Control of the smoldering front temperature in a carbon- and carbonate-containing porous medium in order to limit CO2 emissions

This PhD thesis focuses on the control of the smoldering front propagating in a porous medium containing fixed carbon and carbonates (CaCO3). The main objective is to reduce the front temperature, in situ (oil recovery or gas production from oil shale) or in process (combustion of semicoke), in order to limit the medium decarbonation and the resulting CO2 emissions. The reactive porous medium retained to realize the laboratory experiments is a crushed (0.5 to 2 mm) and pre-pyrolyed oil shale, called semicoke. The front propagates in co-current. The first technique experimentally tested is the addition to the semicoke of an inert material (sand) and/or a reactive material (CaCO3) to vary the contents of fixed carbon and of CaCO3, independently. We show that the increase of the CaCO3 content enables to reduce the temperature to 800 °C, but not below; this does not allow to avoid decarbonation. Bringing down the fixed carbon content enables to reduce the front temperature, see even to reach extinction. In the lowest temperatures of propagation, the decarbonation is strongly limited. On the other hand, the front slows down because it does not use all of the fed oxygen. The second original technique consists in adding CO2 (20 mol.%) to the oxidizer air. We show that for a hot front, the decarbonated fraction is reduced from 100% down to 70%, and the CO production at fixed carbon oxidation is increased; this leads to increase the LCV of the produced gas. For a cold front, the decarbonation which was 20%, is totally avoided by adding CO2. Finely, experiments are proposed in the “reaction trailing” combustion mode, little known and implemented. This mode has the major interest to avoid the reactions of “Lower Temperature Oxidation” prejudicial for oil or gas yields in in situ process. Stable and repeatable experiments are realized with different oxygen fractions in feeding gas. The front temperature is directly linked to this parameter; the decarbonation is clearly limited in this mode of propagation. Two types of modeling are proposed. A mass and thermal balance based on simple analytical expressions enables to evaluate the front temperature and velocity. A numerical model developed by IMFT is based on convective/diffusive heat and mass transfer equations coupled with the oxidation reactions (into CO and CO2) and CaCO3 decarbonation is proposed. It describes in a very satisfactory way the experiments in the “reaction leading” mode with variation of the medium composition (first technique).