Release from a CGH2 tube trailer during hydrogen transfer

A release of hydrogen occurred on a trailer during filling a stationary hydrogen storage tank.
A pressure relief device (burst disk) on one of the tube failed prematurely and released hydrogen. The alarm was given by a person walking by the facility, who heard the noise of escaping gas. Facility personnel were alerted and the tube trailer company dispatched an incident response team.
The team identified the source of the release on the burst disk of one of the tubes and closed the isolation valves on all tubes. The leaking tube was vented to atmosphere using a temporary vent pipe. After the leaking tube was completely vented, the trailer was disconnected from the gas distribution control of the stationary storage facility and moved out of the storage area
Post-incident investigation found out that the disc had ruptured. Evidences of corrosion or contamination were not found. Testing of other pressure relief devices revealed disc rupture at lower-than-expected pressures.

H2TOOLS report

THE INITATING CAUSE of the release was the premature opening of the burst disc.
The probable ROOT CAUSE was the choice of the burst disc material. Inconel #600 is characterised by a high hardness and tensile strength, which makes it more susceptible to stress corrosion cracking. According to H2TOOLS, the degradation of the burst disc was triggered by the combination of exposure to hydrogen and stress associated with the working pressure involved.

burst disc

DESCRIPTION og COMPONENTS
The trailer mounting 26 compressed hydrogen tubes. According to H2TOOLS report, each of them was equipped with an over-pressure-protection burst disc rated at a burst pressure of 4000 psi (275 bar). The disc material was made of Inconel #600 series material, which had been the predominant material at the time the tube trailer was last requalified.

STANDARDS and REGULATIONS
The transport of hazardous materials is regulated in the US by the 49 CFR, the Code of Federal Regulations regarding Hazardous Materials Transportation. According to the 49 CFR (recital §173.301(f)), “a cylinder filled with a gas and offered for transportation must be equipped with one or more pressure relief devices... The pressure relief device must be capable of preventing rupture of the normally filled cylinder when subjected to a fire test conducted in accordance with CGA C-14”.
The CGA C-4 provides the Procedures for Fire Testing of DOT Cylinder Pressure Relief Device Systems. The PRD is designed to operate with a burst disk in combination with a eutectic material which allows the disc to open only in the event of a fire, minimizing the possibility of an unintentional release. The burst disc is designed to burst at 5/3 of the container's pressure. The eutectic material is designed to melt at 100 °C (212 °F).
The US regulations prescribe only the pressure at which they must fail, according to the “DOT rating” of the tubes. This burst pressure must be 5/3 related to the nominal working pressure (service pressure) of the tubes to which the PRD’s are attached. In this case, the service pressure was 2400 psi (165 bar), so that the pressure relief device rating resulted in 2400 x 5/3 = 4000 psi ( 276 bar).

The only consequence was the release of some kg of hydrogen . Assuming a maxtotal transportatio ncapacity of 500 kg, the content of one of the 26 tubes of the trailer would be less than 20 kg

(i) There has been no specific industry guidance or standard prescribing the chemical composition of the pressure relief device (PRD) materials. The US regulations prescribe only the pressure at which they must fail, according to the “DOT rating” of the tubes.

(ii) Before the filling process, facility personnel had completed successfully all procedures and safety checks according to procedures. These included verification of leak. However, these leak detection before to starting the filling process, critical for identifying threaded connections leaks, could not have detected or predicted the premature failure of the burst disk.

(iii) The trailer involved in the incident was using a burst disk based upon a nickel base alloy designated as Inconel 600. All randomly selected PRD’s mounted on the tubes of the same trailer were tested and failed at a pressure below design specification. This shows that all Inconel 600 PRD’s were similarly degrading when exposed to hydrogen and mechanical (variable) stress. It was also found that all the PRD’s in service on the other hydrogen tube trailers of the fleet were using a different material, based on Carpenter 20 (also a nickel alloy).

All frangible disks on the trailer were replaced. Prior to placing tube trailer back into hydrogen service, all tube trailer components were examined for leaks using nitrogen at two succeeding pressures and standard leak detection fluid. A third and final examination was performed at full settled pressure before releasing the tube trailer back into service for the transportation of compressed hydrogen.