Premature opening of a pressure relief valve

A CGH2 tube trailer was transferring hydrogen to a customer cylinder skid. The pressure relief on the customer system opened, the escaping hydrogen ignited and burned for approximately 10-15 seconds, causing a 30 ft (10 m) flame.
This occurred at the end of the filling process, the gauge pressure was approaching 2400 psi (166 bar) and the driver was proceeding to terminate the process. The safety relief valve vented before the driver could complete the process. The driver received burns while trying to close valves. Safety relief valve was designed to activate at 2400 psig (166 bar).

The INITIATING CAUSE was the premature activation of a pressure relief valve.
ROOT CAUSE could be attributed to a (i) defective component (the valve), or (ii) to a erroneous choice of the pressure value for the activation (erroneous design), or (ii) to the transfer procedure which was not allowing for uncertainty in the final pressure of the process (erroneous procedure). A human error cannot be excluded either.
The source does not provide elements for a choice.

Pressure Relief Valve or Device - Non-Reclosing

This incident could be attributed to:
(i) A defective component (the valve) which activated before its set pressure,
(ii) An erroneous choice of the pressure value for the activation (erroneous design),
(iii) The transfer procedure which was not taking into account the possible variation/ uncertainty in the final pressure of the process (erroneous procedure).
(iv) An error of the operator (the driver) who allowed the pressure to achieve the valve activation value.

The report does provide the data and the information necessary to clarify these options: for example, the exact value of the filling pressure in the system when the valve activated and the actual pressure in the trailer tank. The design of the customer tank and the position of the pressure relief system is unknown. This played an important role in the consequence, because the driver was near the activating valve when the release occurred. Could a better knowledge of the customer system and a better design reduce the hazards?
Moreover, it appears that the transfer process consisted in a series of manual steps. Could more automatic steps perhaps help reducing some of the hazards?