Modelisation et simulation de pyrolyse de pneus usages dans des reacteurs de laboratoire et industriel

The present thesis covers an applied study on tire pyrolysis. The main objective is to develop tools to allow predicting the production and the quality of oil from tire pyrolysis.;The first research objective consisted in modelling the kinetics of tires pyrolysis in a reactor, namely an industrial rotary drum operating in batch mode. A literature review performed later demonstrated that almost all kinetics models developed to represent tire pyrolysis could not represent the actual industrial process with enough accuracy. Among the families of kinetics models for pyrolysis, three have been identified: models with one single global reaction, models with multiple combined parallel reactions, and models with multiple parallel and series reactions. It was observed that these models show limitations. In the models with one single global reaction and with multiple parallels reactions, the production of each individual pyrolytic product cannot be predicted, but only for combined volatiles. Morevoer, the mass term in the kinetics refers to the final char weight (Winfinity) that varies with pyrolysis conditions, which yields less robust models. Also, despite the fact that models with multiple parallels and series reactions can predict the rate of production for each pyrolysis product, the selectivities are determined for operating temperatures instead of real mass temperatures, giving models for which parameters tuning is not adequate when used at the industrial scale.;A new kinetics model has been developed, allowing predicting the rate of production of noncondensable gas, oil, and char from tire pyrolysis. The novelty of this model is the consideration of intrinsic selectivities for each product as a function of temperature. This hypothesis has been assumed valid considering that in the industrial pyrolysis process, pyrolysis kinetics is limiting.;The developed model considers individual kinetics for each of the three pyrolytic products proportional to the global decomposition kinetics of pyrolysables. The simulation with data obtained in industrial operation showed the robustness of the model to predict with accuracy in transient regime, tires pyrolysis, with the help of model parameters obtained at laboratory scale, namely in regards of the trigger of production, the residence time of tires (dynamic production) and the amount of oil produced (cumulative yield). It is a novel way to model pyrolysis that could be extrapolated to new waste materials.;The second objective of this doctoral research was to determine the evolution of specific tires specific heat during pyrolysis and the enthalpy of pyrolysis. The origin of this objective comes from a primary contradiction. With few exceptions, it is acknowledged that organic materials pyrolysis is globally an endothermic phenomenon. At the opposite, all experiments led with laboratory apparatuses such as DSC (Differential Scanning Calorimetry) showed exothermic peaks during dynamic experiments (constant heating rate). It has been confirmed by results obtained at the industrial scale, where no sign of exothermicity has been observed. The Hess Law has also confirmed these results, that globally, pyrolysis is indeed a completely endothermic process. An accurate energy balance is required to predict mass temperature during pyrolysis, this parameter being unbindable from kinetics.;An advanced investigation of char first allowed demonstrating that specific heat of solids during pyrolysis decreases with increasing temperature until the weight loss peak is reached, around 400°C, and then starts increasing again. This observation, combined with the fact that the sample loses weight during pyrolysis is considered as the major cause of the apparition of an exothermic peak in laboratory scale experiments. That is, the control system of these apparatuses generates a bias and an unavoidable overheat of the samples producing this exothermic behavior. It would thus be an artifact.;On the base of new data on the evolution of global specific heat during pyrolysis, a model of the energy balance has been developed at the industrial scale to determine the enthalpy of pyrolysis. The simulation has shown that a major part of the heat transferred to the pyrolized mass would make its temperature increase. Next, an enthalpy of pyrolysis dependent of weight loss was obtained. Finally, two other terms of enthalpy have been found, namely an enthalpy for the breakage of sulfur bridges and an enthalpy for the stabilization of char when conversion approaches completion.;This research will have allowed establishing a novel general methodology to determine the enthalpy of pyrolysis. More particularly, new clarifications hasve been obtained in regards to the evolution of specific heat of solids during pyrolysis and new enthalpies of pyrolysis, all endothermic, could be obtained, in agreement with the theoretical expectations.;The third research objective concerned the behavior of sulfur during tires pyrolysis. With as a premise that sulfur is an intrinsic contaminant of many types of waste, it is critical to clarify its fate during pyrolysis, in the present case for waste tires. It has been observed in the literature that some quantitative analyses had been presented, but generally, the mechanisms for the distribution of sulfur within the pyrolytic products remain unclear. Thus, it was then not possible to predict the transfer of sulfur to each of the tire pyrolysis products.;The results taken form literature have been complemented with a series of TGA experiments followed by complete elemental analyses of the residual solids. Mass balances have been performed in order to characterize the distribution of elements within the three products (noncondensable gas, oil, and char). A novel parameter has been created during this research: the sulfur loss selectivity. This intrinsic selectivity is a prediction of the distribution of sulfur within the pyrolysis products as a function of temperature.;Three phenomena has been identified that could affect the sulfur loss selectivity. First, the natural devolatilization of sulfur due to pyrolysis. Next, the sulfur devolatilization due to the desulfurization of the solid matrix by hydrogen and finally, the clustering of sulfur in the solid state due to metal sulfidation (zinc and iron). The results have shown that this selectivity reach a limit value of 1 when pyrolysis is limited by the kinetics and in the absence of metal. When the mass transfer is limiting at low temperature (<500°C) the selectivity will be greater than 1. At a temperature over 350°C with the presence of metals, the selectivity will be lower than 1.;It is a useful tool for industrial pyrolysis processes, being a novel indicator for the distribution of contaminants during the pyrolysis of waste. A better comprehension of these mechanisms allows elaborating a better strategy when designing these industrial processes. For example, in light of this research, it could be preferable to pre-treat the tires at lower temperature to eliminate a significant part of sulfur before pyrolyzing them at high temperature. The resulting pyrolytic products would then necessitate a lighter purification post-treatment, being more efficient and more economical.