Analysis of roll wave characteristics under low liquid loading two‐phase flow conditions

An experimental study is conducted using a 0.152‐m ID facility to investigate the wave characteristics of two‐phase stratified wavy flow in horizontal pipelines. The experiments are conducted under low liquid loading condition, which is very commonly observed in wet gas pipelines. The experiments are conducted with water as the liquid phase, and repeated with 51 wt % of monoethylene glycol (MEG) in the aqueous phase to analyze the effects of MEG presence on wave characteristics. The experimental range of this study covers superficial gas velocity, v_Sg, values of 9–23 m/s and superficial liquid velocity, v_SL, values of 0.01–0.02 m/s. Similar test matrices are completed for the cases with and without MEG in the aqueous phase. A conductivity probe system is used to measure the wave characteristics at the liquid–gas interface. These characteristics include the wave celerity, frequency, amplitude, length, and liquid film thickness. The experimental oil–air wave characteristics data of Gawas et al. (Int J Multiphase Flow. 2014;63:93–104) is also used for comparison purposes. The trends in the resulting wave characteristics with respect to input parameters are investigated, for oil, water, or MEG–water mixture as the liquid phase. Common predictive methods for interfacial wave celerity, including shallow water theory, Watson (Proceedings of the 4th International Conference in Multi‐Phase Flows, Nice, France. 1989:495–512), Paras et al. (Int J Multiphase Flow. 1994;20(5):939–956), Al‐Sarkhi et al. (AIChE J. 2012;58(4):1018–1029), and Gawas et al. (Int J Multiphase Flow. 2014;63:93–104) are evaluated in comparison with the experimental data. The results of the wave frequency correlation of Al‐Sarkhi et al. (AIChE J. 2012;58(4):1018–1029) are also compared with the experimental wave frequency data. Lastly, a correlation is developed to predict the relative wave amplitude, as a function of superficial gas Weber number and liquid velocity number. Most of the commonly used two‐phase stratified flow models are developed with the assumption of steady‐state conditions, and neglect the transient wave effects. This study provides valuable experimental results on wave characteristics of stratified wavy flow for different types of liquid phase. Moreover, a comprehensive analysis of the parameters affecting the wave characteristics of stratified wavy flow is presented. © 2017 American Institute of Chemical Engineers AIChE J, 63: 3177–3186, 2017     

Influence of Gas Injection on the In‐Situ Oil Fraction of an Oil‐Water Flow in Horizontal Pipes

In this work, an experimental study was made on gas injection into an oil‐water flow in horizontal pipes with two unequal pipe diameters. Special attention was given to the influence of gas injection on the average in‐situ oil fraction. Measurements were made for input water flow rates of 1.25–5 m³/h, input oil flow rates of 0–8 m³/h and input gas flow rates of 0–9 m³/h. It was found that gas injection has a considerable influence on the in‐situ oil fraction. In general, a small increase in the rate of air injection leads to greatly decreasing in‐situ oil fractions. The in‐situ oil fraction with gas injection decreases to a greater extent than that without gas injection, at the same input liquid flow rates. At a given input water flow rate, the value of the in‐situ oil fraction in the pipe with the larger diameter is higher than that in the pipe with the smaller diameter. Furthermore, the drift flux models were extended to predict the average in‐situ fractions of the oil phase in the intermittent three‐phase flow regimes. A good agreement is obtained between theory and data, especially for the in‐situ oil fraction range of 0.2–1.0.     

Pressure Drop in Liquid‐liquid Two Phase Horizontal Flow: Experiment and Prediction

The present study is aimed at an investigation of the pressure drop characteristics during the simultaneous flow of a kerosene‐water mixture through a horizontal pipe of 0.025 m diameter. Measurements of pressure gradient were made for different combinations of phase superficial velocities ranging from 0.03–2 m/s such that the regimes encountered were smooth stratified, wavy stratified, three layer flow, plug flow and oil dispersed in water, and water flow patterns. A model was developed, which considered the energy minimization and pressure equalization of both phases.     

