Convective and wall-to-solids heat transfer coefficient in flighted rotary kilns: experimental measurement and modeling

International audienceRotary kilns are gas-solid reactors commonly used in industry to achieve a wide range of material processing operations: mixing, heating or cooling, reacting of coarse, free-flowing or cohesive solids. Therefore rotary kilns are used for applications such as reduction of oxide ore, pyrolysis of biomass or hazardous waste, calcining of petroleum coke, conversion of uranium fluoride for the manufacture of nuclear fuel, and so on. When operated at atmospheric pressure, these units consist of a cylindrical shell that can be inclined, into which the solids burden is fed continuously at one end and discharged at the other. Most of them are equipped with lifting flights or lifters, and/or exit dam at the kiln outlet end. They can be classified into two main heating modes; they can be either directly heated or indirectly heated, depending on the heating source position with respect to the kiln’s tube wall. They usually require very little labor to operate in comparison with other industrial reactors.Though operational cost of these units is usually high, their design and intended operating conditions are often conservative due to the lack of fundamental understanding notably upon the solids flow behavior and the heat transfer mechanisms. Heat transfer in rotary kilns is very complex and may involve the exchange of energy via all the fundamental physical transfer mechanisms that are, conduction, convection, and radiation. There have been quite a few studies dealing with this subject in the literature. Although many researchers studied the main phenomena occurring in the kiln, the heat transfer between the wall and solid particles, or the (free) convection of non forced gas are not yet well understood. The present study investigates the convective gas-to-wall heat transfer coefficient, in the case of a non-forced air flow, and the wall-to-solids heat transfer coefficient. These coefficients were first experimentally determined, and then correlated based on a dimensional analysis, so to be used in a global model for rotary kilns.A series of experiments were carried out on a pilot scale rotary kiln at atmospheric pressure, whether or not equipped with lifters and fitted with a dam at the outlet end. The experimental apparatus, 1.95m in length and 0.01m in (internal) diameter, can be externally heated in two independent consecutive zones by electrical resistance up to 1000°C. Regarding the thermal metrology, thermocouples are positioned at five and four cross-sections, respectively in and outside along the kiln tube. Hence, after turning on the heating system, axial temperature profiles of gas, wall and solids were measured until steady state is achieved. Both coefficients were determined from the temperature profiles measurements data for low and medium wall temperature set point comprises between 100 and 500°C. In particular for the wall-to-solids heat transfer coefficient, the bulk materials used was quartz sand (average size about half a millimeter) with a narrow distributed size fraction; No solids were fed into the kiln when studying the air convection inside the kiln. Two shapes of lifters were compared to determine the influence of lifters presence and their geometry on the heat transfer: straight (one-section) lifters and rectangular (two-section) lifters (RL). The kiln operating conditions examined also include: the rotational speed (2-12 rpm), the mass flow rate (0.8-2.5 kg/h) and the exit dam height at the kiln outlet end (23.5-33.5 mm). An experimental matrix of about eighty experiments was achieved.For the determination of the experimental value, the lumped system analysis and a heat balance accounting for the measured power supplied for the heating are used. Results showed that the wall-to-gas convective heat transfer coefficient is significantly lower that what can be expected for natural convection. Though only small variations were observed, still some trends could be observed in presence of lifters, and when varying the rotational speed. The wall-to-solids heat transfer coefficient was notably lower in presence of lifters. It was also found to increase with the temperature set at the wall and the filling degree, which is imposed by the operating conditions set. Dimensional correlations were developed to describe these two heat transfer mechanisms from the experimental results