The purpose of the present study is to improve the current prediction capabilities of the entrain... more The purpose of the present study is to improve the current prediction capabilities of the entrainment fraction in horizontal gas-liquid flow. Since it is recognized that waves at the gas-liquid interface are the main source of entrainment, an experimental and theoretical work has been carried out to characterize the waves at the gas-liquid interface and to develop a model for entrainment calculations based on such characteristics. The model consists of three sub-models, namely, onset of entrainment, maximum entrainment and entrainment values in between. The onset of entrainment model determines the conditions at which the gas starts shearing the wave crests through a force balance between drag and surface tension forces. The maximum entrainment model provides the maximum fraction of liquid that can be entrained at high gas velocities by integration of the turbulent velocity profile to a determined dimensionless film thickness within the buffer sub layer. The entrainment fraction in between onset and maximum boundaries is calculated from an equilibrium between atomization and deposition rates. The atomization rate is calculated by first determining the wave mass flux in the liquid film and second by calculating the fraction of a single wave that is sheared by the gas through a force balance. The deposition rate is calculated as a linear function of the droplet concentration in the gas. Closure relationships have been developed from data for wave celerity, frequency, amplitude and width which are used in the entrainment model. A review of the most used correlations for calculating the entrainment fraction is presented and their performance evaluated. The present model shows better prediction than available models when compared to the acquired experimental data and the available experimental data in the literature.
Sand transport in multiphase flow has recently gained keen attention of the oil and gas industry ... more Sand transport in multiphase flow has recently gained keen attention of the oil and gas industry owing to the negative effects associated with it. These include partial pipe blockage, pipe corrosion, excessive pressure drop and production decline. To date, no comprehensive literature review and models evaluation have been published, which compare the experimental data collected for the prediction of the critical sand deposition velocity under intermittent flow with the related model predictions. This study can be used by engineers and researchers to determine the conditions under which the developed models perform the best. The intermittent flow critical sand deposition velocity data acquired by Najmi (2015) are presented in detail. Next, the effects of important parameters such as phase velocities, liquid viscosity as well as particle size and concentration on the critical velocity are investigated. The collected data are utilized to evaluate the performance of the models developed by Salama (1998), Hill (2011), Stevenson et al. (2001) and Danielson (2007), in order to determine the best model for the prediction of the sand critical velocity. The experimental data of Najmi (2015) indicate that higher critical velocities are required with increasing the liquid viscosity, particle size and particle concentration. However, the predictions of the models of Salama (1998), Stevenson et al. (2001) and Danielson (2007) demonstrate that these models do not take into account the effect of particle concentration. Depending on the liquid viscosity, Stevenson et al. (2001) model significantly over-predicts or under-predicts the critical velocity over different ranges of the phase velocities, while Salama (1998) model under-predicts the critical velocity under all experimental conditions. An overall comparison of the data with the published model predictions confirms that the Hill (2011) model has the best performance capturing the physical phenomena, including the effects of phase velocities, particle size, particle concentration and liquid viscosity.
Slug flow, which commonly occurs in the petroleum industry, is not always a desired flow pattern ... more Slug flow, which commonly occurs in the petroleum industry, is not always a desired flow pattern due to production operation problems it may cause in pipelines and processing facilities. To mitigate these problems, flow conditioning devices such as multiphase flow manifolds and slug catchers are used, where dissipation of slugs occurs in downward flow or in larger diameter pipe sections. Tee-junctions are important parts of these flow conditioning devices. In this work, Computational Fluid Dynamics (CFD) simulations are conducted using ANSYS/FLUENT 17.2 to investigate slug dissipation in an Enlarged Impacting Tee-Junction (EIT). An Eulerian–Eulerian MultiFluid VOF transient model in conjunction with the standard k-ε turbulent model is used to simulate slug dissipation in an EIT geometry. The EIT consists of a 0.05 m ID 10 m long inlet, which is connected to the center of a 0.074 m ID 5.5 m long section that forms the EIT branches. Moreover, experimental data are acquired on slug dissipation lengths in a horizontal EIT with a similar geometry as in the CFD simulations. The CFD results include the mean void fraction and cross-sectionally averaged void fraction time series in the EIT for different gas and liquid velocities. These results provide the inlet slug length and dissipation length in the EIT branches. The CFD results are evaluated against the experimental data demonstrating that the slug dissipation occurring in EIT branches can be predicted by simulation.
