Advanced RANS model for simulating high-pressure gas leaks and dispersion dynamics

Abstract

This article reports three CFD models, RANS-based analysis, were improved to represent the primary phenomena associated with pressurized gas leakage in pipelines. These models were benchmarked against experimental data from the literature, focusing on depressurization, expansion, and dispersion. The enhancements to the RANS method included using a patch tool to separate the pipeline boundary from the atmosphere, allowing specific initial conditions to be assigned to mesh elements; automatically mesh adaptation to capture regions with significant velocity and pressure gradients accurately; implementing a source term in the transport equation to account for buoyancy effects related to density changes lied to the high-temperature gradient typical present in the referred phenomena — that was crucial for the Peng–Robinson real gas equation effectiveness. Qualitative and quantitative analyses reveal that enhanced models closely matched experimental results, demonstrating significant improvements over classical approaches, with maximum average percentage differences of 12.87 %, 90.52 %, and 53.10 % for depressurization, expansion, and dispersion, respectively. The delineated method provides a valuable approach for rapidly creating a database to predict potential outcomes of pressurized pipeline leaks.