leak from a hose when refuelling a forklift
The incident occurred when refuelling with hydrogen a fuel cell forklift. The hydrogen was provided by a mobile refuelling station. The filling of a second forklift of the day had just begun, when a loud noise was heard and the filling hose broke, releasing hydrogen (approximately 0. 14 m3). No ignition occurred. The excess flow prevention valve was triggered. An operator pressed the emergency stop button on the control panel. The amount of released hydrogen was negliglbe, i nhe order of magniture of 10 g. The accident caused a damage to the filling hose, coupling, and the filling port of the FC forklift, with a total damage of a million of yen. No injury occurred. The damaged filling hose had been used 2,334 times and had been in use for approximately three years, from December 2016 to September 2019 (December 2018 to June 2019), including a suspension period until September. An investigation by the manufacturer revealed wrinkles and abrasions due to bends on the outer surface of the filling hose. These damages were not limited to the area affected by the burst. The inner tube had fractured circumferentially.

The INITIATING CAUSE was the loss of confinement of the filling hose due to the rupture of the inner and outer layers. The investigation found that part of the damage of the hose was coming from its storage during the movement of the mobile refuelling station: wrinkles were caused by contact with the hook of the filling cart used for storage and by the weight of the hose. Scratches were produced by contact with the storage door, valve, instruments, etc. due to vibrations while driving. Another occasion of damage was the refuelling cycle. When the hose was attached and removed by holding only the hose during filling, the weight of the nozzle caused a kink in the filling hose. Repeated bending weakened the strength of the bent area, which was pressurised while in a deformed state. This resulted in a non-uniform pressure load and damage to the reinforcing layer. Based on these findings, it can be concluded that the ROOT CAUSE was related to design shortcoming. It is not clear which role the spring itself plaid in the generation of (or protection from) damage. The durability test was conducted without the outer spring, because it does not contribute to the internal confinement capacity of the hose. Finally, lack of regular inspection did not allow for an early identification of the damages.