Journal of Energy Resources Technology-transactions of The Asme, Feb 27, 2018
In order to assess the critical sand deposition condition, a unique 4-in ID test facility was des... more In order to assess the critical sand deposition condition, a unique 4-in ID test facility was designed and constructed, which enables the pipe to be inclined 1.5 deg upward. Experiments were conducted with air–water-glass beads at low sand concentrations (< 10,000 ppm), and the air and water flow rates were selected to ensure stratified flow regime along the pipe. At constant superficial liquid velocity, the gas velocity was reduced to find the critical sand deposition velocity. Six sand flow regimes are identified, namely, fully dispersed solid flow, dilute solids at the wall, concentrated solids at the wall, moving dunes, stationary dunes, and stationary bed. The experimental results reveal that sand flow regimes under air–water stratified flow are strong functions of phase velocities, particle size, and particle concentration. Also, the results show that air–water flow regime plays an important role in particle transport; slug flow has high capability to transport particles at the pipe bottom, while the stratified flow has high risk of sand deposition. As long as the sand dunes are observed at the pipe bottom, the critical sand deposition velocities slightly increase with concentrations, while for stationary bed, the critical velocity increases exponentially with concentration.
This report was prepared as an account of work sponsored by an agency of the United States Govern... more This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government oor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liabiiity or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or proctss disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recornmendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government orany agency thereof.
This paper presents the comparative study of experimental, modeling, and simulation results that ... more This paper presents the comparative study of experimental, modeling, and simulation results that are performed using commercially available ANSYS Fluent software to analyze the separation kinetics of oil and water in a horizontal separator at various velocities and watercuts. The horizontal pipe separator used in this study has an internal diameter of 0.0762 m and a length of 10.3 m separating oil and water with specific gravities of 1.0 and 0.857 and watercuts ranging from 20 to 90%. The mixture velocities studied are 0.08, 0.13, and 0.20 m/s. Numerical simulations are done using the hybrid Eulerian-Eulerian multifluid VOF model to study the effect of watercut on the creaming of the oil layer and sedimentation of the water layer respectively. As the mixture velocities increased, the initial length of separation increased like experimental results. As the watercut increased, the separation of water enhanced, while the oil creaming improved with the lowering of the watercut as expected. Numerical results showed good agreement for water/dispersion interface predictions for all the conditions studied. The CFD results are compared against experimental results obtained by Othman in 2010 and agree with the trend of separation. The numerical simulations gave insights into the velocity profiles in each of the layers such as creamed oil, sedimented water, and the layer of emulsion that is not separated. Also, the numerical results are validated against the extended Gassies (2008) model incorporating correlation for turbulent time decay and oil volume fraction proposed by Dabirian et al in 2018.
Gas Carry-Under (GCU) is one of the undesirable phenomena that exist in the Gas-Liquid Cylindrica... more Gas Carry-Under (GCU) is one of the undesirable phenomena that exist in the Gas-Liquid Cylindrical Cyclone (GLCC) separators even within the liquid carry-over Operational Envelope (OE). In order to quantify the GCU, it is important to understand the cause of gas entrainment that occurs in the GLCC other than the incoming entrained gas within the liquid medium. The tangential inclined inlet of 27° with reduced area allows the stratified liquid flow to exit the inlet nozzle tangentially along the wall into the vertical lower part of the GLCC, whereby the liquid film spreads along the wall in an asymmetrical shape. The gas moves to the center of the GLCC and escapes through the gas leg. The liquid film flow is complex and turbulent exhibiting unevenness of the film thickness and asymmetrical velocity distribution. Experimental investigations show that the magnitude of liquid wall jet film tangential and axial velocity change as a function of length along the GLCC below the inlet of the GLCC. This wall jet film flowing down along the wall is the cause for gas entrainment and GCU. The experimental results show that the gas entrainment mechanism is not like the conventional jet entrainment as expected to be occurring in GLCC. The change in velocities of the wall jet film at various liquid heights maintained below the inlet results in varying gas entrainment at various inlet liquid levels and for fluid properties. The wall jet phenomena that takes places at the inlet has been discussed in detail and a mechanistic model capable of predicting the wall jet parameters has been presented in this paper. Further, a novel mechanistic model that is developed for the first time is also presented which can predict the gas entrainment at various liquid levels and flow conditions using the wall jet parameters as an input condition.
