Fuel cell forklift explosion

An explosion occured on a forklift, during operation at a company producing detergents, killing one and injuring several workers. The cause of the explosion was the sudden rupture of a glass fibres reinforced plastic composite tank used for hydrogen storage on board of the forklift.

The tank was defective, and the post-incident investigation showed that all installed tanks had design and manufacturing failures which could cause ruptures under service operative conditions well below the required safety margins. The major problem was the poor quantity and poor quality-control of the resin, but more importantly, , the tanks had not been properly certified according to applicable manufacturing standards, despite the claim of having passed all tests prescribed by the standard ISO/TS 15869. Finally, the investigation discovered that the tank supplier falsified documents to show that proper procedures were used to manufacture the subject cylinder, and critical records relating to the cylinder's manufacture were missing.

One of the conclusions of the investigation was that all 3,690 cylinders supplied and used worldwide posed an immediate safety risk.

The INITATING CAUSE was the abrupt rupture of the high-pressure tank. All the tanks supplied and mounted in the forklift fleets, failed at a pressure much lower than the required.
The conclusion of the court highlighted ROOT CAUSES such as wrong manufacturing and/or design and severe legal and managerial shortcomings.
According to the court report, “…the supplier concealed certain testing information and made a grossly negligent or intentional error in design calculations that led to its improper qualification, so that it appeared exempt from some testing and lead to a miscategorized of the tank.”
The conclusion of the court highlighted ROOT CAUSES such as wrong manufacturing and/or design and severe legal and managerial shortcomings.

forklift, fuel cells system, tank

The explosion was limited to the forklift, without consequence for the surroundigs.

The forklift company was sued by the victim’s family and by the company using them.
On its turn, the forklift company sued the tank supplier Worthington which agreed to pay for the cylinders’ replacements and ancillary costs to remove them from service.
The court found the tanks manufacturer guilty of, among others, of Negligent Design, Negligent Manufacturing, Negligent Misrepresentation and Fraudulent Misrepresentation.

US$ 5.5 million in replacement costs were paid, but the ligation continued on additional incurred costs.

This event touches on several important lessons. First, the critical importance to use good technical standards for the certification of materials and components and their quality control. Secondly, the important of an independent assessment of the results of the tests prescribed by the tests. By the event is also a story of the difficulties of the immature supply chains from suppliers for emerging hydrogen technologies.
In this incident several players were involved: a chemical company using fuel cell and hydrogen forklifts for materials movement in stores. The provider of these forklifts was assembling components from various suppliers. The tanks supplier was also relying on a subsidiary for the manufacturing of the composite tanks. This tanks manufacturer had selected the standard ISO/TS 15869 of the International Organization for Standardization, to assess and control the materials, design, construction, prototype testing and routine manufacturing inspections and tests for its tanks. The ISO/TS 15869:2009 - Gaseous hydrogen and hydrogen blends — Land vehicle fuel tanks – (later withdrawn) specified the requirements for cylinders intended for the on-board storage of high-pressure compressed hydrogen gas on land vehicles.

However, the company did not perform all required tests on the specific tank model. They had already done so on a bigger tank with the same characteristics, and the ISO standard stated that a tank could be qualified as a child of a parent design "Only when thickness changes proportional to the diameter and/or pressure change." The exploded tank had been classified as child of the bigger one, but this relied on a basic mathematical error in calculating the parent/child relationship of the subject cylinder.
By this simplified approach, the Accelerated Stress Rupture and the Extreme Temperature Pressure Cycling were not performed. They were designed to ensure that the cylinder maintains its throughout the operative lifetime and under all possible operative conditions. Extreme Temperature Pressure Cycling focuses on cycling performance at the boundary temperature of +85°C and -40°C. When the tank manufacturing company performed this test, upon request of the customer, the cylinder burst at 86°C having completed only 202 cycles of the required 11,250 cycles, which represent a 15-year life of use in commercial heavy-duty vehicles.

Another ISO/TS 15869 requirement regarded the composite resin. Resin glass transition temperature must be measured according to the standard ASTM D 3418 and the test results must within the manufacturer’s specifications. The tank manufacturer claimed that this value was 92°C. But it was not true, the value applied only to the hardener used to cure the epoxy resin. The manufacturer knew that the epoxy resin mixed with the hardener in the 84.4°C to 85.7°C. Using a resin with a glass transition temperature with less than one degree of margin from the 85°C temperature specified in ISO/TS 15869 is poor design judgment, a margin of 30°C is typically used. Moreover, the tank production was affected by extremely poor-quality control. The resin of the failed tank had a transition temperature of only 62 to 69°C. This poor performing resin was responsible for a strong reduction in tank strength at high temperature: a 72% reduction in strength of the composite was measured at 80°C.

Another lesson regards speculations of third, when data and details of an incident are not available. In a scientific article dedicated to the PEMFC, the authors advanced an hypothesis on the cause of the accident: “Although the cause of the forklift incident is not reported in the literature, it is likely that the temporal degradation of the PEMFC components of the forklift over a period of six years (2012e2018) was a root/contributing cause towards this incident.” In Ade et al., IJHE 45 (2020), https://doi.org/10.1016/j.ijhydene.2019.12.011
That this was a completely unfounded conclusion, is demonstrated by facts which became known later.

All the defective composite tanks were replaced. During the removal process, the forklift manufacturer advised all its customers using the defective cylinders to reduce the amount of hydrogen being placed in the cylinders and to operate the trucks at a lower capacity to help avoid a similar explosion.
This precautionary measure caused significant disruption to forklift users because it required more refills of hydrogen operate the FC forklifts, while these had been bought to replace the batteries forklifts and reduce charging times.