Abstract
In-situ combustion (ISC) is an effective thermal enhanced oil recovery method. However, it is still not widely implemented in oilfields. One of the factors limiting the wide application of ISC is the challenge in its simulation and prediction. In this work, the oxidation experiments of maltenes and asphaltenes in reservoir rock were performed in the porous media thermo-effect cell (PMTEC) to establish a simplified reaction model based on non-isothermal measurements and to use it in numerical simulation of ISC process. It was found that the oxidation reaction process of oil fractions can be divided into different regions depending on generated self-energy rate and oxygen consumption rates that is up to the temperature. In order to propagate reactions from one mode to another, a specific oxygen consumption per unit mass of oil fractions is required. The average oxygen requirement for crossing LTOad (low temperature oxidation, oxygen addition reactions) boundary into LTC (low temperature combustion) mode was 64 mgO2/g(maltenes) and 10.4 mgO2/g(asphaltenes). To propagate reactions into HTO mode from the LTC mode, it requires about 646 mgO2/g(asphaltenes) for asphaltenes fraction. Moreover, this characterization seems to be a key tool when designing air injection in field pilots. Additionally, it was revealed that asphaltenes are more exothermic and require lower oxygen uptake per unit of temperature increment in comparison to maltenes. Furthermore, the mass conversion data obtained from non-isothermal measurements of oil fractions allow for the estimation of the stoichiometry coefficients of two low temperature oxidation reactions, i.e. oxidation and cocking processes, which can be included into a numerical simulation model to replicate combustion tube (CT) results. The numerical simulation model reveals that the simplified reaction model from a 6-step into a 3-step reaction scheme can reproduce ignition process, temperature profiles, combustion velocity, and fluid production, which thus makes it suitable for the upscaled modelling of ISC.