Analysis and nomograph development for a leaky pipeline carrying plug flow based on numerical modeling and experimental validation.

Abstract

Leak detection in pipelines transporting gas-liquid multiphase flow remains a challenging task due to complex and inherently unsteady flow behavior, particularly under intermittent regimes such as plug flow. Conventional leak detection techniques, which are well established for single-phase flow, often suffer from reduced sensitivity and false alarms when applied to multiphase systems. Motivated by these limitations, the present study investigates leakage characteristics in a horizontal pipeline conveying gas-liquid plug flow under underwater conditions. A three-dimensional transient numerical model based on the Volume of Fluid (VOF) approach is developed to simulate various leakage scenarios, including different discharge sizes and gas-liquid superficial velocities. The numerical results are validated against experimental data obtained from a dedicated multiphase flow loop. Time-series pressure signals are analyzed using statistical metrics, probability density functions, and continuous wavelet transform techniques to assess their effectiveness in identifying leakage occurrence. Furthermore, a non-dimensional analysis is employed to develop a nomograph for estimating gas release velocity from underwater leaks. The results demonstrate that leakage significantly alters pressure fluctuation characteristics and gas void fraction distribution, with detectability strongly influenced by flow conditions and discharge size. The proposed nomograph predicts gas release velocity with reasonable agreement relative to experimental measurements, highlighting its potential applicability for subsea leak assessment and process safety analysis in multiphase pipeline systems.