Data Analytics and Statistical Hypothesis Testing: Making the Difference between Passing or Failing a Wet Gas Multiphase Flow Meter in a Field Trial Test

Ruvalcaba Velarde, Salvador A. (Saudi Aramco) | Villegas Rodriguez, Ruben (Saudi Aramco) | Asiri, Mohammed A. (Saudi Aramco)


Abstract This paper presents the impact of using statistical hypothesis testing and data analytics to evaluate the performance of a wet gas multiphase flow meter (WGMPFM) against a test separator reference measurement. This study is part of a field evaluation in a wet gas operational environment. The outcome determines whether to approve or reject the WGMPFM for permanent installations and well testing. The methodology focuses on evaluating and ensuring the reliability of the test separator measurement, given that it is used as the reference against which the WGMPFM is compared for performance. The procedure goes through the use of exploratory data analytics for raw data consolidation, flow stability analysis and data validation. This includes quadratic order linear regression curve fitting of wellhead pressure-flow rates relationship. Furthermore, WGMPFM repeatability evaluation is explained for pseudostable flow in field conditions. In this study, it is shown that using only raw data, a WGMPFM fails a field evaluation against a test separator. However, after employing data analytics for validation of test separator reference values and after using statistical hypothesis testing for evaluating the repeatability of the WGMPFM, it actually performs within specifications. This paper provides a robust methodology for multi-phase flow meter performance evaluation in the field. The focus resides on measurement accuracy, reliability and repeatability, through the use of exploratory data analytics and statistical hypothesis testing. The robustness and reliability of reference measurement data in the field is critical, given that flow conditions and stability cannot be controlled as good as in laboratory flow loop tests. Introduction Test separators are the well testing conventional measurement systems, for which consistent industry standards have been developed to ensure their reliability. However, the flow measurement accuracy and reliability of this equipment is highly sensitive to several factors including: well stability conditions, proper separation control, measurement devices calibration and factor determination, and human intervention. Nevertheless, it is common that the validity of the test separator data is not questioned after it was acquired for comparison purposes. This happens because it is assumed that the procedures followed in the field to ensure the stability and calibration of the reference measurement are comprehensive enough to make the process and metering instrumentation uncertainty negligible. However, in the field, a well cannot be considered as a controllable environment in which the process is in full steady state. A well usually flow under pseudostable conditions, where the process actually fluctuates several percentage points in terms of temperature, pressure, and flow rates. Also, the conventional test separators do not guarantee a near-perfect separation and complete process stability inside the vessel. The meters are calibrated against water and adjusted through rounded factoring values, which lead to uncertainty propagation in the reference. Hence, a test separator cannot be deemed as an unquestionable reference.

Duplicate Docs Excel Report

None found

Similar Docs  Excel Report  more

None found