Kinetics of pyrolysis and combustion of a South African coal using the distributed activation energy model

摘要

Most of the energy requirements in South Africa are met by conversion of coal. Kinetics ofudthese coal conversion reactions are necessary for these processes to be run efficiently. In thisudwork, the kinetics of pyrolysis and combustion of a South African coal have been studied.udThermo-gravimetric experiments were carried out on coal and char samples under nitrogenudand oxygen atmospheres, at different heating rates. The results were used with appropriateudmodels in order to determine the reaction kinetics.udThe Distributed Activation Energy Model (DAEM) is commonly used to describe the coaludpyrolysis process. The model states that coal devolatilizes according to a number of firstudorder reactions, each with unique activation energy (E). An algorithm has been developed toudinvert this model in order to calculate the fraction reacting, E and pre-exponential factor (A)udof each reaction using thermo-gravimetric data. The algorithm was tested on pyrolysis dataudfrom real and simulated TGA experiments. The parameters obtained were used to model theudreaction at different heating rates. It was found that the DAEM is suitable to model theudpyrolysis reactionudFurther scrutiny of the inversion algorithm has shown that the calculation of the activationudenergy is a model-free method. The algorithm was therefore applied to real and simulatedudthermo-gravimetric data for coal combustion. Results show that the DAEM can be used as audmodel-free method to calculate the E of coal combustion. However, the calculation of the Audrequires the use of an appropriate structural sub-model. For this particular coal, the shrinkingudcore model was not suitable to describe the combustion reaction. Finding the correct modeluddid not form part of this work.udIn addition, the assumption of a constant heating rate used in the algorithm was investigated.udExamination of TGA data showed that there was a lag between the program temperature andudthe actual sample temperature. This temperature lag, however small, impacted the heatingudrate of the sample. Instantaneous values for the heating rate were used in the algorithm.udAgain, the algorithm proved able to calculate kinetic parameters.udFinally, data obtained from coal and char combustion reactions was compared. Both the Eudand temperature at the maximum devolatilization rate indicate that raw coal is more suitableudfor use in industrial boilers than char.