Scholarlink

<p>To address the extreme operating conditions of the Keshen ultra-deep, ultra-high-pressure gas field—characterized by high gas volume fractions (GVF), low liquid yields, and highly corrosive fluids—this study proposes an optimized gamma-ray/Venturi Multiphase Flow Meter (MPFM) model with localized parameter corrections. The primary innovation lies in a strongly coupled iterative algorithm integrating gamma-ray and Venturi measurements under ultra-high-pressure conditions (up to 30 MPa). Unlike conventional low-pressure approaches that treat subsystems independently, a dynamic closed-loop framework is established to capture the non-ideal behavior of high-pressure gas, whose density approaches that of liquids. The preliminary mixture density obtained from gamma-ray measurements is introduced into the Venturi model, where real-time thermophysical properties are calculated using the AGA8 Equation of State. These properties are used to determine the expansion factor and throat velocity. A high-pressure slip model then evaluates the flow regime and iteratively corrects the true phase fraction. This coupling mechanism effectively mitigates distortions arising from interactions between thermodynamic parameters (e.g., density and isentropic exponent) and hydrodynamic factors (e.g., slip ratio). By synchronously updating the expansion factor and a Weber-number-based slip factor via an Extended Kalman Filter (EKF)-based high-frequency 'compute-then-average' architecture, the proposed method circumvents mathematical rectification errors (Jensen's inequality) induced by transient slugging. Quantitatively, this coupled iteration suppresses prediction residuals profoundly; at 30 MPa, it curtails the Root Mean Square Error (RMSE) of the mass flow calculation from an uncorrected baseline of 129.8 kg/h to 66.7 kg/h, realizing a rigorous 48.6% reduction compared to traditional open-loop models. System reliability was evaluated using the GUM uncertainty framework and small-sample Bland-Altman statistical inference. Rigorous Monte Carlo simulations under the GUM framework demonstrate a relative expanded uncertainty of U = ± 3.0% (k = 2) at 30 MPa. Field validations confirm that the model’s penalized confidence intervals align with the standard API ± 5% tolerance limits of reference separators, establishing a robust, self-calibrating solution for ultra-high-pressure environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimization of a high-pressure multiphase flow metering model for ultra-deep gas fields based on dynamic PVT and slip correction

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月21日

查看全文
  • Guoqiang Han,
  • Qianliang Xue,
  • Xue Tian,
  • Yujie He,
  • Tong Zhou,
  • Bingcheng Ning

摘要

To address the extreme operating conditions of the Keshen ultra-deep, ultra-high-pressure gas field—characterized by high gas volume fractions (GVF), low liquid yields, and highly corrosive fluids—this study proposes an optimized gamma-ray/Venturi Multiphase Flow Meter (MPFM) model with localized parameter corrections. The primary innovation lies in a strongly coupled iterative algorithm integrating gamma-ray and Venturi measurements under ultra-high-pressure conditions (up to 30 MPa). Unlike conventional low-pressure approaches that treat subsystems independently, a dynamic closed-loop framework is established to capture the non-ideal behavior of high-pressure gas, whose density approaches that of liquids. The preliminary mixture density obtained from gamma-ray measurements is introduced into the Venturi model, where real-time thermophysical properties are calculated using the AGA8 Equation of State. These properties are used to determine the expansion factor and throat velocity. A high-pressure slip model then evaluates the flow regime and iteratively corrects the true phase fraction. This coupling mechanism effectively mitigates distortions arising from interactions between thermodynamic parameters (e.g., density and isentropic exponent) and hydrodynamic factors (e.g., slip ratio). By synchronously updating the expansion factor and a Weber-number-based slip factor via an Extended Kalman Filter (EKF)-based high-frequency 'compute-then-average' architecture, the proposed method circumvents mathematical rectification errors (Jensen's inequality) induced by transient slugging. Quantitatively, this coupled iteration suppresses prediction residuals profoundly; at 30 MPa, it curtails the Root Mean Square Error (RMSE) of the mass flow calculation from an uncorrected baseline of 129.8 kg/h to 66.7 kg/h, realizing a rigorous 48.6% reduction compared to traditional open-loop models. System reliability was evaluated using the GUM uncertainty framework and small-sample Bland-Altman statistical inference. Rigorous Monte Carlo simulations under the GUM framework demonstrate a relative expanded uncertainty of U = ± 3.0% (k = 2) at 30 MPa. Field validations confirm that the model’s penalized confidence intervals align with the standard API ± 5% tolerance limits of reference separators, establishing a robust, self-calibrating solution for ultra-high-pressure environments.