UBC Theses and Dissertations
Fluid flow through packed columns of cooked wood chips Lee, Quak Foo
Kraft batch and continuous digesters are used to react wood chips with an alkali based liquor to remove lignin from the wood and release individual pulp fibres for further processing. Current operating strategies rely on contacting chips with liquors of different temperatures and compositions throughout the cook to optimize pulp quality and strength. This requires uniform liquor flow through the chip column. For batch digesters liquor must reach all chips in a timely manner as cooking conditions are changed. For continuous digesters this requires creation of distinct reaction zones (liquor flow zones) in the descending chip mass where effective chipliquor contacting is made. However, recent findings have shown that the liquor flow is likely non-uniform, particularly in larger digesters. Indications of this in continuous digesters include corrosion in the lower digester, temperature non-uniformity around the vessel circumference, and variation in kappa number produced as a function of time. A key parameter affecting liquor flow is the flow resistance through the chip bed as a function of operating conditions. In order to understand this, pressure drop was measured as a function of superficial velocity, void fraction, kappa number, and compacting pressure. White spruce chips were used in this study. The chips were characterized and cooked into four different furnishes defined as follows: 100% accepts, 87.5% accepts + 12.5% pins, 75% accepts + 25% pins, and 100% pins. The main results of these experiments are as follows. First, chips are more easily compressed as the lignin content (kappa number) of the chips decreases. The void fraction of the chip column decreases with decreased kappa number and increased compaction force applied to the chip column. Second, the flow resistance increases with decreased kappa number, increased superficial velocity, and increased compacting pressure. Furthermore, pressure drop was a function of the particle size distribution of the chips. The addition of pin chips to the accept chips could dramatically increase the pressure drop in the chip column. Indeed a mixture of chip types (75% accepts + 25% pins; 87.5% accepts + 12.5% pins) had a pressure drop greater than either pure pin or accept chips alone. Third, it was found that while a modified Ergun equation (according to Harkonen (1987)) could be used to correlate our experimental data, the predicted constants of Ri and R2 were larger than those predicted by the Harkonen equation and differed depending on the chip mixture and kappa number. Further, the use of d.32 (Sauter mean diameter of particle) for the chip particle size distribution did not permit the pressure drop to be correlated by the Ergun equation with constant values of A and B. These difficulties are likely due to the compressibility of the chip column, which is not accounted for in this analysis.
Item Citations and Data