The role of heat transfer and varying physical properties in the laminar flow tubular reactor for homogeneous liquid phase reactions: A numerical study.

Abstract

A numerical model was developed to examine the behaviour of a vertical laminar flow tubular reactor subject to varying physical properties. Analysis of the results of the model using factorial design techniques indicated that the characteristics of the reactor were closely tied to the heat transfer occurring in the reactor. Three regimes of heat transfer behaviour were identified. These three regimes were instrumental in the effect that varying physical properties had upon reactor performance. In the high heat transfer regime it was found that varying properties were only important when there were appreciable changes in the thermal conductivity and density due to changes in composition. The medium heat transfer regime gave similar results; however, under some conditions changes in heat capacity were found to be significant. In the low heat transfer regime physical property variations had the largest effect. Under these conditions temperature dependent properties were capable of affecting reactor performance when large temperature rises occurred. Concentration dependence was also significant. With respect to concentration dependence of the physical properties, density and heat capacity variations were found to be the most significant in this regime. An experimental program was carried out to verify the more obvious effects determined in the theoretical program. Although the experimental results could not completely be modelled through the developed numerical model, some of the major findings of the numerical study were verified. In the experimental investigation the reaction was carried out in an adiabatic reactor. Some of the differences between theoretical and experimental conversions were directly attributable to the sensitivity of the system to the dimensionless heat of reaction and to free convection. To a smaller extent the effect of a varying heat capacity was observed. The correspondence between numerical and experimental results could be improved if the following modifications were made: (1) Modify the numerical model to allow for the possibility of negative or stagnant flows. (2) In conjunction with the above recommendation the range of values for the buoyancy term that are solvable by the numerical model should be increased. Convergence difficulties were encountered for larger values of $\beta\sb{U}$. These were quite likely due to the manner in which the system pressure was solved. A different formulation of the momentum equation may be necessary. (3) Investigate a reaction in which all of the pertinent information regarding the kinetics and physical property data are readily available. Some of the error between the numerical and experimental results could be attributed to the contradictory literature on the hydrolysis of acetic anhydride. The experimental results also indicated that the heat of reaction, heat capacity and density should be as accurate as possible.