Distillation and absorption are so far the only high-capacity processes for the separation of the mixtures in the systems liquid-gas and liquid-vapor and their importance in the near future is therefore undoubtable. Both the absorption and especially the distillation are very energy-consuming processes and their design needs to be carefully optimized, which requires precise design methods.
At present, about half of these processes are performed in packed columns which in comparison with tray columns provide the advantage of higher capacity and lower pressure drop. In principle there exist two approaches to the design of those columns – stage wise design and rate-based approach.
Our research activities target to the improvement of the rate-based approach to the absorption and distillation column design through the usage of the reliable mass-transfer coefficients and proper description of the hydrodynamics of the gas/vapor and liquid phases. In order to reach this goal we continuously develop and improve reliable methods for the measurement and estimation of the hydraulic (pressure drop, liquid hold-up), hydrodynamic (axial mixing) and mass-transfer (kL, kG, aeff) characteristics of structured and random packings under absorption and distillation conditions. The experimental data required for our work are measured on two sets of absorption (DN 150.6 and DN 297) and distillation (DN 150 and DN 300) columns.
Absorption measurements
Standard methods for measurement of the mass-transfer characteristics
The volumetric mass-transfer coefficients, kLa and kGa, and the effective interfacial area, aeff are measured using the proved absorption systems:
- desorption of oxygen from water into the stream of nitrogen for kLa measurements
- absorption of sulfur dioxide from the air to the aqueous solution of sodium hydroxide for kGa measurements
- absorption of carbon dioxide from the air to the aqueous solution of sodium hydroxide for the measurements of the effective interfacial area
The above-mentioned methods have been thoroughly tested. Their advantages and detail description is given in the paper
These methods are regularly used for the measurements of the mass-transfer characteristics of mainly structured packings. The data are used for the improvement of the existing mass-transfer models. The recent results of our research have been published in papers
Rejl, F.J., Valenz, L., Haidl, J., Kordač, M., Moucha, T., 2015. On the modeling of gas-phase mass-transfer in metal sheet structured packings. Chem. Eng. Res. Des. 93, 194–202. Rejl, F.J.; Valenz, L.; Haidl, J.; Kordac, M.; Moucha, T., 2015. Hydraulic and mass-transfer characteristics of Raschig Super-Pak 250Y. Chem. Eng. Res. Des. 99, 20-27.
In the last two years, the standard methods have been improved to enable the measurements with absorption systems with physical properties similar to distillation ones. The kLa measurements have been performed utilizing desorption of oxygen from primary alcohols into the nitrogen broadening the spectra of the physical properties of the liquid phase, namely the surface tension. The kGa measurements have been performed with helium and sulfur hexafluoride as the gas phase providing the wide range of gas phase densities and gas phase Schmidt number.
Axial mixing
Axial mixing is an undesirable phenomenon as it degrades the efficiency of the separation equipment through a reduction of the mass-transfer driving force, and in the limit, a column behaves like a single equilibrium stage (Henley, 2011). In packed columns, the axial mixing can be anticipated in both phases as a result of non-uniform velocity profiles across the column diameter. Its effect on the separation efficiency pronounces with the number of transfer units in the corresponding phase. In our experimental setup, the axial dispersion of gas and liquid phases in packed columns was measured using a dynamic method based on injecting the step concentration impulse of a tracer into the inlet stream and subsequent monitoring of the response patterns at two different points downstream from the injection point. Helium and an aqueous solution of NaCl were used as the inert tracers in the gas and the liquid phase, respectively. The results revealed the significant influence of the liquid-side axial mixing on the separation efficiency of both the random and structured packings.
In the following research, we study the axial mixing effect on the distillation column separation efficiency.
Pressure drop of structured packings
The pressure drop is one of the main characteristics of structured and random packings. It is usually measured either under absorption conditions using air/water system or under distillation conditions and the total reflux conditions. We aim at the development of the general hydraulic model of structured packing applicable for the columns of various diameters and the wide range of operating conditions (absorption and distillation). To reach this goal we are building a broad database of the hydraulic data of structured packings.
Distillation experiments
Overall separation efficiency and the pressure drop
Although the rate-based modelling of the distillation columns provides more possibilities than the conventional stage-wise approach the high uncertainty in the available mass-transfer coefficients (experimental or estimated according to some general model) decreases the reliability of the rate-based column design. For the wider acceptance of the experimental data we therefore also perform the measurements of the overall separation efficiency in the form of HETP and the pressure-drop data measured under the total reflux conditions and the atmospheric pressure. These measurements are performed either with cyclohexane-heptane (C6/C7) or methanol-propanol test systems. After the distillation site was rebuilt it is also possible to perform these measurements under significantly lower (~10 kPa) or slightly higher (200 kPa) pressure.
