The relevance of various oxidation reactions to the modelling of in-situ combustion in heavy oils was studied in three sets of experiments.
The first set of tests involved temperature-programmed thermal gravimetric scans of three heavy oils and their SARA fractions. A comparison between the scans made in the presence and in the absence of air revealed the temperature regions in which each of the SARA fractions underwent oxygen uptake and then combustion.
In the second set of experiments the effect of various levels of oxygen exposure on the subsequent high-temperature coke formation was investigated. The results clearly showed that low-temperature oxidation had significant and sometimes dramatic effects on the amount of cake formation.
Finally, a series of autoclave tests allowed the products of low-temperature oxidation to be identified within the SARA based reaction scheme. These tests indicated also that asphaltenes apparently underwent low-temperature oxidation more rapidly than the other fractions-a characteristic that could not be detected by thermal gravimetric analysis.
The observed reaction characteristics provided guidelines for visualizing a SARA-based reaction model that would be realistic for the prediction and enhanced control of the in-situ combustion process.
After many years of enhanced oil recovery (EDR) experience. the use of in-situ combustion to produce heavy oils still falls far short of its apparent potential. The process is highly efficient in delivering heat to a reservoir. it has a strong tendency to display gravity-stabilized gas override. and it has an inherently strong gas drive. These features make the process particularly attractive in pressure-depleted reservoirs that are either very thick or very thin.
The application of in-situ combustion has been restricted in part by difficulties in predicting its performance in new reservoirs or under new operating conditions. A major root of this problem is the lack of reliable kinetic parameters for the oxidation reactions ranging from low temperature oxidation (LTD) through the auto ignition temperature (AIT) region. This is especially true for liquid or solid hydrocarbons. for which we know of no systematic LTD and AIT study.
The use of SARA (saturates, aromatics, resins, asphaltenes) fractions as a basis for a chemical reaction model offers a practical option for filling this void in the chemical kinetics. Earlier studies l.2 provided some evidence that each SARA fraction oxidizes at a different rate, and that perhaps characteristically different reactions can be distinguished through the study of the separate fractions. The coke/residue formation by individual SARA fractions was shown to depend strongly on the type of mineral surface present, and was reservoir specific. However, the main contribution to the amount of coke/residue available for combustion came from the preceding oxidation reactions.
The characteristics of the various oxidation reactions in heavy oils were studied through three very different sets of tests. The first set involved normal then no gravimetric analysis, whereby the oxidation behaviour of each fraction was inferred over the entire temperature range of interest. In the second set of experiments, the measurements focused on the amount of coke/residue remaining after a SARA fraction had been heated under oxygen-poor conditions.