Title : Predictive polynomial mapping framework for hydrogen mass flow rate characterisation in full scale major hazard testing
Abstract:
The precise measurement and prediction of gas mass flow rate is essential in the oil and gas industry, particularly as the sector transitions towards net zero. Hydrogen and hydrogen-blended gases offer significant opportunities to decarbonise hard-to-electrify energy sectors such as manufacturing and transport. However, hydrogen’s wide flammability range and low ignition energy requires accurate characterisation of the flow behaviour, particularly in the scenarios involving releases, ruptures and velocity design.
This paper presents a nonlinear predictive method for the determining the mass flow rate of gaseous hydrogen using polynomial mapping. By using thermodynamic property data generated from real-gas equations of state, polynomial representations were constructed of nonlinear hydrogen characteristics, such as density, viscosity, and compressibility across a wide range of pressures and temperatures relevant to real-world experiments.
The method was evaluated against and extensive experimental programme involving the controlled release of hydrogen in the scenario of a high-pressure pipeline rupture. Results showed that the polynomial-mapping approach accurately captured nonlinearities in flow conditions and prediction errors were typically within acceptable values calculated by measurements.
Finally, the technique is being applied to live mass flow measurement using orifice plate and pitot-static assemblies for another experimental test programme with hydrogen. This test programme is designed to determine the correlation between gas velocity and pipe debris transportation, erosion and noise and vibration. Therefore, accurate and real-time characterisation of the mass flow rate across a range of pressure tiers is essential for the accurate quantification of these effects. Early findings from the velocity design experiments reveal that the model provides stable real-time estimates without the need for repeated calls to full thermodynamic solvers. Overall, the results to date demonstrate the potential for polynomial mapping as a computationally efficient and practical tool for hydrogen mass flow prediction in safety critical applications.
