Shamsuzzaman Farooq

A predictive model for a kinetically controlled pressure swing adsorption separation process

Abstract The micropore diffusion model, including the concentration dependence of the micropore diffusivity, originally developed to simulate pressure swing adsorption air separation for nitrogen production on a carbon molecular sieve, has been applied to a similar process using the zeolite molecular sieve RS-10. The equilibrium and kinetic parameters for the model have been determined from single-component breakthrough curve measurements. It is confirmed that the concentration dependence of the micropore diffusivity has a large effect on the performance of a zeolite-based kinetically controlled PSA separation process. When this effect is allowed for, the model gives performance predictions which are in close agreement with published experimental data.

Analysis of a piston PSA process for air separation

Abstract The piston driven PSA process offers the potential for achieving productivity improvement by rapid piston action. In the present work, experiments were performed on a laboratory scale piston driven PSA test rig with provisions to vary all the important operating variables, namely, phase angle configuration, stroke length, cycling speed duration and angles of feed introduction and product withdrawal. Air separation on 13X zeolite was chosen as the model experimental system. Experiments with adsorbent particles of two different sizes confirmed that mass transfer resistance is important and may significantly affect the separation performance. A mathematical model was developed to simulate the process. The numerical solution was verified by simulating limiting conditions that had analytical solutions. Some basic model assumptions related to piston motion were verified by comparing with experiments conducted at well-defined limiting conditions like empty column and total recycle. Flow resistance in the connecting tubes seemed to explain the observed difference in both phase and amplitude of the pressure profiles measured at the two ends of the column. The model predictions were generally in good agreement with the experimental observations. The model was used to perform a parametric study in the operating regions that were not covered in the experiments. General inferences are made regarding the operating configurations that are expected to improve system performance.

Effect of velocity variation due to adsorption-desorption on equilibrium data from breakthrough experiments

Adsorption and desorption breakthrough experiments for methane and adsorption breakthrough experiments for ethane were carried out in an activated carbon bed. The experiments were done at two different temperatures and the fraction of the adsorbable component was varied in the range 0.3-63%. An approximate way of accounting for the effect of velocity variation has been proposed which, although has somewhat restricted applicability for the desorption runs, is adequate for the adsorption runs over a wide range

Estimation of overall effective coefficient of heat transfer for nonisothermal fixed-bed adsorption

One- and two-dimensional models for nonisothermal fixed-bed adsorption have been numerically solved. Concentration and temperature profiles obtained from the two-dimensional model are presented. By matching the exit concentration and temperature profiles from the two models, an empirical correlation has been proposed which allows estimation of the overall effective coefficient of heat transfer for the one-dimensional model.

Effect of sorbate–sorbate interaction on micropore diffusion in steady-state adsorption processes

Abstract Most studies dealing with micropore diffusion have assumed zero sorbate–sorbate interactions with good success. However, a recent study (van de Graaf, Kapteijn & Mouliin (1999). Modeling permeation of binary mixtures through zeolite membranes. A.I.Ch.E. Journal, 45 (3), 497–511) has shown that the interactions can have a significant effect. In this paper, effects of sorbate–sorbate interaction in a steady-state microporous membrane process have been studied under a variety of operating conditions and system parameters. The generalized Maxwell–Stefan (GMS) models with and without sorbate–sorbate interaction have been analyzed to develop several useful analytical results. These models are then used to simulate numerically a tubular membrane process with a binary mixture. A detailed comparison of the two models shows that quantitative agreement between them results only under very restrictive conditions of either similar component diffusivities or operation in the linear range of isotherms.

Maxwell–Stefan theory for macropore molecular–diffusion-controlled fixed-bed adsorption

Abstract Maxwell–Stefan theory was applied for simulating a multicomponent, macropore molecular–diffusion-controlled fixed-bed adsorption process. Simulations were also made based on the conventional Fick formulation. By comparing the simulation results with experimentally measured column dynamics for the adsorption of O 2 –N 2 –He mixture on 5A zeolite, it was found that both sets of simulations agree with each other and are in good agreement with the experimentally observed behaviour.

Dispersed plug flow model for steady-state laminar flow in a tube with a first order sink at the wall

Abstract Steady state, laminar flow transport in a tube with a first order sink at the wall involves two dimensions—radial and axial. In this paper, a novel iterative technique has been proposed for reducing such a two-dimensional model to an equivalent one-dimensional dispersed plug flow model. The latter yields an analytical expression for the equivalent axial dispersion and a simple, closed form, but approximate, analytical solution of the original two-dimensional problem. The operating range in which this analytical solution is useful has been investigated for a system with mass transfer at the wall.

Dynamic modelling and optimization of an LNG storage tank in a regasification terminal with semi-analytical solutions for N-2-free LNG

Abstract A comprehensive dynamic model for an LNG storage tank in a typical regasification terminal, operating in holding mode, is presented. The model incorporates LNG recirculation for cooling the transfer lines for loading/unloading. It assumes a hypothetical thin vapour interface in equilibrium with the liquid to compute LNG evaporation, which allows the boil-off gas to be hotter than the tank LNG, as observed in practice. A numerical procedure based on the secant method is implemented in MATLAB for solving the governing ordinary differential equations. For the special case of N 2 -free LNG, a semi-analytical solution is proposed to solve the dynamic model for this relatively complex system and compute the amount of boil-off gas (BOG). The semi-analytical solution is subsequently used to optimize the LNG storage tank and recirculation loop designs.

Optimizing the PSA process of propylene/propane using Neuro-Fuzzy modeling

Abstract Cryogenic distillation is the common method for separating propylene/propane mixtures, but this is highly energy intensive. Some 8-ring silica zeolites, especially pure silica chabazite (SiCHA), are known to show high diffusivity ratio for propylene over propane. In this work, we study the separation of propylene/propane using pure silica chabazite (SiCHA) in a simple 4-step pressure swing adsorption process. An isothermal isobaric micropore diffusion model with concentration-dependent diffusivity has been developed to simulate this kinetically controlled separation. It is first developed and implemented in the multi-physics software COMSOL to simulate different modes of PSA process. In this study, we present a sequential optimization strategy based on neuro-fuzzy model and genetic algorithm (GA) with synergistic combination of COMSOL simulation model to maximize the purity of propylene and propane productions.

Analysis of a tubular zeolite membrane process

Models for analyzing the adsorption and diffusion of a sorbate from an inert carrier gas through a zeolite membrane grown on a porous tubular support are proposed. For a system with negligible support resistance, analytical solutions are developed for the cases of constant and concentration-dependent diffusivities. A Langmuir isotherm model has been assumed and chemical potential gradient is considered as the driving force for diffusion in order to represent the concentration dependence of crystal diffusivity. Possibility of estimating meaningful kinetic parameters from permeance data is also discussed in detail.