Subscribe to our newsletter to receive the latest news and events from TWI:

Subscribe >
Skip to content

Monitoring systems for safe usage of hydrogen power

TWI is part of a consortium that has completed a two-year collaborative project that successfully developed a condition monitoring technique to improve the safety of hydrogen fuel storage tanks.

The SafeHPower project, funded by the European Commission, addressed the need for safe hydrogen (H2) fuel storage tanks by successfully developing two prototypes based on the acoustic emission (AE) technology. The prototypes were designed to continuously monitor tanks throughout the entirety of the storage, distribution and usage phases in the hydrogen transport economy, ensuring that all defects that could cause eventual leaks are detected in time to avoid catastrophic failures.

The project also included the development of a portable neutron radiography system to conduct imaging of tanks at all points in the H2 supply chain. However, the prototype was unsuccessful in imaging tank flaws using the neutron flux from a portable deuterium-tritium generator and a real-time digital imaging system, so the system is not included in this case study.

Hydrogen: a fuel for the future

Hydrogen is a promising alternative to fossil fuels whose benefits include abundance, efficiency, a low carbon footprint and absence of other harmful emissions. However, there are safety issues surrounding its use, especially when stored at the high pressures required for it to be economical. Higher pressure means increased risk of tank rupture during storage or transportation.

Hydrogen can easily pass through interstices between atoms of many structural materials. In metals, this can cause degradation by a process called hydrogen embrittlement. Tanks also experience continuous cyclic loading between ambient and storage pressure, which stresses the tank material. Additional, residual stresses may also remain from manufacture, and this mix of stresses can result in the development of fatigue cracks. The combination of fatigue cracks and the process of hydrogen embrittlement become significant over time, resulting in accelerated crack growth rates. Several accidents and injuries have been caused by this going undetected in the past.

Storage tanks made from a composite material lined internally with a polymer (Type IV) or metals such steel or aluminium (Type III) are generally used in vehicles, whereas tanks for stationary storage and transport applications are more commonly made from steel (Type I).

SafeHPower project activities

1. Miniature AE sensor system for continuous monitoring in composite tanks (Type IV)

The SafeHPower project developed a miniature AE sensor prototype for continuous monitoring and detection of fatigue cracks in Type IV fuel tanks for vehicles. The prototype comprises an AE sensor, control box and embedded software. The system was designed and developed with several features to enhance its suitability for installation in a vehicle:

  • All power needs are satisfied by the vehicle’s battery.
  • The control box can act as a Faraday cage to provide protection against electromagnetic interference produced by the external electronic components, as well as being resistant to water and dust particles.
  • The embedded software can permanently monitor all AE events and makes a yes/no decision to activate a vehicle dashboard warning.

Successful laboratory and field trials have been conducted on composite Type IV and Type III tanks respectively. During the field trials, the AE sensor prototype experienced conditions representative of real-world operation in terms of temperature and exposure to water, dust particles, humidity and vibrations. The results of these field trials showed that the fuelling cycle generates little AE activity (because the tank on the vehicle did not have any defects) compared to the laboratory tests which showed significantly higher activity (because flaws were manually introduced in these tanks).

Figure 1 AE sensor system installed under the vehicle
Figure 1 AE sensor system installed under the vehicle
Figure 2 Typical AE signal detected using the miniature AE sensor system
Figure 2 Typical AE signal detected using the miniature AE sensor system
Figure 3 AE events from a healthy vehicle tank (red circles show the periods of refuelling marked by sharp increase in the AE events)
Figure 3 AE events from a healthy vehicle tank (red circles show the periods of refuelling marked by sharp increase in the AE events)

mailto:membership.dutymgr@twi.co.uk 2. AE array system for continuous monitoring of large steel tanks (Type I)

The project also developed an AE array system to continuously monitor large storage steel tanks and, crucially, locate the position of any defects with high accuracy. It comprises an array of sensors (three or more), a PXI platform to provide a basis for building electronic test equipment, automation systems and modular instruments, and advanced signal processing software. This system is capable of providing an early warning of fatigue cracks and hydrogen damage before a defect becomes critical.

The main feature of the system is the bespoke AE signal processing firmware, which had to be designed to cope with a large volume of data due to there being several sensors in the array and a need to monitor longer load cycle periods. Advanced signal processing tools investigate and cover different aspects of the problem. The developed software operates in stages, by performing data loading, normalisation, analysis, thresholding and triangulation. A graphical user interface allows the operator to set up various parameters, view the AE signals from different sensors and locate defects with high accuracy.

Laboratory validation trials in which water (instead of hydrogen) at a pressure of 220 bar was stored in a tank containing a flaw showed that the developed AE array system is capable of detecting AE events generated by crack growth (50x2x1mm3 and 50x0.5x5mm3). The triangulation method used by the software successfully located the flaw.

Field trials were conducted by installing an array of six sensors, for one month, on a large stationary steel tank with a capacity of 50m3. The results obtained from these tests were comparable to the laboratory results obtained on a sample with flaws.

Further work

The AE prototypes are robust, with a technology readiness level of five to six. However, several improvements and refinements are planned. These include improved communication capability between the AE sensors and the vehicle control system, to provide information about when the tank is being pressurised. This would make the AE system more reliable, as it would avoid false alarms produced by AE sources other than crack initiation and propagation. Causes of such false positives include tank vibrations that make the tank impact another component, or water drops, which can each generate similar AE events to cracking. This extra input could also provide information about the start and final pressure during the refuelling, allowing the results of AE activity during the pressurisation to be normalised according to the increase in pressure.

The consortium believes that after one more year of field trials, the AE products will be ready for commercialisation in 2017. The extended trials would also generate data to that would allow identification of the exact patterns of defective and non-defective tanks, which would in turn enable adjustments to the software to initiate the alarm process with a higher degree of accuracy.

For more information on SafeHPower visit the project website at www.safehpower.eu or email contactus@twi.co.uk

Figure 4 One of six AE sensors placed on a large H2 storage tank
Figure 4 One of six AE sensors placed on a large H2 storage tank
Figure 5 Bespoke defect localisation map showing the location of the six sensors (blue) along with the estimated defect locations (red)
Figure 5 Bespoke defect localisation map showing the location of the six sensors (blue) along with the estimated defect locations (red)
Avatar Malini Vieyra Senior Project Leader – Non-Destructive Testing

}