Production equipment, such as pumps and chokes, cause shear in oil/water-mixture flow and form sm... more Production equipment, such as pumps and chokes, cause shear in oil/water-mixture flow and form smaller droplets in the system, which might lead to operational and separation problems. Our objective in this paper is to provide new data on droplet sizes under different shear conditions in oil/water flow and compare these data with model predictions. New experimental data on droplet-size distribution are acquired for a gear pump (configured in a closed loop) with an inline camera, using the bead-sizing tool. The acquired data capture the effects of the dispersed-phase volumetric fraction and shear intensity. Two prediction models for a centrifugal pump, the Pereyra (2011) and Kouba (2014) models, are modified to enable droplet-size-distribution predictions for a gear pump. These models use the log-normal-distribution and Rosin-Rammler-distribution methods, which have been found to match the acquired data relatively well. The results demonstrate that mean droplets tend to decrease in diameter when the shear intensity increases. Moreover, the higher the dispersed-phase volumetric fraction is, the larger the resulting droplets will be. Comparisons between the acquired data and the proposed droplet-size-distribution prediction, which uses the modified d max model, are very good for both Rosin-Rammler and log-normal distributions, with normalized root-mean-square deviation (NRMSD) of less than 7%.
A novel non-fibrous filter media is evaluated for in-line oil-water separation. Outside-in-crossf... more A novel non-fibrous filter media is evaluated for in-line oil-water separation. Outside-in-crossflow configuration incorporating the filter media is utilized in order to test the filter. All experiments are conducted with a hydrophilic-olephobic filter for water-continuous flow with low oil concentrations. The collected experimental data include permeate flow rate and purity as well as pressure drop. Values of permeate flow rate and pressure drop are averaged over the duration of the experiments, which is about 5 minutes, constituting the “initial average” of the permeate flow rate and the corresponding pressure drop. Totally twelve experimental runs are conducted for mixture velocities of 0.038 m/s, 0.055 m/s and 0.066 m/s, and oil concentrations of 0.6%, 0.83%, 1.1%, 7.9% and 9.1%. Permeate samples are analyzed for oil content, demonstrating a high separation efficiency of 98 ± 2%. The permeate flux across the filter cartridge ranges between 0.0739 (L/h)/cm2 to 0.216 (L/h)/cm2 owing to the low pressure drop across to filter. Oil concentration in to permeate water samples shows consistently increasing trend with an increase in inlet oil content, while maintaining high separation efficiency for all runs. The pressure drop across the membrane under flowing conditions ranges from 0.35 psid to 0.6 psid for flow rates between 0.1 L/min and 0.29 L/min, respectively. Also the data confirm that the filter membrane breakthrough pressure is 0.35 psid.
A novel and unique experimental facility was design, constructed and instrumented, enabling slug ... more A novel and unique experimental facility was design, constructed and instrumented, enabling slug flow splitting data acquisition in both parallel and looped lines. A total of 81 experimental test runs were carried out for various superficial gas and liquid velocity combinations. Uneven split conditions data were acquired by utilizing either a choke valve on one of the lines or different pipe diameters. For the symmetrical configurations of both parallel and looped equal diameter lines, the phases split equally in the lines. When the parallel and looped lines are of different resistance to flow (utilizing a choke installed on one of the lines or different line diameters), uneven split of the phases occurs. The gas-phase flows preferentially into the smaller resistance line. This results in different gas–liquid ratios in the two lines, which are different from the gas–liquid ratio at the inlet. A mechanistic model has been developed for the prediction of the uneven gas and liquid splitting and the pressure drop in two-phase looped lines. Good agreement is observed between the model predictions and the experimental data, with an average error of about 15 % in the phase splitting and pressure drop.