Liquid‐liquid flow patterns in reduced dimension based on energy minimization approach

Study of liquid‐liquid flow patterns in reduced dimensions is relevant under the current trends to miniaturize process equipment. The phase distribution results from interplay of surface (dominant in microchannels) and gravity forces (dominant in larger dimensions). The proposed analysis, based on minimization of total system energy comprising of kinetic, surface, and potential energy, unravels the influence of wetting properties and predicts the range of existence of annular and plug flow as well as the onset of stratification with increasing conduit dimension. Unlike existing models marking abrupt transitions, the proposed methodology can predict zones of transition where interfacial distributions gradually evolve with change of operating conditions—the predictions agreeing closely to experimental and literature data. The analysis illustrates the coupled effect of diameter, contact angle, and inlet composition on flow distribution and defines the transition from macrodomain to microdomain (millichannels) in terms of Bond number as 0.1 < Bo < 10. © 2015 American Institute of Chemical Engineers AIChE J, 62: 287–294, 2016     

Oil–water two‐phase flow in microchannels: Flow patterns and pressure drop measurements

This paper investigates oil–water two‐phase flows in microchannels of 793 and 667 µm hydraulic diameters made of quartz and glass, respectively. By injecting one fluid at a constant flow rate and the second at variable flow rate, different flow patterns were identified and mapped and the corresponding two‐phase pressure drops were measured. Measurements of the pressure drops were interpreted using the homogeneous and Lockhart–Martinelli models developed for two‐phase flows in pipes. The results show similarity to both liquid–liquid flow in pipes and to gas–liquid flow in microchannels. We find a strong dependence of pressure drop on flow rates, microchannel material, and the first fluid injected into the microchannel.     

Bubble/droplet formation and mass transfer during gas–liquid–liquid segmented flow with soluble gas in a microchannel

Microchannels have great potential in intensification of gas–liquid–liquid reactions involving reacting gases, such as hydrogenation. This work uses CO₂–octane–water system to model the hydrodynamics and mass transfer of such systems in a microchannel with double T‐junctions. Segmented flows are generated with three inlet sequences and the size laws of dispersed phases are obtained. Three generation mechanisms of dispersed gas bubbles/water droplets are identified: squeezing by the oil phase, cutting by the droplet/bubble, cutting by the water–oil/gas–oil interface. Based on the gas dissolution rate, the mass transfer coefficients are calculated. It is found that water droplet can significantly enhance the transfer of CO₂ into the oil phase initially. When bubble‐droplet cluster are formed downstream the microchannel, droplet will retard the mass transfer. Other characteristics such as phase hold‐up, bubble velocity and bubble dissolution rate are also discussed. The information is beneficial for microreactor design when applying three‐phase reactions. © 2016 American Institute of Chemical Engineers AIChE J, 63: 1727–1739, 2017     

A new mechanistic model to predict gas–liquid interface shape of gas–liquid flow through pipes with low liquid loading

Devising a new mechanistic method to predict gas–liquid interface shape in horizontal pipes is concerned in this article. An experiment was conducted to find the pressure gradients of air–water flow through a 1‐in. pipe diameter. Comparing results of model with some experimental data available in the literature demonstrates that the model provides quite better predictions than existed models do. This model also predicts flow regime transition from stratified to annular flow better than Apparent Rough Surface and Modified Apparent Rough Surface models for both 1‐ and 2‐in. pipe diameters. The model also leads to reliable predictions of wetted wall fraction experimental data. Although one parameter of new model was evaluated based on air–water flow pressure loss experimental data for 1 in. pipe, it was considerably successful to predict pressure drop, liquid holdup, stratified‐annular transition and wetted wall fraction for other gas–liquid systems and pipe diameters. © 2014 American Institute of Chemical Engineers AIChE J, 61: 1043–1053, 2015     

Inception and termination of the core‐annular flow pattern for oil‐water downflow through a vertical pipe

Different flow patterns for lube oil–water and for kerosene‐water downflow through a vertical glass tube have been analyzed with the help of flow visualization. Core‐annular flow is the dominant flow regime, with oil forming the core, and water is forming the wall film. When the velocities are increased, transition to slug flow and transition to dispersed flow are found. The waves found during the transition to slug flow depend on oil viscosity: axisymmetric bamboo waves are seen in kerosene‐water downflow and the waves are asymmetric in case of lube oil–water flow where they have a cork‐screw shape. Based on the experimental observations, simple mathematical models have been proposed for predicting the flow pattern transition curves. © 2011 American Institute of Chemical Engineers AIChE J, 2012     

