The nonlinear equilibrium isotherm relationship for carbon adsorption has been linearized by expanding it using a Taylor series. The Taylor series expansion is truncated so that the resulting linear form can be substituted into the equation obtained by applying orthogonal collocation technique to the differential equation for adsorption process. Finally, a set of algebraic formulas was derived by using Laplace transform for calculating bulk liquid concentration and interface liquid concentration. The Taylor series replacement iterative procedure is recommended for calculating the true values of bulk liquid concentration. The results from the new approach were verified by comparison with those obtained from Runge-Kutta integration using experimental data reported by other investigators. The proposed algebraic method is applicable for all types of batch adsorption experimental data without any restrictions. Water Environ. Res., 65, 781 (1993).
KEYWORDS:carbon adsorption, surface diffusion, bulk liquid concentration, interface liquid concentration, Taylor series, algebraic formulas.The homogeneous surface diffusion model (HSDM) has been widely applied to simulate the dynamics of the adsorption process for wastewater. It has been developed and analyzed by several researchers (Weber and Chakravorty, 1974;Mathews and Weber, 1975;Crittenden and Weber, 1978;Hand et al., 1983;Traegner and Suidan, 1989) to describe the process of adsorption on granular activated carbon (GAC). The partial differential equation to describe the process of adsorption on carbon in spherical coordinates is linear, but the boundary equation from the isotherm equilibrium condition is nonlinear. Therefore, the mathematical formulation of the HSDM requires numerical evaluation and iterative calculation. The equations for the HSDM involve physical parameters such as the liquid film mass transfer coefficient k f and the surface diffusion coefficient D s , which cannot be measured directly by analytical means. The usual way of determining these kinetic parameters involves the process of iteration to generate the model prediction, and then minimizing the sum of square of the residuals between the model calculations and actual experimental data to obtain K f and D s .In the past, numerical solution of HSDM involved using orthogonal collocation method to discretize the partial differential equation of the HSDM to ordinary differential equations. Those ordinary differential equations and the nonlinear isotherm equation were then solved using numerical methods such as the Runge-Kutta technique. This numerical solution involved a stepby-step iterative procedure. Another approach proposed by Hand et al. ( 1983) involved designing the experiment in such a way as to eliminate the liquid film transfer coefficient. The results of such experiments could be then analyzed to obtain D s , and the liquid film transfer coefficient K f , when necessary, could be calculated from appropriate correlation reported in literature Letterman et al (1974). However, it should be ...