Liquid Holdup in Wet-Gas Pipelines

Abstract

Summary. An experimental study of two-phase flow was conducted to investigate liquid holdup in wet-gas pipelines. The liquid-holdup data were obtained by passing spheres through a 1,333-ft [406.3-m] -long, 3.068-in. [77.93-mm] -ID horizontal pipe and measuring the liquid volumes removed. Three different two-phase mixtures were used. The holdup data were compared with predicted holdup values and were used to evaluate a mechanistic model for stratified flow. None of the methods could accurately predict liquid holdup in this low-holdup region. Two new empirical liquid-holdup correlations for horizontal flow were proposed. The first is strictly for wet-gas pipelines (0 less than yL less than 0.35); the second is for any horizontal pipeline (0 less than yL less than 1.0). Introduction The simultaneous flow of gas and liquid in pipes is commonly encountered in the oil and gas industry. The complexity of this two-phase flow, however, has hindered the development of an accurate and completely theoretical model for predicting pressure drop and liquid holdup. As a result, numerous empirical correlations have been published and widely used. These correlations are usually published and widely used. These correlations are usually based on experimental data, taken in small-scale facilities operating at relatively low pressures and using different fluids from the ones encountered in the field. The accuracy of the correlations during simulation of behavior in large-diameter, high-pressure, and long pipelines is questionable. The liquid present from retrograde condensation in wet-gas transmission lines is usually small, leading to low liquid-holdup values. Accurate prediction of liquid holdup for these lines is not possible because all existing correlations were based on relatively high liquid-holdup data. Accuracy of the data diminishes rapidly as measured holdups approach zero. The inaccuracy is a result of the liquid-holdup measurement techniques used by investigators. In this study, a horizontal test loop 3.068 in. [77.93 mm] in diameter and 1,333 ft [406.3 m] long was used to investigate liquid holdup in the low-liquid-holdup region. Three different two-phase mixtures were used: kerosene/air, water/air, and water/surfactant/air. The liquid holdup was averaged over the entire length of the pipeline, because liquid was removed by pigging with rubber pipeline, because liquid was removed by pigging with rubber spheres. The collected data were used to evaluate the correlations of Dukler et al., Eaton et al., Beggs and Brill, and Mukherjee and Brill, as well as the Taitel and Dukler theoretical model for predicting liquid level in stratified flow. Experimental Program Fig. 1 is a schematic of the experimental facility used in this study. A brief description of the various components of the system and the testing procedure follows. A more detailed description was given in Ref. 6. Experimental Equipment. Air was supplied by a compressor rated at 800 Mscf/D [22.65 std m3/d] at 125 psig [861.8 kPa]. A pressure-regulator valve was located downstream of three connected storage tanks to maintain the pressure at 80 psig [551.6 kPa]. The air was then piped through one of two meter runs before going to the mixer. The temperatures were read from a thermometer installed upstream from the orifice. Kerosene and water were stored in two separate steel tanks. Each liquid had its own single-stage centrifugal pump with a 6,800-B/D [1081-m3/d] capacity. After the pump with a 6,800-B/D [1081-m3/d] capacity. After the liquid went through the pump, it was metered by either an orifice meter or a rotameter, depending on the flow rates. A quick-closing ball valve was provided just before the mixing tee. The two-phase test section consisted of a horizontal flow loop 3.068 in. [77.93 mm] in diameter and 1,333 ft [406.3 m] long. The length was measured from the pig launcher to the pig catcher. Special joints were used to ensure constant-diameter connections. Two transparent PVC sections located at 200 and 656 ft [61 and 200 m] from the pig launcher enabled visual observation of flow pattern pig launcher enabled visual observation of flow pattern and liquid level. At the outlet end, a pig catcher that consisted of a cylinder with baffle plates was used to catch the pig. Then quick-closing ball valves allowed deviation of the flow into either the separator or the weighing tank. Pressure transducers and pressure gauges were located at 14, 200, 656, and 1,299 ft [4, 61, 200, and 396 m] from the pig launcher and at the separator. The transducer signals were demodulated and recorded on an oscillograph. Direct pressure reading was also possible by use of a multimeter pressure reading was also possible by use of a multimeter with the demodulators. A tension load cell was used to weigh the amount of liquid in the weighing tank. This cell was connected to a digital electronic weight indicator. Testing Procedure. The data points were selected so that they would be evenly distributed on the Mandhane et al. flow-pattern map. For each test, air and liquid were allowed to flow at the desired flow rates through the separator until a steady-state condition was reached. SPEPE P. 36