Liquid‐Liquid Stratified Flow through Horizontal Conduits

The stratified configuration is one of the basic and most important distributions during two phase flow through horizontal pipes. A number of studies have been carried out to understand gas‐liquid stratified flows. However, not much is known regarding the simultaneous flow of two immiscible liquids. There is no guarantee that the information available for gas‐liquid cases can be extended to liquid‐liquid flows. Therefore, the present work attempts a detailed investigation of liquid‐liquid stratified flow through horizontal conduits. Gas‐liquid flow exhibits either smooth or wavy stratified orientations, while liquid‐liquid flow exhibits other distinct stratified patterns like three layer flow, oil dispersed in water, and water flow, etc. Due to this, regime maps and transition equations available for predicting the regimes in gas‐liquid flow cannot be extended for liquid‐liquid cases by merely substituting phase physical properties in the equations. Further efforts have been made to estimate the in‐situ liquid holdup from experiments and theory. The analysis considers the pronounced effect of surface tension, and attempts to modify the Taitel‐Dukler model to account for the curved interface observed in these cases. The curved interface model of Brauner has been validated with experimental data from the present work and those reported in literature. It gives a better prediction of liquid holdup in oil‐water flows and reduces to the Taitel‐Dukler model for air‐water systems.     

Experimentally‐based constitutive relations for co‐current gas‐liquid flow in randomly packed beds

Experimental Observations on average pulse velocity and frequency in concurrent gas‐liquid (down) flow through randomly packed beds are used to extract constitutive relations for the gas‐liquid interaction and mean curvature terms that appear in a recently proposed volume‐averaged two‐fluid model for bubbly flow. The proposed closures lead to a reasonably quantitative prediction of the average pressure drop and liquid saturation under bubbly flow conditions and in the near pulse regime. In addition, the proposed closures provide realistic estimates for the location of the bubble‐to‐pulse transition in microgravity and in 1g down‐flow and predict the disappearance of the bubbly flow pattern at low liquid fluxes in 1g down‐flow. © 2016 American Institute of Chemical Engineers AIChE J, 63: 812–822, 2017     

Plasma Polymerization Modified Polyvinylidene Fluoride (PVDF) Membrane Development and Characterization for Degumming of Soybean Oil

Unmodified and surface‐modified polyvinylidene fluoride (PVDF) membranes were tested for their ability to degum soybean crude oil and crude oil miscellas. The membrane was modified with 1,1,1,3,3,3‐hexafluoro‐2‐propanol or hexamethyldisiloxane (HMDSO) by radio‐frequency plasma polymerization at 10–100 W glow discharge power and 1–30 min contact time. The membranes were characterized by contact angle measurements, attenuated total reflectance Fourier transform infrared spectroscopy, atomic force microscopy, and scanning electron microscopy. Modification of the PVDF membrane with HMDSO at 60 W power for 5 min increased the interfacial free energy between water and solid surface from 30 ± 2 to 64 ± 2 mJ/m². This membrane was tested for permeate flux and phospholipid rejection with crude oil and different concentrations of miscella. Although formation of the polymer film on the membrane tended to decrease membrane pore size, the modified membrane had an oil flux as good as the unmodified membrane did. In addition, the modified‐membrane improved the phospholipid rejection and removed 76 % of the phospholipids from the crude oil and 81–90 % of the phospholipids from crude oil miscellas.     

Rapid prediction of solvation free energy and vapor pressure of liquid and solid from molecular dynamics simulation

We show that the solvation free energy and vapor pressure, important thermodynamic properties of pure substances in liquid or solid states, can be obtained from short, about 20 ps, molecular dynamics simulations. The method combines the determination of free energy of a chemical in vacuum using the normal‐mode analysis (NMA, energy minimization), and in the condensed phase using the two‐phase thermodynamic (2PT) model. We have examined the calculation results for common liquids and solids, including water, alcohol, acid, aromatics, and alkanes. The results, referred to as 2PT‐NMA, is comparable to those calculated from thermodynamic integration (TI) for liquids, and is readily applicable to solids, where simple TI is not applicable. Furthermore, the free energy from 2PT‐NMA converges (20 ps) much faster than that from TI (1 ns). The new method could be a very useful tool for fast screening of condensed phase pressure from the trajectory of MD simulations. © 2015 American Institute of Chemical Engineers AIChE J, 61: 2298–2306, 2015     

Flowsheet simulation of aqueous two‐phase extraction systems for protein purification

