A contribution to the modeling of kraft pulping

Abstract

This paper presents a mathematical model for pulping that considers the effect of alkali diffusion and chemical reaction. An approximation for 3-dimensional diffusion in wood chips is tested against pulping data. A method for predicting pulp viscosity is also presented. The model simulates the effect of key pulping variables on pulp kappa number, viscosity, yield, residual alkalinity and pulping uniformity. Experimental data from a number of sources are successfully predicted by the model.On presente dans cet article un modele mathematique pour la mise en pite qui prend en consideration l'effet de la diffusion des alcalins et de la reaction chimique. Une approximation de la diffusion tridimensionnelle dans des copeaux de bois est testee par rapport a des donnees de mise en pite. Une methode de prediction de la viscosite de la pite est egalement presentee. le modele simule I'effet de variables cles de mise en pite sur le Kapa de la pite, la viscositt, le rendement, I'alcalinite residuelle et l'uniformite du traitement. Des donnees experimentales de differentes sources sont predites avec succ&s par le modele. Keywords: pulping model, difhsion, pulp uniformity, pulp viscosity, simulation.iffusion of alkali from pulping liquor into the wood D chips plays an important role in kraft pulping. The overall pulping rate is determined by the rates of reaction and diffusion. It has been shown that the lignin content of cooked chips varies across chip thickness (Gullichsen et al., 1992). Lignin content for cooked chips is highest at the chip center and lowest at the surface. For chips less than 2 mm thick, pulping is uniform. For thicker chips, alkali consumption rate exceeds the rate of diffusion at pulping temperature. The apparent reaction rate is diffusion-limited. Diffusion limitation leads to nonuniform pulping. The final pulp contains fibers with a nonuniform distribution of kappa number (kappa number is proportional to % lignin in pulp). Layered pulping of 2 to 10 mm thick chips has shown that 2 mm chips produce pulp of kappa 15 while 10 mm chips give kappa 70. The large change in kappa with thickness is mainly due to the role of alkali diffusion. Pulping nonuniformity increases the average kappa of pulp and eventually leads to high screening rejects. Clearly, it is necessary to consider the effect of difision in order to quantitatively predict haft pulping performance. This paper presents a comprehensive model for pulping that considers the combined effect of chemical reaction and alkali diffusion. Some improvements are made to an earlier reaction-diffusion based model by Gustafson et al. (1983). A modified treatment for diffusion is used in this work. A method for predicting pulp viscosity is also presented. Model prediction for four different quantities -kappa, yield, viscosity and residual alkali are compared with experimental data from three sources. Some of the data shown here is reproduced from an earlier paper (Agarwal et al., 1994) to make a more complete presentation of model capabilities. Diffusion in wood...

Modeling intra-particle convection in heterogeneous non-catalytic reacting systems

“…The chemical degradation and dissolution of the solid structure of the wood is responsible for the increase of the porosity of the chips. This has been ignored in previous attempts to describe the intra-particle process heterogeneity (Johnsson, 1970;Tyler, 1982;Gustafson et al, 1983;Gustafson, 1988;Agarwal, 1993;Agarwal, Gustafson, & Arasakesari, 1994;Agarwal & Gustafson, 1997;Li, Phoenix, & Macleod, 1997). In these works a considerable discrepancy between the experimental values and the predicted space averaged proÿles does exist and the range of values anticipated for non-uniformity is far from reality.…”

“…In these works a considerable discrepancy between the experimental values and the predicted space averaged proÿles does exist and the range of values anticipated for non-uniformity is far from reality. This inaccuracy subsists even when di erent parameters for different data sets are employed in the simulation (Gustafson et al, 1983;Agarwal, 1993;Agarwal et al, 1994;Agarwal & Gustafson, 1997). The thicker the chips, the more evident is the weakness of these models.…”

“…The thicker the chips, the more evident is the weakness of these models. Agarwal (1993) (see also Agarwal et al, 1994;Agarwal & Gustafson, 1997) tries to overcome the di culties described in Gustafson et al (1983) by introducing the concept of an "equivalent sphere" with the same speciÿc surface area as the chip but the results, although better, are still unsatisfactory.…”

Sustainable design of biorefinery processes: existing practices and new methodology

“…In these cases, the original processes – a Kraft process, for example – are often enhanced in order to change the activities. So, it can be useful to use Kraft process models from the literature . Moreover, another fundamental issue is the difficulty to establish models which take into account the heterogeneity of biomass raw materials.…”

“…So, it can be useful to use Kraft process models from the literature. 160 Moreover, another fundamental issue is the diffi culty to establish models which take into account the heterogeneity of biomass raw materials. Indeed, feedstocks are generally a mix of diff erent species of biomass.…”