Signal-Based Identification of the Critical Liquid-Loading Condition in Gas–Liquid Two-Phase Flow

Accurate diagnosis of liquid loading in gas wells is hindered by inconsistent criteria for identifying the critical liquid-loading condition and by reliance on subjective observation during the development of physical models. To address this issue, controlled laboratory experiments were conducted to investigate pressure fluctuations in gas–liquid two-phase flow under different flow regimes, with the aim of establishing a quantitative criterion to identify such critical conditions. High-frequency pressure signals were collected and analyzed using complementary ensemble empirical mode decomposition (CEEMD). Characteristic parameters describing slug flow, annular flow, and the critical liquid-loading condition were extracted accordingly, including signal variance, intrinsic mode function energy entropy, and kurtosis. The results demonstrate that the critical liquid-loading state exhibits distinctive pressure fluctuation features compared with slug and annular flow regimes. Evidence is provided that, by integrating statistical indicators with fractal-based analysis, the proposed method enables reliable identification of the critical liquid-loading condition.