No abstract

There has been very little knowledge about the state of gas dispersion in the oxygen delignification process, even though this has a major impact on the performance of the reactor. This paper presents a new continu-ous inline method for measuring oxygen bubble size distribution in the reactor, as well as results from studies con-ducted in softwood and hardwood lines. This new measurement worked well, and new information about oxygen bubble size, as well as how different reactor conditions affected the distribution, was obtained. For example: • In the softwood line, the mean volume-weighted bubble size was about 0.1 mm, whereas in the hardwood line, this size was almost 10 times higher. For both lines, there was considerable variation in the measured bubble size over the long term. • For both lines, an increase in mixer rotation speed caused a discernible decrease in the bubble size, and an increase in oxygen charge caused a discernible increase in the bubble size. • In the softwood line, no coalescence of the bubbles in the reactor was observed, but in the hardwood line, some coalescence of the larger bubbles occurred. • In the test conducted in the hardwood line, the use of brownstock washer defoamer caused a discernible increase in oxygen bubble size. • In the hardwood line, reactor pressure had a noticeable effect on the amount of delignification, which indicated that improving mass transfer of oxygen (e.g., by decreasing the oxygen bubble size, in this case) should also have an increasing effect on the delignification.

Oxygen delignification is an essential part of the pulp production process. Delignification occurs with the aid of alkali and dissolved oxygen. Dissolved oxygen is obtained by dispersing oxygen gas into the pulp suspension by using efficient mixers. Little is known about the state of oxygen gas dispersion and its effect on oxy-gen delignification kinetics and efficiency. This paper will present the results for the effect of gas bubble size on the performance of oxygen delignification. The results are mainly based on detailed studies made in a Finnish hardwood mill where the oxygen bubble size distribution could be altered at the feed of the reactor. An essential aspect of these studies was the use of a new continuous inline gas bubble size measurement system to simultaneously determine the bubble size distribution at the feed and top of the reactor. Information about oxygen consumption in the reactor could also be obtained through the bubble size measurements. Accordingly, these studies quantify the effect of oxygen bubble size on the kappa reduction of the pulp. The effect of different chemical factors on the oxygen bubble size is also studied. Finally, the relationship between the gas bubble size and the liquid phase oxygen mass transfer coefficient (kLa) is presented. This connects the bubble size to the kappa reduction rate. Based on the presented modeling approach and the evaluation of practical factors that are not taken into account in the modeling, it was concluded that the volumetric average oxygen bubble size should preferably be smaller than 0.2 mm in practice. The information obtained with the new gas bubble size measurement system and the presented modeling approach give a very new basis for understanding, monitoring, adjusting, and designing oxygen delignification processes.

No abstract