Experimental Investigation and Correlation of the Effect of Carbon Nanotubes on Bubble Column Fluid Dynamics: Gas Holdup, Flow Regime Transition, Bubble Size and Bubble Rise Velocity

Highlights

• Effect of carbon nanotube on gas holdup, flow regime transition, bubble size and bubble rise velocity were determined.

• The effect of carbon nanoparticles on flow regime transition is examined.

• The effect of carbon nanoparticles on gas holdup and bubble size were visually investigated.

• Two accurate correlations were developed for prediction of gas holdup data.

Abstract

The gas holdup is an important parameter in industrial applications for design and operation of bubble columns as well as for calculation of the pressure gradient in fluid flow through pipelines. This study presents new experimental gas holdup data by using kerosene as the liquid phase, CO₂ and N₂ as the gas phase, and multiwall carbon nanotubes (one of them without functionalized group and another with carboxyl functionalized group) as the solid phase. Results show that the un-functionalized multiwall carbon nanotube (MCNT) decreases the gas holdup until the concentration of 17 ppm and after that increasing the concentration of nanoparticle increases the gas holdup. The maximum gas holdup for kerosene/CO₂ and kerosene/N₂ systems was 0.3 and 0.29, respectively. The MCNT at 17 ppm concentration decreased the maximum gas holdup to 0.22 and 0.25 for kerosene/CO₂ and kerosene/N₂ systems. Results also show that the multiwall carbon nanotube with carboxyl functionalized group (MCCNT) increases the gas holdup until the concentration of 17 ppm and after that increasing the concentration of nanoparticle decreases the gas holdup. The MCCNT at 17 ppm concentration increases the maximum gas holdup to 0.37 and 0.33 for kerosene/CO₂ and kerosene/N₂ systems. The analysis of transition point of flow from homogeneous regime (bubbly flow regime) to heterogeneous regime shows that the transition point is not significantly affected by the addition of both types of nanoparticles; however, the type of gas has a strong effect on the transition point. Two correlations were also developed for prediction of experimental gas holdup data for two types of nanoparticles in both homogeneous and heterogeneous regimes and analysis of predicted data showed the acceptable accuracy of the developed correlations.

Introduction

A bubble column is an instrument which is usually filled with a liquid phase with or without solid additives in which bubbles of gas phase are flowing through it to trigger a chemical or biochemical reaction. Bubble columns can be used in homogeneous or heterogeneous flow regimes and they are currently widely utilized in petrochemical and chemical industrial processes due to their simple design, easy operations and flexibility in terms of the residence time of liquid phase.

Gas holdup is one of the key parameters in different industrial processes including construction of bubble columns and in design problems and liquid metering calculations in the field of petroleum and chemical engineering. Gas holdup is the gas fraction of a unit volume in which the gas is flowing. This parameter is required to calculate the average linear velocities of individual phases and the difference of their velocities which is named “slip velocity”. The energy transfer between liquid and gas phases in the interface of them is due to the gas slippage over liquid (Eaton et al., 1967).

Various investigations have been found in the literature on experimental measurement of gas holdup in various systems. Most of reported works focus on gas holdup behavior in two phase gas-liquid systems which are applicable in other fields such as chemical and biochemical engineering.

Fig. 1 illustrates that three main flow regimes can be detected in these systems (Anastasiou et al., 2013). According to this figure which typical gas holdup versus the superficial gas velocity is shown, the homogeneous regime is happened at low gas flowrates. After that the flow regime changed to heterogeneous regime by increasing gas flowrate. But this change in flow regime from homogeneous regime to heterogeneous regime does not occur suddenly. It means after homogeneous regime; transition regime follows where the rate of gas reached. The visual observations illustrate that in homogeneous regime, coalescence between bubbles is negligible and various size of bubbles are formed (Anastasiou et al., 2013).

Literature works suggests that adding various additives and particles to liquid phase has many influences on the gas holdup values in bubble columns.

Saleh (1997) investigated the effect foaming phenomena on the flow behavior in water/air system. They observed that the addition of foam decreases the superficial gas velocity at the onset of liquid fall back (Saleh, 1997). Duangprasert et al. (2008) studied the effect of addition of surfactant on flow regimes in water/air system. They concluded that surfactant has no significant effect on boundaries of churn and annular flow regimes, but reduces the pressure drop in the churn-slug flow regime at superficial gas velocities smaller than 1 m/s. Van Nimwegen et al. (2015) studied the effect of addition of surfactant on liquid holdup and the dynamics of pressure gradient in annular and churn flow through vertical pipe in water/air system. They observed three characteristic behavior depending on the gas and liquid flow rates.

