Fracturing/pressure pumping

Study Models Onset of Liquid Loading in Large-Diameter Deviated Gas Wells

The complete paper describes an experimental study performed to investigate the onset of liquid loading in a 6-in. production casing at various inclination angles.


Multiple challenges related to liquid loading have been observed during flowback after hydraulic fracturing and during the production phase. These challenges are further aggravated by high-inclination angles found in deviated wellbores. The complete paper describes an experimental study performed to investigate the onset of liquid loading in a 6-in. production casing at various inclination angles. A unified mechanism model for the onset of liquid loading is developed for a large-diameter production casing.

Previous Studies on Onset of Liquid Loading

Multiple experimental studies can be found in the literature that have successfully determined the onset of liquid loading. However, most of these experiments have been conducted in pipes with small diameters. A summary of the experimental studies for two-phase pipe flow along with the defining parameters is shown in Table 1 of the complete paper.

During these previous studies, estimation of critical gas velocity has been performed by either the liquid-droplet-removal model or a film-reversal mechanism. Multiple experimental studies conducted in the past few years have shown that liquid-film behavior is the ­primary mechanism behind the onset of liquid loading for inclined pipe, instead of the liquid-droplet-removal mechanism. However, liquid-droplet models are still used widely because of their simplicity. Both liquid-droplet-removal and film-reversal models are detailed in the complete paper.

Experimental Study

The system consists of industry-grade Coriolis flowmeters, thermal mass flowmeters, an electrical submersible pump, a progressive cavity pump, an air blower, and a three-phase separator to handle the fluids. The pipe section is connected to a truss system attached to a hoist that enables the straight pipe to incline at a range of 0–90° from horizontal. Because the current study focuses on the onset of liquid loading for two-phase flow, only the water and gas systems are used.

The test section is a 32-ft-long, 6-in.-inner-diameter (ID) acrylic pipe. Gas- and liquid-injection lines are connected to the inlet of the test section. The outlet of the test section is connected to a vertical 6-in.-ID return line, which leads the fluid flow back to the separator. The test section is instrumented with eight capacitance probes, three pressure transmitters, three temperature transmitters, and one differential pressure transducer. In addition, two injection and two drainage points are available to inject or remove desired liquid volumes. The test section can change the inclination angle from 0° (horizontal) to 90° (vertical).

Experimental Results and Analysis

Effect of Inclination Angle on Critical Gas Velocity. Pipe inclination angle is an important variable to be ­considered while determining the onset of liquid loading. Previous studies show that the critical gas velocity first increases, then decreases with increasing inclination angle, with the maximum occurring between approximately 40° and 60°. The trend is similar with the 3-in. pipe, which increases first and then decreases.

This trend could be explained by understanding film behavior as well as gravitational forces acting in the system. Liquid-film reversal starts at the bottom of the pipe, where the maximum liquid-film thickness is located. The liquid film being formed at the bottom of the pipe, or the maximum liquid-film thickness, changes as inclination angle increases. At the same time, the increasing gravitational gradient exceeds the influence of thicker film on liquid loading. In such cases, higher gas velocities are required to provide enough force and prevent the falling of the liquid film. This is the reason for the increase in critical gas velocity as the inclination angle increases from 0° (horizontal position) to approximately 40°. When the inclination angle is higher than 50°, the critical gas velocity begins to drop, causing liquid loading to occur at reduced velocity. This is because of the maximum liquid-film thickness, which decreases as the inclination angle increases, causing a reduced critical gas velocity.

Another important observation involves liquid velocities, where experiments run at higher liquid velocity achieve critical gas velocity at higher rates for the same inclination angle. The liquid-flow-rate effects on the critical gas velocity are diminished when the inclination angle approaches to 0° or 90° but become significant at medium inclination angles.

Effect of Pipe Diameter on Critical Gas Velocity. Pipe or tubing ID plays an important role in the estimation of critical gas velocity, which has been observed to increase with an increase of tubing ID. This is primarily because of smaller tubing having a smaller film thickness and thus smaller critical gas velocity. Experimental results confirm that the onset of liquid loading occurs at higher gas velocities for larger-diameter pipe.

