Modeling the properties of hydrogen gas mixtures for predicting leakage parameters from distribution networks

Abstract

Simple engineering mathematical models have been developed and validated to determine the key thermodynamic properties that form the basis for improving the accuracy of predicting leakage parameters of hydrogen-natural gas mixtures under the conditions of gas distribution networks. To create a reliable benchmark database for model development, a comprehensive array of values was generated in the PVTsim Nova software package using the fundamental GERG-2008 equation of state. The study encompassed fifty base component compositions of natural gas, reflecting the variability of its quality in the real networks of Ukraine. The modeling was performed for absolute pressures from 0.1 to 1.3 MPa and temperatures from 243.15 to 323.15 K, which fully covers the operating range of distribution networks. The molar content of hydrogen in the gas mixture varied from 0 to 20%. Based on the obtained data, power-law regression equations were developed to calculate the compressibility factor, adiabatic index, and speed of sound as functions of pressure, temperature, the relative density of the mixture with respect to air, and the molar concentration of hydrogen. Validation of the developed models was conducted by comparison with the benchmark data, establishing the high accuracy and adequacy of the proposed equations throughout the entire investigated range. It was determined that for the compressibility factor model, the maximum absolute relative error is 0.44%, and the coefficient of determination is 0.995. For the adiabatic index, the corresponding values are 0.93% and 0.914; for the speed of sound – 1.5% and 0.998. An analysis of the error distribution confirmed the stability of the models and revealed only minor systematic biases for the adiabatic index and speed of sound, which depend on the hydrogen content. The proposed mathematical models have a simple analytical structure and do not require iterative procedures, allowing for their direct implementation into engineering methodologies for the rapid calculation of leakage parameters and production-technological expenditures and losses of gas in distribution networks.