This theoretical study shows that a liquid phase can be developed in and passed through a fixed-bed adsorber during thermal regeneration. The liquid results from passing hot gas from the inlet end of the bed, where the adsorbate is desorbed, towards the cool outlet end, where condensation of the solute occurs. The stage model of an adsorption bed is applied to two examples, the regeneration of an activated carbon bed with adsorbed benzene using hot nitrogen and the regeneration of a 4A molecular sieve bed with adsorbed water using hot methane. DAVID K. FRIDAY andM. DOUGLAS LEVAN Department of Chemical EngineeringUniversity of Virginia Charloitesville, VA 22901 SCOPE Displacement of adsorbed compounds from a fixed-bed adsorber often results from the passage of a hot gas into the column. This process, called thermal regeneration, finds wide application in solvent recovery, dehydration, and liquid treating operations. Following the heating step, the bed is cooled and is then ready for reuse.This work considers theoretically the thermal regeneration of an adiabatic adsorption column with the development of a liquid phase in the bed. The objectives of this paper are, first, to show that the liquid condensate phase can be formed during thermal regeneration and, second, to present a simple yet versatile model that can be used to predict regeneration behavior for various systems and operating conditions. Thorough treatment will lead to improved methods for the design of adsorption systems and to a better fundamental understanding of coupled heat and mass transfer in adsorptive media.The phenomenon of solute condensation has not been investigated previously although a basis for it has been described by Basmadjian et al. (1975). In their study of thermal regeneration of 5A molecular sieve with adsorbed carbon dioxide using nitrogen, a plateau of readsorbed carbon dioxide was predicted for which the adsorbed-phase concentration exceeded the initial loading. Basmadjian et al. referred to this phenomenon as roll-up.Several methods are available that can be applied to this problem. Many recent theoretical analyses of adiabatic processes in fixed-bed adsorbers have been based on the method of characteristics 1965;Rhee et al., 1970;Pan and Basmadjian, 1971;Banks, 1972;Basmadjian et al., 1975 Ikeda, 1979), the packed bed is modeled as a number of stages, plates, mixing cells, or tanks in series. Our model is obtained by discretizing in the axial direction the partial differential equations for conservation of mass and energy in a fixed bed to obtain ordinary differential equations for each stage. As in the method of characteristics approach, the effects of heat and mass transfer resistances are neglected. Since the model used here is based on finite differences, as the number of stages is increased the results for concentration and temperature profiles in the bed approach those given by the method of characteristics.Two systems are examined: the passage of hot nitrogen into a fixed bed of activated carbon with adsorbed benzene, a...

No abstract

Adsorption equilibrium data have been measured for four refrigerant vapors (R-113, R-ll, R-318, and R-22) on BPL activated carbon at three temperatures over a wide range of partial pressures using an automated isotherm apparatus. Special emphasis is placed on the design and operation of the apparatus to ensure the quality of these data. Isotherm data for each refrigerant vapor are analyzed for correct Henry's law behavior and correlated using threeand four-parameter functions with specified temperature dependencies. The functions employed in this study include (1) the Langmuir equation, (2) the Dubinin-Astakhov equation, (3) the virial equation, and (4) the modified Antoine equation. The Dubinin-Astakhov equation provides the best (or very comparable) fits, based on the variance, for R-113, R-ll, and R-318. However, the Dubinin-Astakhov equation does not correctly describe the behavior of the R-22 data, resulting from its incorrect approach to Henry's law. Overall results show that, for adsorption of these refrigerants on BPL activated carbon, those correlations with exponents on the adsorbed-phase concentration terms greater than 1 generate the smallest variances.

Nonisothermal effects are known to be generally important in gas-phase adsorption processes. This paper considers the role of a weakly adsorbed carrier gas in pressure swing adsorption for purification. A combination of experimental and mathematical modeling results is presented in order to understand and describe the influence of the carrier gas on the behavior of the bed. Dry air is fed to beds of BPL activated carbon. Variations consider a feed of pure helium and a packing of glass beads or an empty column. The role of the heat capacity of the end regions of the bed is emphasized. For air with activated carbon, which adsorbs weakly with insignificant separation of nitrogen and oxygen, periodic state behavior leads to subcooling of the feed-inlet end of the bed and, for cycles of moderately short duration, to temperature rectification, or the establishment of a permanent oscillating temperature wave in the bed. The effect of the volumetric purge-to-feed ratio on the extent of this region and degree of subcooling is examined.

Activated carbon beds with adsorbed n-octane are regenerated using hot air. Effects of pressure, feed temperature, and initial loading on temperature and concentration breakthrough curves are studied. For moderate and high values of these variables, effluent vapors are found to be saturated with hydrocarbon during part of the heating step, indicating the presence of liquid within the bed. Regeneration behavior is predicted with reasonable accuracy using a nonisothermal stage model. D. K. FRIDAY and M. D. LeVANDepartment of Chemical Engineering University of Vlrginia Charlottesville, VA 22901 SCOPECyclic adsorption systems used for solvent recovery, dehydration, and some bulk separations rely on in situ thermal regeneration to remove the adsorbate from the adsorbent. This is accomplished by passing either a hot purge gas or steam into the bed.There are few published experimental studies of hot purge gas regeneration for which the phase equilibria relation is known. We have discussed elsewhere LeVan, 1982, 1984) much of the previous work on nonisothermal adsorption and desorption, including the drying of fibers and agricultural products, for which temperature variations are mild. Of more immediate relevance to our work is the experimental study by Basmadjian et al. (1975) in which carbon dioxide and ethane were desorbed from activated carbon beds using hot purge gas. Measurements of concentration and temperature breakthrough curves compared favorably with predictions of an adiabatic model. Jacob and Tondeur (1983) have made similar comparisons for pentanes adsorbed on 5A molecular sieve. LeVan (1982, 1984) have carried out modeling studies which show that a liquid phase can be developed in and passed through a fixed-bed adsorber during thermal regeneration with hot purge gas, and that increases in initial loading of the adsorbent, the regeneration pressure, and the regeneration feed temperature should increase the likelihood that the liquid phase is formed. Basmadjian (1984) has investigated the role that transfer resistances may play on water condensation in gas dehydration beds and has also found that the formation of a liquid phase should be favored by high regenerant temperature and pressure.This paper reports experimental work on hot purge gas regeneration with emphasis on the condensation phenomenon. Results are obtained for the regeneration of a fixed-bed of Calgon type BPL activated carbon with adsorbed n-octane using hot air. Pressure, feed temperature, and initial loading of the adsorbate are treated as variables. Adsorption isotherm data are also measured, correlated, and used to predict regeneration behavior for comparison with observations.The mathematical model used here to interpret breakthrough behavior is a nonisothermal stage model. We are concerned principally with factors that determine the heights of plateaus and the general shapes (abrupt and gradual) and locations of transitions, namely, material balances, the energy balance including heat losses, and phase equilibria relations. The stag...