- 3:37 PM
581b

Transversal Hot Zones in a Shallow Packed-Bed Reactor during Single or Multiple Reactions

Sandhya Sundarram, University of Houston, 4800 Calhoun, Chemical Engineering Department, Building D, Room S 222, Houston, TX 77204 and Dan Luss, Chemical Engineering, University of Houston, Houston, TX 77204.

The evolution and dynamics of hot zones on the top of a shallow packed-bed reactor (10 cm diameter) were studied using infrared thermography. The catalyst bed was packed with one layer of 0.5 wt.% Pd/Al2O3 spherical, 2-3 mm diameter catalyst pellets with the active metal deposited as a thin surface shell. The experiments consisted of the atmospheric oxidation of either propylene or of a mixture of propylene and carbon monoxide under conditions for which steady state multiplicity existed. Preliminary single pellet studies were conducted to select the possible candidates for the study. During the oxidation of propylene, a stream consisting of 1.5% C3H6, 28.5% N2 & 70% O2 was fed to the reactor. The bed was heated till 275C when it reached a uniform ignited state with a conversion of 93%. Slow cooling of the reactor vessel from this state led to the formation of a stationary hot region, separated by a sharp temperature front from the adjacent colder region (ΔT~50C). Additional cooling (flow rate of 1200cc/min), led to the formation of a novel hot zone motion, referred to as a pulse. It consisted of a hot zone that periodically formed from a uniformly extinguished state, grew in size and then contracted till the bed was again uniformly quenched. At a lower flow rate of 1000cc/min, a breathing motion was observed. Here the hot zone alternately expanded and contracted but did not extinguish between two consecutive cycles. As the temperature of the reactor was decreased, the frequency of the oscillations decreased till a uniform quenched state was reached. An interesting long transient of a rotating hot zone was observed when the reactor temperature was rapidly decreased from 250 → 200C (order of minutes). The hot zone that evolved from a uniform ignited state, rotated around the bed at an almost constant angular velocity of about 9.8/min. This angular velocity was almost an order of magnitude faster than the 1.72/min observed during CO oxidation, which was previously studied in the same experimental system (Marwaha et al, 2004). The oxidation of mixtures of CO & C3H6 were conducted in the same shallow packed bed to determine whether the interaction between the two reactions would introduce new behavioral features, different from those exhibited by a moving hot zone when only one reaction is carried out. During these experiments, initially a single reactant was fed to the reactor, either 6% CO or 1.5% C3H6. The reactor temperature was then slowly increased till ignition occurred (at 160C for CO or 275C for C3H6). After a uniform high temperature state was attained, the reactor vessel temperature was slowly decreased and the second reactant was introduced into the feed. The feed composition of the mixture was 6%CO, 1.5% C3H6, 22.5% N2 & 70% O2 in all these experiments. Three qualitatively different types of hot zone motions breathing, anti-phase, hopping, were observed during the oxidation of the feed containing both CO and C3H6. These were similar to those observed during CO oxidation (Marwaha et al, 2004) and different from those observed during C3H6 oxidation, in which only one qualitative hot zone motion existed at a fixed flow rate. When the feed contained a single limiting reactant, a moving hot spot and a corresponding oscillatory reaction rate were observed only close to extinction temperature. The use of the two reactants mixture largely expanded the region in which a hot zone existed and shifted the extinction temperature to lower temperatures. The large difference in the period of the motion of the hot zones formed during the oxidation of CO, C3H6 and the mixture of CO & C3H6 indicate that the speed of the hot zone motion is affected mainly by the kinetics of the reactions and not by the thermal or flow properties through the reactor. It also indicates that the very slow hot zone motion encountered during CO oxidation is a feature of that reaction and not one to be expected in general. Hot zones may form in packed bed reactor due to non-uniform catalytic activity, non-uniform flow, spontaneous symmetry breaking or global coupling. It is difficult to conclusively differentiate between the impacts of each of these phenomena on the observed hot zones. While the impact of non-uniform properties of the bed and of global coupling can be readily understood, there is still a need to gain a better understanding if and when spontaneous symmetry breaking may generate such hot zones, and their magnitude and impact. Further understanding of the above mechanisms is necessary to establish clear operating procedures to circumvent the occurrence of stationary & moving hot zones.