Composite hydrogen tanks rarely burst. When hydrogen escapes or burns, the cause is more often something attached to the tank: a pressure relief device, a fitting, a plug or a vent line. So the first job in any investigation is to establish which part failed, before anyone forms a theory about the tank itself. These are the questions we work through first.
1. Was it the tank, or something attached to it?
Two well-documented incidents show why this comes first. In February 2018 in Diamond Bar, California, hydrogen vented and ignited on a tube trailer carrying 25 Type 3 carbon composite cylinders. The NTSB found that a requalification technician had installed a pressure relief device with the wrong rating, which opened at normal transport pressure, and that loose vent line fittings let the fire spread. None of the cylinders lost mechanical integrity. In June 2019 at Kjørbo, Norway, a hydrogen station exploded because the inner bolts of a plug on a steel storage tank had not been adequately torqued. Neither event began with a composite overwrap failing.
2. Did the pressure relief device work, and was it the right one?
Thermally activated pressure relief devices (TPRDs) are what stop a tank from rupturing in a fire. Investigators check the part number and rating against the design, whether the device activated, and where the fire was. A fire that heats one part of a tank, away from the TPRD, is the difficult case: research for NHTSA (DOT HS 811 303, 2010) found that in localized fires a cylinder can be weakened to the point of rupture before the TPRD activates. The engulfing bonfire test long required by CNG standards does not reproduce that, which is why standards bodies drafting hydrogen requirements asked for a localized fire test.

3. What happened to the liner?
In a Type 4 tank the plastic liner holds the gas and the carbon overwrap carries the load. The liner has failure modes of its own. In NHTSA laboratory cycling of 70 MPa tanks (DOT HS 811 832, 2013), Type 4 liners blistered and pulled away from the end boss at elevated temperature, and none of it was visible from outside or with a borescope before the tanks were sectioned. Liners can also buckle when a tank that has been holding hydrogen is fully vented to atmospheric pressure, as can happen when a relief device opens. CT scanning can show liner collapse without cutting the tank, so it belongs early in the inspection protocol, before any destructive work.

4. What did the fill and temperature history look like?
Vehicle fueling under SAE J2601 is designed to keep gas temperature between −40 °C and 85 °C, with a maximum dispenser pressure of 87.5 MPa for a 70 MPa tank. Station fill logs, ambient conditions, the number of fills and any fast depressurization events all help explain liner damage, leaks or overheating. For tube trailers and stationary storage, the fill and offload records play the same role.
5. What do the overwrap and fracture surfaces say?
If the composite did fail, the fracture surfaces and fiber break patterns record where the failure started and how it ran: impact damage, a cut or abrasion, chemical attack, heat, or long-term stress rupture of the fibers. That record is fragile. Fragments must be collected, kept dry and handled as little as possible until all parties agree how to examine them. Our note on what to preserve after a cylinder failure covers the first steps.
6. Which standard was the tank built and approved to?
The answer sets the expected performance and the service life. Vehicle tanks are typically qualified to ANSI/CSA HGV 2, SAE J2579, UN GTR No. 13 or ISO 19881. Transportable cylinders fall under DOT rules in the United States, either as UN pressure receptacles built to ISO 11119-3 or under a DOT special permit, which has commonly been the route for large composite tubes on trailers. The label, the approval documents and the requalification records tell investigators what the tank was supposed to withstand, and whether it was still within its permitted life.
Investigating a cylinder failure?
Contact us before any destructive examination or testing. We can help agree an inspection protocol, preserve the evidence and scope the failure analysis.
This is general technical guidance, not legal advice. Every matter depends on its own facts and evidence.
Sources
- NTSB: Hydrogen release and fire, Diamond Bar, California (HZM-19-02)
- Nel: Status and Q&A regarding the Kjørbo incident
- NHTSA DOT HS 811 303: Localized fire protection assessment for vehicle compressed hydrogen containers
- NHTSA DOT HS 811 832: Cumulative fuel system life cycle and durability testing
- US DOE: Introduction to SAE hydrogen fueling standardization (J2601)