Investigations on oil shale particle reactions

Oil shale research and development has grown in the shadow of the petroleum industry. The uncertainty of petroleum prices, its growing worldwide consumption and limited availability have, motivated many oil shale rich countries to investigate means to produce and use shale oil as an alternative. On the other hand, high shale oil costs, its processing complexities and relatively stable petroleum prices have hampered the establishment of the shale oil industry. Oil is recovered from shale via endothermic reactions, heat for which is generated by combustion of the residual carbon in the spent shale. Oil shale pyrolysers and combustors have generally been designed on an empirical basis. The objective of this work was to produce working mathematical models of raw shale pyrolysis and spent shale combustion, adequate to describe the mechanism by which these reactions occur within oil shale particles, and to investigate the parameters involved. Among these, the most relevant and difficult to obtain are the kinetic ones. Verified models for single particles can then be used to describe oil shale particle reactions in any reactor configuration. A three-dimensional model was developed to describe the transient temperature profile within a cubic shaped shale particle. Also a model for shale devolatilization is presented, based on an unreacted core mechanism. Both models are especially apt for large particles, of the type used in moving bed reactors. A thorough investigation was conducted about the equipment and methods used to obtain pyrolysis kinetic parameters. A standard thermogravimetric apparatus was used to generate these data for two shales: New Brunswick shale, and shale from the Irati Formation in Brazil. The potential of a first order—on kerogen concentration—rate equation to represent shale devolatilization was assessed. A one-dimensional model was developed to describe the transient temperature profile and carbon and oxygen concentration within a particle of spent shale undergoing combustion. The model assumed that oxygen was able to access any part of the particle's interior. Kinetic parameters for shale combustion were also obtained by thermogravimetry using Irati shale. The first order dependence of the combustion process on oxygen concentration was confirmed, and kinetic parameters as a function of temperature were extracted from the results. The models were solved using the method of lines, a standard numerical method for solving sets of parabolic partial differential equations. It was implemented in conjunction with the finite difference method. Models for larger particles were verified by heating and devolatilization experiments with 1.3 cm wide particles suspended in a tube furnace. Most of the experimental work addressed two different shales; one from New Brunswick, Canada, and the other from the Irati Formation, in Brazil.Applied Science, Faculty ofChemical and Biological Engineering, Department ofGraduat