occurred while the fourth fuel cell vehicle of the day was being filled with hydrogen gas (-37°C, 63MPa). Ten vehicles in total had been refuelled since the start of the operations).

DESCRIPTION OF THE FACILITY
The mobile HRS was delivering 70 MPa. The hydrogen in a hydrogen cartridge was pressurised by a hydraulically driven gas booster and stored in a medium-pressure storage at 40 MPa and a high-pressure storage at 82 MPa. The refuelling of the vehicles at the dispenser was occurring by differential pressure. The dispenser was equipped with hydrogen precooling able to deliver at -40 C. The HRS capacity was 69,107 m³/day (approximately 6 kg/d).

This is case at the border between a minor incident and a near miss. The safety systems in place worked as designed with no consequence to human or installation, except the time lost during the shut-down and the neeed to change the design.

The safety systems in place worked as designed with no consequence to human or installation

(1) Due to the frequent occurrence of leaks at joints of screw-type and flange-type, it is desirable to minimise as few joints as possible in the high-pressure gas components of a hydrogen refuelling stations.
(2) When using O-rings in the sealing parts of high-pressure gas equipment at hydrogen stations, it is important to select O-rings with appropriate sealing properties, considering the fluctuations in temperature and pressure during operation.

The ice pack around the shut-off valve was removed.
(1) Since the replacement of the O-ring with one of a different material able to resist lower temperatures had already been tried unsuccessfully at HRS, the whole shutoff valve was replaced: in place of a plug type valve, an integrated shut-off valve was used, which does not use a ring.