BACKGROUND: Aqueous two‐phase extraction (ATPE) has many advantages as an efficient, inexpensive large‐scale liquid–liquid extraction technique for protein separation. However, the realization of ATPE as a protein separation technology at industrial scales is rather limited due to the large, multidimensional design space and the paucity of design approaches to predict phase and product behavior in an integrated fashion with overall system performance. This paper describes a framework designed to calculate suitable flowsheets for the extraction of a target protein from a complex protein feed using ATPE. The framework incorporated a routine to set up flowsheets according to target protein partitioning behavior in specific ATPE systems and a calculation of the amounts of phase‐forming components needed to extract the target protein. The thermodynamics of phase formation and partitioning were modeled using Flory‐Huggins theory and calculated using a Gibbs energy difference minimization approach. RESULTS: As a case study, suitable flowsheets to recover phosphofructokinase from a simple model feedstock using poly(ethylene glycol)‐dextran (PEG6000‐DxT500) and poly(ethylene glycol)‐salt (PEG6000‐Na₃PO₄) two‐phase systems were designed and the existence of feasible solutions was demonstrated. The flowsheets were compared in terms of product yield, product purity, phase settling rate and scaled process cost. The effect of the mass flowrates of phase‐forming components on product yield and purity was also determined. CONCLUSION: This framework is proposed as a basis for flowsheet optimization for protein purification using ATPE systems. Copyright © 2010 Society of Chemical Industry     

Bubble splitting under gas–liquid–liquid three‐phase flow in a double T‐junction microchannel

Gas–aqueous liquid–oil three‐phase flow was generated in a microchannel with a double T‐junction. Under the squeezing of the dispersed aqueous phase at the second T‐junction (T2), the splitting of bubbles generated from the first T‐junction (T1) was investigated. During the bubble splitting process, the upstream gas–oil two‐phase flow and the aqueous phase flow at T2 fluctuate in opposite phases, resulting in either independent or synchronous relationship between the instantaneous downstream and upstream bubble velocities depending on the operating conditions. Compared with two‐phase flow, the modified capillary number and the ratio of the upstream velocity to the aqueous phase velocity were introduced to predict the bubble breakup time. The critical bubble breakup length and size laws of daughter bubbles/slugs were thereby proposed. These results provide an important guideline for designing microchannel structures for a precise manipulation of gas–liquid–liquid three‐phase flow which finds potential applications among others in chemical synthesis. © 2017 American Institute of Chemical Engineers AIChE J, 63: 376–388, 2018     

Experimental Investigation on the Holdup Distribution of Oil‐Water Two‐Phase Flow in Horizontal Parallel Tubes

An experimental investigation was conducted to study the holdup distribution of oil and water two‐phase flow in two parallel tubes with unequal tube diameter. Tests were performed using white oil (of viscosity 52 mPa s and density 860 kg/m³) and tap water as liquid phases at room temperature and atmospheric outlet pressure. Measurements were taken of water flow rates from 0.5 to 12.5 m³/h and input oil volume fractions from 3 to 94 %. Results showed that there were different flow pattern maps between the run and bypass tubes when oil‐water two‐phase flow is found in the parallel tubes. At low input fluid flow rates, a large deviation could be found on the average oil holdup between the bypass and the run tubes. However, with increased input oil fraction at constant water flow rate, the holdup at the bypass tube became close to that at the run tube. Furthermore, experimental data showed that there was no significant variation in flow pattern and holdup between the run and main tubes. In order to calculate the holdup in the form of segregated flow, the drift flux model has been used here.     

Recovery and Recycling of Isopropyl Alcohol Used in Biodiesel Production From Yellow Mustard Oil

The recovery of solvents used during biodiesel synthesis is an important factor in the economic feasibility and sustainability of the entire process. In this study, we looked at the use of isopropyl alcohol (IPA) for oil extraction and biodiesel production, as well as its potential for recovery and recycling. We found that multistage extraction improved oil recovery, with up to 86% oil yield using four stages of extraction at an IPA:mustard flour (volume:weight) ratio of 1.5:1 at room temperature. Using acid–base‐catalyzed transesterification, 99% of the mustard oil was converted to biodiesel. At the end of this process, IPA was recovered from the azeotrope by salting out using potassium carbonate or sodium carbonate. The solubility behavior of the components was evaluated by means of ternary‐phase diagrams of IPA/water/sodium carbonate and IPA/water/potassium carbonate, which determined their liquid–liquid–solid equilibrium constants at ambient pressure and at room temperature. Using 20% (w:w) potassium carbonate, 95% of the IPA was recovered at 99% purity from a starting mixture of IPA containing 13% water. Azeotropic distillation of the IPA–water azeotrope with 10% potassium carbonate resulted in the recovery of 99% of the IPA at 94% purity. These results suggest that IPA is not only a suitable solvent for mustard‐oil extraction but also for salt‐enhanced azeotropic distillation resulting in near‐complete recovery from aqueous solutions.     

