Measurement of gas phase characteristics in vertical oil-gas-water slug and churn flows
Introduction
Oil-gas-water three-phase flow widely exists in the oil well production and oil-gas transportation. Due to the significant differences of physical properties and mutual interactions among three phases, the flow structures are more complicated compared with gas-liquid or oil-water two phase flow, which brings great difficulties to measure flow parameters in oil-gas-water three-phase flow. Understanding the gas characteristics of oil-gas-water three-phase flow is of significant importance to flow measurement as well as to develop oil-gas-water three-phase flow models.
In early studies of oil-gas-water three-phase flow, Tek (1961) regarded the two immiscible liquids as an equivalent single phase and predicted the pressure loss. Shean (1976) developed flow regime maps and established drift-flux models for vertical oil-gas-water flows to obtain the global volumetric phase fractions. Lahey et al. (1992) proposed a generalized drift-flux expression for horizontal three-phase conduit flow. Thorn et al. (1997) summarized the principal strategies and technologies of three-phase flow measurement. Fordham et al. (1999) utilized local fiber-optical sensors for immiscible-fluid discrimination and determining volume fraction profiles in oil-gas-water three-phase. Ghorai et al. (2005) developed a model to predict the values of holdup and pressure gradient for three-phase stratified flow prevailing in a horizontal pipeline. Spedding et al. (2007) investigated oil-gas-water three-phase flow in a horizontal 0.0259 m ID pipe, they found that liquid holdup data depended on flow regimes, and presented a new model for the prediction of three-phase liquid holdup. Silva et al. (2007) firstly implemented the local complex permittivity measurements of a three-phase bubbly flow. Salgado et al. (2009) proposed new methodology based on nuclear technique and artificial neural network for volume fraction predictions in oil-water-gas multiphase systems. Cazarez et al. (2010) propounded a two-fluid mathematical model to predict pressure, temperature, volumetric fraction and velocity profiles in heavy oil-gas-water bubbly flow. Hoffmann and Johnson (2011) used a traversable dual-energy gamma instrument to measure phase fractions at different positions. Henry et al. (2013) applied Coriolis metering for a three-phase (oil-gas-water) mixture. Their experimental results demonstrated the potential of using Coriolis mass flow metering in three-phase flow measurement. Sun and Yang (2015) imaged the three-phase flows based on ECT/ERT dual-modality and calculated the holdup of each phase in a three-phase flow with the image fusion results. Karami et al. (2017) utilized an isokinetic sampling probe system to measure liquid phase entrainment fraction. Pietrzak et al. (2017) developed new methods to calculate phase fraction and total pressure drop in oil-gas-water three-phase flow.
As for experimental observations and flow pattern classification, Chen (1991) investigated three-phase flow characteristics in an upward vertical pipe and classified flow patterns into oil in water or water in oil type flow. Açikgöz et al. (1992) performed extensive flow visualization studies in a 19 mm diameter pipe and defined horizontal oil-gas-water three-phase flow regimes. Woods et al. (1998) conducted three-phase flow experiment in a 26 mm ID vertical perspex pipe and concluded nine flow patterns. Spedding et al. (2000) pointed out that flow regime maps, liquid holdup and pressure drop were all different between the vertical and near vertical three-phase upflow. Oddie et al. (2003) investigated steady-state and transient multiphase flow in a large diameter, inclined pipes and detailed flow pattern maps were generated over the entire range of flow rates and pipe inclinations. Descamps et al., 2006, Descamps et al., 2007 studied the influence of gas injection on phase inversion in oil-water flow through a vertical tube. In addition, other impact factors, such as, the processes of mixing of the liquid phases (Hewitt, 2005), the properties of pipe geometry, liquid viscosity and surface tension (Wegmann et al., 2007), played an important role in the flow pattern transition. Bannwart et al. (2009) observed three-phase flow patterns in horizontal and vertical 2.84 cm ID glass pipes and assessed the changes of pressure drop induced by water injection. S. Wang et al. (2013) investigated the effects of water injection and viscosity on flow parameter measurement. More recently, intelligent recognition of flow regime based on the signals of gamma-ray by using artificial neural networks has been a research focus (Salgado et al., 2010, Roshani et al., 2015, Roshani et al., 2016, Roshani et al., 2017, Nazemi et al., 2016).
