A STUDY OF LIQUID-LIQUID TWO PHASE-FLOW THROUGH MINI-SERPENTINE REACTOR

Abstract

Liquid-liquid flow finds applications in chemical, petrochemical, food, and pharmaceutical industries. In addition, there is a growing interest to explore liquid liquid flow in miniature geometries. The reason can be attributed to the successful utilization of miniaturization as the tool of process intensification. Miniature reactors provide larger interfacial area that increases the rate of transport processes which in turn results in less residence time. At the same time, it requires less inventory, hence, considered safe for hazardous reactions. Therefore, liquid-liquid two-phase flow in micro and mini systems finds wide applications in chemical and biochemical industries. One such liquid-liquid application is the transesterification of triglycerides and methanol/ethanol to produce biodiesel. Biodiesel is defined as mono-alkyl esters of long chain fatty acids. It is a mass transfer limited reaction. As a result, the reaction time needed for the desired conversion is much higher. There is a growing number of investigations carried out to produce biodiesel in miniature reactors. A thorough review of literature on liquid-liquid flow and their applications in miniature geometries are mentioned in chapter 1. It is noted that despite the increasing applications, the literature on liquid-liquid flow reveals no systematic information about how yield varies with different flow patterns in serpentine reactors. Understanding of liquid-liquid two-phase flow in mini channels, helps in designing such kind of miniature reactor. This is the major objective of the present work. In order to do so experiments with different non-reactive and reactive liquid pairs are planned. Detailed experimental procedure and instruments used are discussed in chapter 2. The experiments have been performed in test sections made up of glass capillary with different radius of curvature to diameter (R/D) ratios. In addition, a straight tube reactor of the same length and diameter is considered. Also, a serpentine annulus configuration is tested. As liquid phase properties can play a significant role, experiments are performed with three different liquid pairs namely, kerosene-water, toluene-water, and jatropha oil-methanol. As the viscosities of the oils are widely xvii different, it is expected that such a study will bring out the effect of viscosity on the hydrodynamics of flow. The effect of phase properties on interfacial dynamics of liquid-liquid non-reactive flow has been discussed in chapter 3. It is observed that the range of appearance of flow patterns is governed by the viscosity ratio of the phases. Placing a hydrophilic wire inside such geometry enhances the coalescence of slug and intermittent slug, and subsequent breaking of the drop from tail leads to form slug droplet region for all liquid pairs. In other words, the effect of phase properties is found to be diminished if a hydrophilic wire is placed inside the serpentine geometry. It enhances slug coalescence and subsequent break-up and leads to the formation of slug droplet and dispersed droplet flow patterns across all the liquid pairs. Further, a unified approach is proposed to estimate the pressure drop, which shows an acceptable agreement with the experimental results. Next, the role of flow pattern on the yield of transesterification reaction is discussed in chapter 4. It has been noted that flow regimes play an important role in Fatty Acid Methyl Ester (FAME) yield. Percentage FAME yield depends on the predominance of long slug structure or dispersed drops in later tubes. In the first case, FAME yield decreases and in the latter case FAME yield increases. Maximum FAME yield is achieved for an optimum condition of these two parameters. Both are responsible for influencing the FAME yield. Slug regime was observed to have the highest yield for all types of reactors. For a straight reactor, the optimum yield is 69%. While that in the serpentine reactor with R/D ratio 6 is 87%. Yield increases almost 1.26 times in presence of bend as it induces secondary flow in the liquid film surrounding the slug. Reverse flow in bend generates eddies surrounding the slug. Optimum yield in the serpentine reactor for the slug regime further increases to 98 % with a decrease of R/D ratio from 6 to 3. However, further reduction of the R/D ratio increases the coalescence of slug in downstream tubes. In future, this study can be extended with different other non-edible oils. Also a CFD analysis may be taken up to give better insight of flow dynamics as mentioned in chapter 5.