Production of bio‐crude from forestry waste by hydro‐liquefaction in sub‐/super‐critical methanol

Hydro‐liquefaction of a woody biomass (birch powder) in sub‐/super‐critical methanol without and with catalysts was investigated with an autoclave reactor at temperatures of 473–673 K and an initial pressure of hydrogen varying from 2.0 to 10.0 MPa. The liquid products were separated into water soluble oil and heavy oil (as bio‐crude) by extraction with water and acetone. Without catalyst, the yields of heavy oil and water soluble oil were in the ranges of 2.4–25.5 wt % and 1.2–17.0 wt %, respectively, depending strongly on reaction temperature, reaction time, and initial pressure of hydrogen. The optimum temperature for the production of heavy oil and water soluble oil was found to be at around 623 K, whereas a longer residence time and a lower initial H₂ pressure were found to be favorite conditions for the oil production. Addition of a basic catalyst, such as NaOH, K₂CO₃, and Rb₂CO₃, could significantly promote biomass conversion and increase yields of oily products in the treatments at temperatures less than 573 K. The yield of heavy oil attained about 30 wt % for the liquefaction operation in the presence of 5 wt % Rb₂CO₃ at 573 K and 2 MPa of H₂ for 60 min. The obtained heavy oil products consisted of a high concentration of phenol derivatives, esters, and benzene derivatives, and they also contained a higher concentration of carbon, a much lower concentration of oxygen, and a significantly increased heating value (>30 MJ/kg) when compared with the raw woody biomass. © 2009 American Institute of Chemical Engineers AIChE J, 2009     

Low Inclination Oil‐water Flows

Two‐phase liquid flows at +5° inclination from the horizontal were studied experimentally for mixture velocities between 0.7 and 2.5 m/s and input oil fractions between 10% and 90%. The results were compared with a two‐fluid model that includes entrainment. The investigations were performed in a 38‐mm ID stainless steel test section, with water and oil as test fluids. Dual continuous flow (both phases remain continuous with inter‐dispersion) prevailed, while the two‐phase pressure gradient was found lower than the single‐phase oil or water. At low mixture velocities the velocity ratio increased with oil fraction while at high ones it decreased. Compared to horizontal flow, water holdup was higher and frictional pressure gradient lower.     

Linking micro‐scale predictions of capillary forces to macro‐scale fluidization experiments in humid environments

The effects of increasing relative humidity (RH) on fluidization/defluidization are investigated experimentally and understood via particle‐level predictions for the resulting capillary force. Experimentally, defluidization is found to be more sensitive to small changes in RH than fluidization. This sensitivity is captured by a new defluidization velocity U_df, which characterizes the curvature of the defluidization plot (pressure drop vs. velocity) observed between the fully‐fluidized (constant pressure drop) and packed‐bed (linear pressure drop dependence on velocity) states; this curvature is indicative of a partially‐fluidized state arising from humidity induced cohesion. Plots of U_df vs. RH reveal two key behaviors, namely U_df gradually increases with a relatively constant slope, followed by an abrupt increase at RH ～55%. Furthermore, the bed transitions from Group A to Group C behavior between RH of approximately 60–65%. From a physical standpoint, these macro‐scale trends are explained via a theory for capillary forces that, for the first time, incorporates measured values of particle surface roughness. Specifically, a model for the cohesive energy of rough surfaces in humid environments shows the same qualitative behavior as U_df vs. RH for RH <55%, unlike predictions of the cohesive force. Furthermore, the abrupt transition at RH ～60–65% is explained via the previously observed onset of liquid‐like water adsorption, rather than crystal/ice‐like adsorption, onto glass surfaces. © 2016 American Institute of Chemical Engineers AIChE J, 62: 3585–3597, 2016     

Experimental Investigation and Simulation of a Gas‐agitated Sieve Plate Column

The extraction of hydrogen peroxide by means of deionized water from anthraquinone working solution via anthraquinone process was carried out in a gas‐agitated sieve plate extraction column. The effects of the superficial velocity of air, dispersed phase and continuous phase on the overall plate extraction efficiency have been investigated. The corrections for the prediction of the overall plate extraction efficiency were presented. The correction proposed to predict the overall plate extraction efficiency in the air, water, anthraquinone working solution three‐phase system agreed satisfactory with experimental data with a maximum absolute deviation of 5.6 %. A new design method for gas‐liquid‐liquid three‐phase extractors is developed based on the multistage countercurrent extraction model. The calculated data by the model agreed well with experimental data and the average relative deviation was less than 10 %. Moreover, the model was used to predict a gas‐agitated sieve plate extraction column for industrial production of hydrogen peroxide. The results show that the plate numbers of gas‐agitated sieve plate extraction column are 30–40 % less than that of liquid‐liquid sieve plate column.