No abstract

The (bulk) removal of carbon-dioxide (CO 2 ) from industrial gases, e.g. natural gas, is usually realized with a reactive absorption technique in which (non-)aqueous solutions of alkanolamines are used.From the absorption rate point of view, primary or secondary amines are preferred. However, in case the costs of regeneration are also taken into account, tertiary amines are much more attractive. In order to combine the specific properties of tertiary and primary/secondary alkanolamines respectively, mixtures of both types of compounds are used. A well known example is the activated methyl-di-ethanol-amine (MDEA)-process in which MDEA is mixed with (small amounts) of piperazine.In this paper mixtures of MDEA with several activators, being primary and secondary amines, are studied with respect to the performance of CO 2 removal from natural gas. The absorption process in a tray column has been simulated. For a number of default cases the impact of the activator on the total number of trays has been calculated. From these simulations the optimal number of trays in combination with the amount of activator-addition can be established. Furthermore, insight is obtained on the mechanism of the absorption steps in mixed amine solutions. It is demonstrated that the working action of the accelerator, the fast reacting amine, is substantially influenced by the partial pressures of carbon dioxide in the gas mixture. Moreover, this effect is strongly depending on the molar fraction of the accelerator.

Frequently applied micro models for gas-liquid mass transfer all assume the presence of a liquid bulk. However, some systems are characterized by the absence of a liquid bulk, a very thin layer of liquid flows over a solid surface. An example of such a process is absorption in a column equipped with structured packing elements. The penetration model was slightly modified, so that it can describe systems without liquid bulk. A comparison is made between the results obtained with the modified model and the results that would be obtained when applying the original penetration theory for systems with liquid bulk. Both physical absorption and absorption accompanied by first and second order chemical reaction have been investigated. It is concluded that the original penetration theory can be applied for systems without liquid bulk, provided that the liquid layer has sufficient thickness (d. d à pen). For packed columns this means, in terms of Sherwood number, Sh ! 4. In case of a 1,1-reaction with Ha. 0.2 an additional second criterion is Sh ! 4 ffiffiffiffiffiffiffiffiffiffiffiffiffi ffi D b =D a p For very thin liquid layers (Sh , 4 or Sh , 4 ffiffiffiffiffiffiffiffiffiffiffiffiffi ffi D b =D a p), the original penetration model may give erroneous results, depending on the exact physical and chemical parameters, and the modified model is required.

No abstract

No abstract