The purpose of the present study is to improve the current prediction capabilities of the entrain... more The purpose of the present study is to improve the current prediction capabilities of the entrainment fraction in horizontal gas-liquid flow. Since it is recognized that waves at the gas-liquid interface are the main source of entrainment, an experimental and theoretical work has been carried out to characterize the waves at the gas-liquid interface and to develop a model for entrainment calculations based on such characteristics. The model consists of three sub-models, namely, onset of entrainment, maximum entrainment and entrainment values in between. The onset of entrainment model determines the conditions at which the gas starts shearing the wave crests through a force balance between drag and surface tension forces. The maximum entrainment model provides the maximum fraction of liquid that can be entrained at high gas velocities by integration of the turbulent velocity profile to a determined dimensionless film thickness within the buffer sub layer. The entrainment fraction in between onset and maximum boundaries is calculated from an equilibrium between atomization and deposition rates. The atomization rate is calculated by first determining the wave mass flux in the liquid film and second by calculating the fraction of a single wave that is sheared by the gas through a force balance. The deposition rate is calculated as a linear function of the droplet concentration in the gas. Closure relationships have been developed from data for wave celerity, frequency, amplitude and width which are used in the entrainment model. A review of the most used correlations for calculating the entrainment fraction is presented and their performance evaluated. The present model shows better prediction than available models when compared to the acquired experimental data and the available experimental data in the literature.
Sand transport in multiphase flow has recently gained keen attention of the oil and gas industry ... more Sand transport in multiphase flow has recently gained keen attention of the oil and gas industry owing to the negative effects associated with it. These include partial pipe blockage, pipe corrosion, excessive pressure drop and production decline. To date, no comprehensive literature review and models evaluation have been published, which compare the experimental data collected for the prediction of the critical sand deposition velocity under intermittent flow with the related model predictions. This study can be used by engineers and researchers to determine the conditions under which the developed models perform the best. The intermittent flow critical sand deposition velocity data acquired by Najmi (2015) are presented in detail. Next, the effects of important parameters such as phase velocities, liquid viscosity as well as particle size and concentration on the critical velocity are investigated. The collected data are utilized to evaluate the performance of the models developed by Salama (1998), Hill (2011), Stevenson et al. (2001) and Danielson (2007), in order to determine the best model for the prediction of the sand critical velocity. The experimental data of Najmi (2015) indicate that higher critical velocities are required with increasing the liquid viscosity, particle size and particle concentration. However, the predictions of the models of Salama (1998), Stevenson et al. (2001) and Danielson (2007) demonstrate that these models do not take into account the effect of particle concentration. Depending on the liquid viscosity, Stevenson et al. (2001) model significantly over-predicts or under-predicts the critical velocity over different ranges of the phase velocities, while Salama (1998) model under-predicts the critical velocity under all experimental conditions. An overall comparison of the data with the published model predictions confirms that the Hill (2011) model has the best performance capturing the physical phenomena, including the effects of phase velocities, particle size, particle concentration and liquid viscosity.
Slug flow, which commonly occurs in the petroleum industry, is not always a desired flow pattern ... more Slug flow, which commonly occurs in the petroleum industry, is not always a desired flow pattern due to production operation problems it may cause in pipelines and processing facilities. To mitigate these problems, flow conditioning devices such as multiphase flow manifolds and slug catchers are used, where dissipation of slugs occurs in downward flow or in larger diameter pipe sections. Tee-junctions are important parts of these flow conditioning devices. In this work, Computational Fluid Dynamics (CFD) simulations are conducted using ANSYS/FLUENT 17.2 to investigate slug dissipation in an Enlarged Impacting Tee-Junction (EIT). An Eulerian–Eulerian MultiFluid VOF transient model in conjunction with the standard k-ε turbulent model is used to simulate slug dissipation in an EIT geometry. The EIT consists of a 0.05 m ID 10 m long inlet, which is connected to the center of a 0.074 m ID 5.5 m long section that forms the EIT branches. Moreover, experimental data are acquired on slug dissipation lengths in a horizontal EIT with a similar geometry as in the CFD simulations. The CFD results include the mean void fraction and cross-sectionally averaged void fraction time series in the EIT for different gas and liquid velocities. These results provide the inlet slug length and dissipation length in the EIT branches. The CFD results are evaluated against the experimental data demonstrating that the slug dissipation occurring in EIT branches can be predicted by simulation.
