Characterization of two-phase flow in a typical metrological test flow loop.

Abstract

The understanding of the flow behaviour such as the flow regimes is important in multiphase flow metering for verification of the test meters especially during the reproducibility tests, as the meter could be transferred among different test flow loops or moved from one location to another within a flow loop. As the pipe geometry and configurations may vary for different testing laboratories and on the field, proper understanding of effect of geometrical variances on multiphase flow behaviour is deemed important for proper assessment of multiphase flow meter (MPFM) performance and as well developing testing protocols for commercial flow meters. To improve the performance assessment of MPFM, adequate understanding of the influence of pipe configurations on multiphase flow behaviour in a typical multiphase flow loop is important in order to design a flow loop for the purpose of calibration and validation of MPFM. To obtain this knowledge, a systematic study of flow characteristics transitioning from the horizontal to the vertical section in a typical MPFM testing installation with varying upstream and downstream configuration is needed to provide guidance on proper designing of MPFM calibration flow loop. To this aim, an experimental study was carried out in a typical MPFM flow loop which consists of 19.2 m long horizontal section followed by a 2.6 m long vertical section. All the sections are at industrial scale, being made of inner diameter (ID) of 0.077 m clear PVC pipe that allows for gas-liquid two-phase flow behaviour to be observed and determined. The alteration of upstream and downstream geometries of the flow loop are also carried out to investigate the effect of geometrical variances on the flow. Air and water are the fluids used for this study. The result of the study showed that the pipe configuration has significant effect on smooth stratified flow. The stratified flow regime observed in conventional straight pipe in horizontal section for low superficial velocities was observed to be absent in the present work. Instead, unstable wavy-slug (UWS) flow regime was observed. None of typical horizontal flow regime maps considered in this work were able to correctly predict UWS flow regime. The void fraction in the horizontal section was observed to be influenced by the pipe configuration due to liquid accumulation in the horizontal section. This could contribute to measurement uncertainties of phase fractions in the horizontal section. Analyses of the experimental results showed that no significant change in flow regimes was observed in the horizontal section with different development lengths of 100D and 200D (D is the pipe diameter) from the gas injection points. This suggests that a length of 100D may be sufficient development length for air-water two-phase flow in the horizontal section for such flow loop. Furthermore, more liquid accumulation is observed in 200D as compared with 100D case, which leads to lower void fraction in 200D development length. Downstream effect of the pipe configuration due to backward flow of the liquid phase was noticed to have significant effect on the flow structure in the horizontal section as observed in the probability density function (PDF) signature of the flows. The experimental investigation of effect of blind tee length on pressure fluctuation has shown that the 90-degree bend (equivalent to a blind length of 0D) has the highest-pressure fluctuation while the blind tee with 0.154 m clearance (2D length) has the lowest pressure fluctuation. The magnitude of pressure fluctuation is observed to be higher for intermittent flows than that of separated flows. The influence of blind tee length on pressure fluctuation tends to decrease with distance away from the blind tee in straight pipes. A set of guidelines for the MPFM test flow loop were proposed based on the outcome of the current studies.