As multiphase flow is a typical nonlinear system, some progress have been achieved in applying nonlinear analysis method to analyze the dynamic behavior of multiphase flow. Fan et al. (1990) analyzed pressure fluctuations in a gas-liquid-solid fluidized bed under different batch operating conditions based on the concept of fractals. Kikuchi et al. (1996) reported the chaotic motions of bubbles and particles in gas-liquid-solid fluidized bed. Yano et al. (1999) investigated the scale-up effect on the dynamic behavior of gas-liquid-solid three-phase reactors using deterministic chaos analysis. Wu et al. (2000) obtained characteristics vectors of various flow regimes in terms of fractal theory. Fraguío et al. (2007) classified flow regimes in three-phase fluidized beds based on chaos theory. In our research group, fractals and chaos (Jin et al., 2001), chaotic attractor morphological description and complexity measures (Wang et al., 2010), complex network (Gao and Jin, 2011), multi-scale long-range magnitude and sign correlations (Zhao et al., 2015) and multi-scale weighted complexity entropy causality plane (Zhuang et al., 2016) have been utilized to investigate the nonlinear dynamic characteristics of three-phase flow.
The significant differences of physical properties and interface interactions among three phases result in more complicated flow structures in oil-gas-water three-phase flow compared with gas-liquid or oil-water two phase flow. So far, studies mainly focus on flow parameters measurement or flow pattern recognition. Knowledge of flow structure and its local flow characteristics in oil-gas-water three phase flow is still relatively lack especially for slug and churn flows. In this study, gas phase characteristics of oil-gas-water slug and churn flows in a vertical upward pipe with 20 mm inner diameter (ID) are investigated and local gas velocity, local gas holdup, flow structure and bubble size at different radial positions are obtained using traversable bi-optical fiber probe. To understand the nonlinear dynamic characteristics of slug and churn flows, the signals of a high-resolution half-ring conductance sensor and optical fiber probes are analyzed using multi-scale cross entropy (MSCE) algorithm to characterize the nonlinear dynamics in slug and churn flows.
Section snippets
Traversable bi-optical fiber probe sensing system
The measurement principle of optical fiber probe is based on the refraction and reflection law. Phases can be discriminated by detecting the intensity of reflected light due to the different refractive indexes of oil, gas and water. The cone angle of the probe tip is very important for gas phase discrimination, and it could be determined by theoretical derivation and experimental test. According to the Fresnel equations (Saleh and Teich, 1991)
Optical fiber probe signal processing
The bi-optical fiber probe fluctuating signals of slug and churn flows at seven measurement positions are shown in Fig. 5. “Position nU, Position nW” respectively represent the leading and rear probe signals at position n. The voltage of output signal is high when the probe tip is in gas phase whilst it is low in liquid phase.
Fig. 5(a) shows the raw signals of slug flow. It can be seen that at measurement position 1 and 7, gas phase near the pipe wall is mainly in the form of gas bubble. Taylor
Nonlinear dynamics analysis for slug and churn flows
In attempt to reveal the nonlinear dynamic characteristics of slug and churn flow and validate the results of gas phase characteristics obtained from bi-optical fiber probe signals, multi-scale cross entropy (MSCE) algorithm is applied to analyze signals of high-resolution half-ring conductance sensor and bi-optical fiber probe to uncover the dynamical instability of slug and churn flows at different positions. The high-resolution half-ring conductance sensor has an advantage on flow pattern
Conclusions
In this study, a traversable bi-optical fiber probe is designed to measure local gas phase characteristics of oil-gas-water slug and churn flows in a 20 mm ID vertical pipe. We also utilize multi-scale cross entropy (MSCE) algorithm to analyze signals of high-resolution half-ring conductance sensor and bi-optical fiber probe to uncover the nonlinear dynamic characteristics of slug and churn flows at different positions. Our conclusions can be stated as follows:
- 1.
Taylor bubbles are surrounded by
Acknowledgments
This study was supported by National Natural Science Foundation of China (Grant Nos. 51527805, 11572220).
References (65)
- et al.
An experimental study of three-phase flow regimes
Int. J. Multiph. Flow
(1992) - et al.
Experimental investigation on liquid–liquid–gas flow: flow patterns and pressure-gradient
J. Petrol. Sci. Eng.
(2009) - et al.
Modeling of three-phase heavy oil–water–gas bubbly flow in upward vertical pipes
Int. J. Multiph. Flow
(2010) - et al.
Slug to churn transition in upward vertical two-phase flow
Chem. Eng. Sci.
(1997) - et al.
Slug flow regime identification from dynamic void fraction measurements in vertical air-water flows
Int. J. Multiph. Flow
(1997) - et al.
The experimental signals analysis for bubbly oil-in-water flow using multi-scale weighted-permutation entropy
Physica A
(2015) - et al.
Influence of gas injection on phase inversion in an oil–water flow through a vertical tube
Int. J. Multiph. Flow
(2006) - et al.
Experimental investigation of three-phase flow in a vertical pipe: local characteristics of the gas phase for gas-lift conditions
Int. J. Multiph. Flow
(2007) - et al.
Classifying flow regimes in three-phase fluidized beds from CARPT experiments
Chem. Eng. Sci.
(2007) - et al.
Numerical modeling of three-phase stratified flow in pipes
Chem. Eng. Sci.
(2005)