(677a) Predicting and Mitigating Retro-Condensation in TSA Dryers Using Aspen Adsorption

Dryers based on the temperature swing adsorption (TSA) concept, using molecular sieves such as zeolite 4A, are commonly used in the process industries to remove water from process streams beyond what can be achieved by condensation and knockout. While the technology is mature and generally well understood, choices of design parameters and details of cyclic operation are usually presented in general terms, using heuristic recommendations, because the detailed modeling of these cyclic systems is complex and time consuming, often requiring specialized tools. Yet the heuristic guidelines do not apply universally to all TSA configurations and operations, so there is an incentive to model specific implementations. Aspen Adsorption^® (AA), a tool available to many process engineers, provides a convenient means of designing and understanding cyclic adsorption processes. However, one phenomenon that does not lend itself to ready modeling in AA is retro-condensation occurring during the hot regeneration step of a TSA cycle, in which water removed from a hot part of the bed travels to a cooler part of the bed and condenses to form a liquid phase. Depending on the extent of condensation and the temperature at which it occurs, this phenomenon can cause progressive damage to the sorbent, limiting its useful service life. Since AA does not account for phase transitions within the sorbent bed, the tool must be used creatively to represent retro-condensation. In a case study of air drying, we model water condensation using a custom extension of the adsorption isotherm, which allows estimation of the liquid holdup (and associated temperature) as it varies in time and space during regeneration. Starting with a base case featuring significant retro-condensation, we show how changing regeneration parameters such as regeneration gas flowrate, pressure, and the progression of its temperature can reduce the severity of (or ultimately eliminate) the phenomenon.