Limitations of the LDF/equimolar counterdiffusion assumption for mass transport within porous adsorbent pellets

Further Validation of the Quartic Concentration Profile Approximation for Describing Intraparticle Transport in Cyclic Adsorption Processes

“…Hence, it is clear that except for very strong adsorbate systems that exhibit a high degree of mathematical linearity, as defined by Botte et al (1998), the accuracy of an approximate model (or the reliability of a correlation for the cycle time dependence of the LDF mass transfer coefficient), may indeed depend on the type of wave function boundary condition utilized in the ensuing analysis. This important result is in agreement with that reported by Todd and Webley (2002).…”

“…Others have recently shown that the LDF approximation provides satisfactory predictions of intraparticle transport over a wide range of conditions during cyclic operation, especially for long cycle times that are typical of most commercial PSA processes (Rodrigues and Dias, 1998;Sircar and Hufton, 2000;Todd and Webley, 2002). In fact, Ritter and co-workers (Zhang and Ritter, 1997;Botte et al, 1998Botte et al, , 1999Ritter, 2002a, 2002b) proposed a novel approach that involved summing together the gas and adsorbed phase concentrations inside the particle and then applying either parabolic (Zhang and Ritter, 1997;Botte et al, 1998Botte et al, , 1999Gadre and Ritter, 2002a) or quartic (Gadre and Ritter, 2002b) profile approximations to obtain approximate yet more robust modifications of the simple LDF model.…”

“…The point being made here and elsewhere (Todd and Webley, 2002) is subtle but important, as it deals with the fact that in reality an adsorbent particle hardly ever experiences a step change at its surface, with the magnitude corresponding to the feed concentration, especially after achieving periodic behavior. As shown recently by Todd and Webley (2002) the sole use of a square wave function has resulted in some serious errors in the predictive ability of the expressions and correlations developed for the cycle time dependence of the LDF mass transfer coefficient when compared to the exact numerical analyses of, for example, rapid PSA cycles.…”

“…As shown recently by Todd and Webley (2002) the sole use of a square wave function has resulted in some serious errors in the predictive ability of the expressions and correlations developed for the cycle time dependence of the LDF mass transfer coefficient when compared to the exact numerical analyses of, for example, rapid PSA cycles. The same limitations may result when comparing any new approximate model for intraparticle transport to the exact numerical solution when only square wave functions are used.…”

Macropore diffusion dusty-gas coefficient for pelletised zeolites from breakthrough experiments in the O2/N2 system

“…A trade-off between reasonable simulation time and model accuracy had to be made. Errors of only a few percent in working capacity arise when comparing discretisation levels of N r = 5 and 30 over an individual sorbent pellet operated under isothermal (Todd and Webley, 2002) or non-isothermal (Todd, 2003) RPSA conditions. This provides some justification that N r = 5 will be a suitable discretisation level for a first approach simulation of RPSA pilot plant data with the DPM.…”