Volumetric mass-transfer coefficients
The experimental HETP values show only the integral mass-transfer characteristics of the packing and therefore lack the information about the local mass-transfer intensity along the packed bed, i.e. the shape of the concentration profile. Moreover, it is known fact that the HETP depends on the concentration range it is measured in and as a consequence on the height of the packed bed used for its measurement. The causes and consequences of this dependence have been described in the paper
The disadvantages of the stage-wise approach to the distillation column modelling and design can be avoided if the rate-based approach is used instead. Our research group developed the unique method for the evaluation of the volumetric mass-transfer coefficients, kLa and kGa, from the concentration profiles measured directly in the distillation column. This so-called profile method is based on the dependence of the on the change of the mass-transfer resistance distribution with the change of the liquid composition along the column. Due to this change, it is possible to analyze the experimental concentration profiles and deduce the contributions of the individual liquid- and vapor-side volumetric mass-transfer coefficients to the overall separation efficiency. The method was developed in the last 10 years and applied to the experimental data measured with methanol-ethanol, methanol-propanol and ethanol-propanol distillation systems on random and structured packings. The experimental data revealed not the negligible influence of the liquid-side mass transfer resistance. The original method has been derived with the assumption of the plug-flow of the phases. In our later work, the influence of the axial mixing phenomenon on the mass-transfer process has been involved in the method. The profile method is briefly described here or in the detail in our papers.
Our recent research activities are focused on the possibility of the evaluation of the axial mixing effect on the packing separation efficiency. The possibility arises from the fact that by the distillation performed under the total reflux conditions the concentration profiles measured in the liquid and in the vapor phase coincide. The axial mixing of the phases causes the separation of those profiles and by analyzing the spacing of the profiles it seems to be possible to quantify the axial mixing for example by means of the Bodenstein number.
Absorption and distillation analogy
The absorption and distillation are usually treated as analogical processes of the mass transfer between the liquid and gas/vapor phase. The origin of this idea is obvious, in both the processes the liquid flowing down the packing is in the contact with the gas or vapor flowing counter-currently and if the composition of phases is not in the equilibria the mass transfer occurs between the phases. In the rate-based approach, the mass-transfer intensity is expressed as the product of the mass-transfer coefficient and the process driving force – difference the concentration between the phase bulk and the phase interface. The common dependencies of the mass-transfer coefficients on the process conditions (namely flow rates) and system physical properties are expected enabling to re-calculate the mass-transfer coefficients measured under absorption conditions to distillation ones and vice versa. Although the analogy is tacitly assumed its validity has been never verified. Moreover the results of our absorption and distillation experiments shown significantly different dependencies of the mass-transfer coefficients on the corresponding phase flow-rates. For these reasons, we started the project of verification of the absorption and distillation analogy. The absorption and distillation experiments have been performed in the wetted wall column, a device with the known interfacial area which allowed to evaluate the mass-transfer coefficients instead of the volumetric ones eliminating possible effects of aeff on the results. The absorption mass-transfer characteristics have been measured using desorption of oxygen from water, methanol, ethanol and propanol into the nitrogen at various temperatures and absorption of the sulfur dioxide from helium, air and sulfur hexafluoride into the solution of sodium hydroxide providing the conditions similar to those of distillation systems. The experimental data have been correlated in a dimensionless form of liquid and gas Sherwood number.
The concentration profiles along the wetted wall column under distillation conditions have been measured and evaluated by means of the profile method. Atmospheric total-reflux distillation of methanol-propanol and ethanol-propanol test mixtures has been utilized. The evaluated values of distillation kL and kG have been compared with the values predicted using the absorption correlations. The difference of only about 8 % was found for the kG values prooving the concept of the analogy for the gas/vapor phase. For the liquid phase, the difference of about 30 % has been found suggesting that there can be also other phenomena affecting the liquid-side mass-transfer resistance.
General mass-transfer model for structured packings
Our research work aims to the development of the reliable hydraulic and mass-transfer model of structured packings suitable bor both the absorption and distillation conditions. The model will be based and validated on the broad database of absorption and distillation data measured with various test systems. Special attention is given to the influence of non-ideal hydrodynamic, i.e. axial mixing, on the packing separation efficiency and on the possibility to predict its effects using widely used engineering packages like Aspen Plus or ChemCAD. On this topic, we collaborate with multiple European universities and also with packing vendors and engineering companies.