Investigations on gas-liquid ejectors

Abstract

A gas–liquid ejector offers relatively higher interfacial area and can be seriously considered as a cocurrent gas–liquid contactor. A model has been proposed, which assumes the contactor comprises a series of stirred vessels followed by a plug flow reactor. The bubble size in each stirred vessel is assumed to be constant, but varies from vessel to vessel. The gas hold-up hardly varies in the contactor for the systems studied and can be approximated by the ratio of volumetric gas flow rate to the total gas plus liquid flow rate. This may be true for many electrolytes which hinder coalescence of gas bubbles. The model has been specifically derived for fast pseudo–first-order reaction, but can be modified for other situations easily. The absorption of Carbon dioxide in Sodium hydroxide solution has been used to test the model. Various parameters have been varied during experiments and their effect on conversion has been studied. The results can be summarised as: The conversion increases: – as the liquid flow rate to the system is increased, – as the gas concentration to the system is increased, – as the liquid concentration to the system is increased. The length of uniform cross?section tube of the ejector also improves the conversion, as the length increases. FUTURE SCOPE The present model is a very crude one and needs to be strengthened to develop a more robust one. The specific tasks are:

The phenomenon of bubble formation is to be explained and a reasonable model developed. No attempt has been made at understanding the coalescence phenomenon in the divergent section. Gas hold-up has to be measured for a variety of systems containing electrolytes and an adequate model developed for its variation. Though the gas hold-up varies significantly in highly coalescing systems, there is no model which can predict it correctly. Thus empirical expressions are being used. The idea of idealising the contactor as a series of stirred vessels followed by a plug flow reactor seems to function well, but a sounder method for deciding the number of stirred vessels needs to be evolved. This will involve investigations on the liquid and gas mixing inside the contactor.

It would appear that the basic model developed in this investigation has a number of useful features. It needs a series of improvements based on future work to make a more robust model which could be used with a high degree of confidence.