Hikita et al. (1980) investigated the fractional gas holdup in a single nozzle gas distributor system in the presence of various gases and pure, aqueous electrolyte and con-electrolyte liquid phases. They also developed a correlation for prediction of fractional gas holdup based on their experimental data. Jamialahmadi and Müuller‐Steinhagen, 1993 investigated the gas holdup and bubble coalescence behavior with plastic particles as solid phase. They investigated the effect of solid particle size, concentration, wettability and density on the gas holdup and bubble coalescence. They found that in the presence of non-wettable particles in the air/water mixture upholds the bubble coalescence and decreases the gas holdup and the presence of wettable particles overturns bubble coalescence and increases gas holdup. Kojima et al. (1997) investigated the effect of pressure and diameter of single nozzle gas distributor on the gas holdup and mass transfer coefficient with various liquid phases. They observed that the increase in pressure increases the gas holdup and mass transfer coefficient. They also observed that the effect of pressure on gas holdup and mass transfer coefficient increases with a decrease in the diameter of single nozzle. Moshtari et al. (2009) investigated the gas holdup and bubble behavior with a porous plate sparger and a perforated sparger with the same porosity with different gas velocity. They observed that the increase in gas velocity increases the gas hold up. In addition to abovementioned experimental works, a number of researchers also studied the effect of different parameters in systems (Lau et al., 2013; McClure et al., 2015; McClure et al., 2014a; McClure et al., 2014b; Pourtousi et al., 2014).

Lau et al. (2013) work showed a transition in bubble size from a unimodal distribution to a bimodal distribution when the superficial gas velocity increases. They suggested that the change in the bubble size distribution can be related to the increase in coalescence or break-up. McClure et al. (2014a) investigated the effect of surfactant on the performance of bubble columns. They concluded that the presence of surfactant changes the overall hold up due to the increase in drag force on bubbles. They also observed a decrease in the mean bubble size in solutions containing surfactants. They reported that the accumulation of surfactants at the interface of air and water effectively inhibit coalescence of bubbles which will lead to reduction in bubble size. McClure et al. (2014b) developed a CFD model for simulation of bubble column bioreactors. They tried to obtain better description and modeling of the bubble induced turbulence in bubble columns by employing more terms to the turbulence models. However, it they observed that the new models exhibit more prediction and simulation errors and hence are not appropriate to the current systems. McClure et al. (2015) have concluded that identifying the influence of the measurement location on the macroscopic mixing time is difficult and challenging. Pourtousi et al. (2014) reviewed the impact of interfacial forces and turbulence models for the purpose of investigating flow patterns behaviors in bubble columns. They noticed that that the lift force has higher influence on the flow regimes compared to other interfacial forces. They concluded that the lift force enhanced the accuracy in estimation of the turbulent kinetic energy, axial liquid velocity, and radial gas hold-up, and the lift coefficient (denoted by C_L) lies within the range between 0.1 to 0.5 for bubbly flow regime of small spherical bubbles.

With the increasing applicability of nanoparticles in different fields, they have been also utilized as the additive particles in the bubble column systems. Su et al. (2009) investigated the effect of SiO₂ nanoparticles on bubble dynamics where they used needles to inject bubbles into the liquid column. They observed that the size of bubbles in pure water are larger than their size in nanofluids. In another work, the evolving and departure characteristics of a single bubble in a boiling process with nanofluid containing Al₂O₃ and pure water was studied. Dousti et al. (2021) investigated the influence of SiO₂, Al₂O₃ and Fe₃O₄ nanoparticles on bubble characteristics in bubble columns. They also observed that the presence of nanoparticles reduces the bubble size compared to pure water.

To the best of author's knowledge there are few papers in the literature concerning the gas holdup, flow regime transition, bubble size and bubble rise velocity behavior in kerosene/CO₂ and kerosene/N₂ systems with and without addition of Multiwall Carbon Nano-Tube (MCNT). In our previous work, we evaluated the gas holdup in kerosene/CO₂ and kerosene/N₂ systems. In this work, we also used an intelligent model named adaptive neuro fuzzy inference system trained with particle swarm optimization (PSO-ANFIS) to predict gas holdup data in various systems by using literature experimental data as well as our own measured experimental results (Barati-Harooni et al., 2017).

Hence, to fill this gap, this study highlights the effect of MCNT particles with two different hydrophobicity on gas holdup, flow regime transition, bubble rise and bubble size behavior by using the kerosene as an oil phase and CO₂ and N₂ as a gas phase.