Journal of Energy Resources Technology-transactions of The Asme, Feb 27, 2018
In order to assess the critical sand deposition condition, a unique 4-in ID test facility was des... more In order to assess the critical sand deposition condition, a unique 4-in ID test facility was designed and constructed, which enables the pipe to be inclined 1.5 deg upward. Experiments were conducted with air–water-glass beads at low sand concentrations (< 10,000 ppm), and the air and water flow rates were selected to ensure stratified flow regime along the pipe. At constant superficial liquid velocity, the gas velocity was reduced to find the critical sand deposition velocity. Six sand flow regimes are identified, namely, fully dispersed solid flow, dilute solids at the wall, concentrated solids at the wall, moving dunes, stationary dunes, and stationary bed. The experimental results reveal that sand flow regimes under air–water stratified flow are strong functions of phase velocities, particle size, and particle concentration. Also, the results show that air–water flow regime plays an important role in particle transport; slug flow has high capability to transport particles at the pipe bottom, while the stratified flow has high risk of sand deposition. As long as the sand dunes are observed at the pipe bottom, the critical sand deposition velocities slightly increase with concentrations, while for stationary bed, the critical velocity increases exponentially with concentration.
This report was prepared as an account of work sponsored by an agency of the United States Govern... more This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government oor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liabiiity or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or proctss disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recornmendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government orany agency thereof.
This paper presents the comparative study of experimental, modeling, and simulation results that ... more This paper presents the comparative study of experimental, modeling, and simulation results that are performed using commercially available ANSYS Fluent software to analyze the separation kinetics of oil and water in a horizontal separator at various velocities and watercuts. The horizontal pipe separator used in this study has an internal diameter of 0.0762 m and a length of 10.3 m separating oil and water with specific gravities of 1.0 and 0.857 and watercuts ranging from 20 to 90%. The mixture velocities studied are 0.08, 0.13, and 0.20 m/s. Numerical simulations are done using the hybrid Eulerian-Eulerian multifluid VOF model to study the effect of watercut on the creaming of the oil layer and sedimentation of the water layer respectively. As the mixture velocities increased, the initial length of separation increased like experimental results. As the watercut increased, the separation of water enhanced, while the oil creaming improved with the lowering of the watercut as expected. Numerical results showed good agreement for water/dispersion interface predictions for all the conditions studied. The CFD results are compared against experimental results obtained by Othman in 2010 and agree with the trend of separation. The numerical simulations gave insights into the velocity profiles in each of the layers such as creamed oil, sedimented water, and the layer of emulsion that is not separated. Also, the numerical results are validated against the extended Gassies (2008) model incorporating correlation for turbulent time decay and oil volume fraction proposed by Dabirian et al in 2018.
Gas Carry-Under (GCU) is one of the undesirable phenomena that exist in the Gas-Liquid Cylindrica... more Gas Carry-Under (GCU) is one of the undesirable phenomena that exist in the Gas-Liquid Cylindrical Cyclone (GLCC) separators even within the liquid carry-over Operational Envelope (OE). In order to quantify the GCU, it is important to understand the cause of gas entrainment that occurs in the GLCC other than the incoming entrained gas within the liquid medium. The tangential inclined inlet of 27° with reduced area allows the stratified liquid flow to exit the inlet nozzle tangentially along the wall into the vertical lower part of the GLCC, whereby the liquid film spreads along the wall in an asymmetrical shape. The gas moves to the center of the GLCC and escapes through the gas leg. The liquid film flow is complex and turbulent exhibiting unevenness of the film thickness and asymmetrical velocity distribution. Experimental investigations show that the magnitude of liquid wall jet film tangential and axial velocity change as a function of length along the GLCC below the inlet of the GLCC. This wall jet film flowing down along the wall is the cause for gas entrainment and GCU. The experimental results show that the gas entrainment mechanism is not like the conventional jet entrainment as expected to be occurring in GLCC. The change in velocities of the wall jet film at various liquid heights maintained below the inlet results in varying gas entrainment at various inlet liquid levels and for fluid properties. The wall jet phenomena that takes places at the inlet has been discussed in detail and a mechanistic model capable of predicting the wall jet parameters has been presented in this paper. Further, a novel mechanistic model that is developed for the first time is also presented which can predict the gas entrainment at various liquid levels and flow conditions using the wall jet parameters as an input condition.
