Influence of gas density on void fraction and pressure drop in upward vapor–liquid flow

Abstract

The influence of gas density on the void fraction, flow map, and pressure drop was examined for an upward vapor–liquid flow. The knowledge of the flow map is necessary to ensure a safe operational condition in a nuclear zone at high pressure. Moreover, the pressure gradient and its volumetric flow rate are related to the profitability of petroleum wells. Considering these scenarios, a modified cascade refrigeration cycle was constructed to generate and measure single-phase streams with different liquid and vapor density ratios. These single-phase liquid and vapor streams were mixed and injected into an upward vertical test section with an inner diameter of 26.64 mm and length of 5.8 m. The pressure ranged from 1.7 to 2.3 MPa, liquid–vapor density ratio ranged from 10.2 to 15.2, and total mass flux ranged from 733.1 to 882.9 kg.s−1.m−2. The void fraction, pressure gradient, and frictional pressure component were measured using quick-closing valves, temperature probes, and absolute and differential pressure sensors. The experimental data and transition line behavior of the flow map at high pressures in an upward two-phase flow as well as the proposed correction factors for the pressure gradient and its frictional component reflect the novelty of this study. Most existing studies have focused only on the influence of the gas density on the horizontal flow. A key observation of our study is that the liquid–vapor density ratio significantly changes the churn and annular transition lines. According to a vertical flow map found in the literature, the disappearance of the churn region can be foreseen and an annular transition occurs at lower gas velocities. The liquid–vapor density ratio does not influence the void fraction or slip velocity ratio; both increase with the superficial vapor velocity. Correction factors for the pressure gradient and its frictional component were proposed. The mean deviation for the pressure gradient was reduced from 27 % to 16 % after application of the correction factor. Half of the experimental points presented absolute relative errors lower than 30 % for the frictional pressure gradient. A cross-validation between the experimental data reported by other authors and the proposed correction factors was conducted to assess their response. In this cross-validation, 126 additional experimental points were used for the pressure gradient whereas 1270 points were used for its frictional component. Overall, the mean deviation of the pressure gradient was reduced from 92.7 % to 31.9 % after the application of the correction factor. Similarly, the mean deviation of the frictional gradient was reduced from 70.7 % to 31.3 % after the application of the correction factor. There was no evident influence of the liquid–vapor density ratio on the total pressure gradient and drift relationships for Froude numbers spanning from 2.6 to 7.4.