Blending hydrogen into the natural gas (NG) network could provide an efficient pathway for decarbonising the NG system through power-to-gas technologies. However, due to the presence of potentially multiple and intermittent hydrogen injection sources, the gas blended throughout the network would be neither homogenous nor at a constant mole fraction. The above features are not captured by the current transient modelling techniques. To bridge this gap, this work presents a transient analysis model that enables the tracking of gas compositions and particularly hydrogen fractions in real-world meshed networks with multiple NG sources, non-pipe elements, and multiple and intermittent hydrogen injection sources. A time-varying compressibility factor is also introduced to account for the variable gas composition across the network. Moreover, numerical techniques are adopted for improving the stability of the Eulerian numerical calculation, and a specific grid size threshold Δxmax is introduced for selecting the stable mesh grid to alleviate convection-dominated oscillations caused by the hydrogen fraction tracking. The case study based on the well-known 20-node Belgian gas network validates the effectiveness of the method in solving practical-scale problems, whereas the unsuitability of steady-state models is also discussed and highlighted. The results clearly demonstrate the effect and importance of introducing variable compressibility factor, hydrogen fraction tracking, and variable gas demand. The impacts of hydrogen blending on pressures and linepack of the network are further investigated.
Read full abstract