Effect of Pressure on Critical Gas Velocity. The pressure inside the system also dictates the onset of liquid loading. Critical gas velocity is found to be lower for higher-pressure systems. For two different pressures using the same fluid, the critical gas velocity decreases as gas density increases. This is mainly because of the higher interfacial shear stress acting at the liquid interface for a higher gas density. In other words, for a higher gas density, the gas can drag the liquid phase more efficiently. The higher interfacial shear stress leads to smaller liquid-film thickness, resulting in lower critical gas velocity.

Model Development

In this paper, a new model is developed based on the liquid-film reversal at the bottom of the pipe. The critical velocity is the gas velocity, which corresponds to a zero liquid wall shear stress at the pipe bottom.

Previous experimental results indicate that the decrease of the critical gas velocity for high inclination angles (greater than 50°) is caused mainly by the reduction of the maximum liquid-film thickness at the pipe bottom. To account for the inclination-angle effects on the critical gas velocity, a model or correlation must be developed to capture the inclination-angle effects on the maximum liquid-film thickness.

The new model is based on the assumption that the liquid film is composed of two parts, circular liquid film around the pipe wall and accumulated liquid film at pipe bottom with a flat interface. Fig. 1 schematically shows the assumption of the current correlation for the inclination-angle effects on the liquid-film distribution. Previous and current studies have shown that flat interface geometry is dominated for small inclined pipe flow. For vertical annular flow, the liquid phase is almost uniformly distributed around the pipe wall. For the same liquid holdup, there is a gradual change of the maximum liquid-film thickness from flat interface geometry to uniform distribution (i.e., from the leftmost plot to the rightmost plot in Fig. 1). Equations and solutions contributing to the model are provided in the complete paper.

Fig. 1—Inclination-angle effects on liquid-film distribution. δ=liquid-film thickness.


The new model investigates the phenomenon that the maximum film thickness is close to the flat interface assumption for small inclination angles and the uniform liquid-film distribution when the pipe is 90° vertical. The results seem more realistic than those of previous studies, especially for low-inclination angles.

Model Evaluation

A comprehensive model evaluation was carried out using multiple data sets listed in Table 1 of the complete paper, along with the conditions under which those measurements were taken. Much of the current study is devoted to comparison of the new model prediction and previous experimental measurements for high- and low-pressure data sets, respectively. The authors write that the new model captures the pressure effect on critical gas velocity quite well in addition to inclination-angle effects. It also captures the pipe-diameter effect on critical gas velocity.

The authors write that previous liquid-droplet models, with one exception, cannot capture the inclination-angle and pipe-diameter effects on critical gas velocity. In general, the liquid-film model provides a better prediction compared with the liquid-droplet model, which is consistent with the findings from previous studies. Overall, the new model gives the best prediction compared with all other available models.


Based on the experimental observations from the current and previous studies, a new unified mechanics model is developed that considers deviation-angle, liquid-flow-rate, gas- and liquid-density, viscosity, and pipe-diameter effects on the onset of liquid accumulation. Additionally, a new model was developed to predict the maximum liquid-film thickness at the pipe bottom as a function of the inclination angle. In general, the new model gives the best prediction for the critical gas velocity.

The new mechanics model fills the knowledge gap to enhance accuracy when predicting the onset of liquid accumulation, especially for deviated and large-diameter wells. The model described in this paper is also applicable to gas-condensate pipelines where lower-inclination angles are more commonly encountered.

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 196130, “Experimental Investigation and Modeling of Onset of Liquid Accumulation in Large-Diameter Deviated Gas Wells,” by Ayush Rastogi and Yilin Fan, SPE, Colorado School of Mines, prepared for the 2019 SPE Annual Technical Conference and Exhibition, Calgary, 30 September–2 October. The paper has not been peer reviewed.