Production equipment, such as pumps and chokes, cause shear in oil/water-mixture flow and form sm... more Production equipment, such as pumps and chokes, cause shear in oil/water-mixture flow and form smaller droplets in the system, which might lead to operational and separation problems. Our objective in this paper is to provide new data on droplet sizes under different shear conditions in oil/water flow and compare these data with model predictions. New experimental data on droplet-size distribution are acquired for a gear pump (configured in a closed loop) with an inline camera, using the bead-sizing tool. The acquired data capture the effects of the dispersed-phase volumetric fraction and shear intensity. Two prediction models for a centrifugal pump, the Pereyra (2011) and Kouba (2014) models, are modified to enable droplet-size-distribution predictions for a gear pump. These models use the log-normal-distribution and Rosin-Rammler-distribution methods, which have been found to match the acquired data relatively well. The results demonstrate that mean droplets tend to decrease in diameter when the shear intensity increases. Moreover, the higher the dispersed-phase volumetric fraction is, the larger the resulting droplets will be. Comparisons between the acquired data and the proposed droplet-size-distribution prediction, which uses the modified d max model, are very good for both Rosin-Rammler and log-normal distributions, with normalized root-mean-square deviation (NRMSD) of less than 7%.
A novel non-fibrous filter media is evaluated for in-line oil-water separation. Outside-in-crossf... more A novel non-fibrous filter media is evaluated for in-line oil-water separation. Outside-in-crossflow configuration incorporating the filter media is utilized in order to test the filter. All experiments are conducted with a hydrophilic-olephobic filter for water-continuous flow with low oil concentrations. The collected experimental data include permeate flow rate and purity as well as pressure drop. Values of permeate flow rate and pressure drop are averaged over the duration of the experiments, which is about 5 minutes, constituting the “initial average” of the permeate flow rate and the corresponding pressure drop. Totally twelve experimental runs are conducted for mixture velocities of 0.038 m/s, 0.055 m/s and 0.066 m/s, and oil concentrations of 0.6%, 0.83%, 1.1%, 7.9% and 9.1%. Permeate samples are analyzed for oil content, demonstrating a high separation efficiency of 98 ± 2%. The permeate flux across the filter cartridge ranges between 0.0739 (L/h)/cm2 to 0.216 (L/h)/cm2 owing to the low pressure drop across to filter. Oil concentration in to permeate water samples shows consistently increasing trend with an increase in inlet oil content, while maintaining high separation efficiency for all runs. The pressure drop across the membrane under flowing conditions ranges from 0.35 psid to 0.6 psid for flow rates between 0.1 L/min and 0.29 L/min, respectively. Also the data confirm that the filter membrane breakthrough pressure is 0.35 psid.
A novel and unique experimental facility was design, constructed and instrumented, enabling slug ... more A novel and unique experimental facility was design, constructed and instrumented, enabling slug flow splitting data acquisition in both parallel and looped lines. A total of 81 experimental test runs were carried out for various superficial gas and liquid velocity combinations. Uneven split conditions data were acquired by utilizing either a choke valve on one of the lines or different pipe diameters. For the symmetrical configurations of both parallel and looped equal diameter lines, the phases split equally in the lines. When the parallel and looped lines are of different resistance to flow (utilizing a choke installed on one of the lines or different line diameters), uneven split of the phases occurs. The gas-phase flows preferentially into the smaller resistance line. This results in different gas–liquid ratios in the two lines, which are different from the gas–liquid ratio at the inlet. A mechanistic model has been developed for the prediction of the uneven gas and liquid splitting and the pressure drop in two-phase looped lines. Good agreement is observed between the model predictions and the experimental data, with an average error of about 15 % in the phase splitting and pressure drop.
Uploads
Papers by Ram Mohan