Effects of capillary pressure-fluid saturation-relative permeability relationships on predicting carbon dioxide migration during injection into saline aquifers

摘要

Mathematical modeling and numerical simulations on carbon dioxide plume migration in confined saline aquifers have become important approaches to evaluate storage capacity and efficiency for carbon dioxide sequestration and to predict potential risks of leakage during and after carbon dioxide injection into subsurface formations. The effects of the characteristic functions relating capillary pressures and relative permeabilities to fluid saturations are critical in predicting carbon dioxide plume migration in saline aquifers. In this paper, four well known models, i.e. Brooks-Corey-Burdine model (BCB), Brooks-Corey-Mualem model (BCM), Van Genuchten-Burdine model (VGB), and Van Genuchten-Mualem model (VGM), were incorporated with an immiscible displacement process model to simulate carbon dioxide injection into saline aquifers. Parametric analysis was conducted to identify the influences of parameters in the Brooks-Corey based model on the predictions of carbon dioxide plume migration in a confined saline aquifer. A comprehensive comparison was carried out to reveal the different performances of the models with respect to the evolutions and patterns of carbon dioxide plume in the confined saline aquifer. The simulation results indicate that the saturation distributions of carbon dioxide are drastically sensitive to the capillary entry pressure . However, the parameters and in the Brooks-Corey based model, which are related to pore size distribution and tortuosity ratio of a porous medium, respectively, have little effects on predicting carbon dioxide plume migration in the saline aquifer. Moreover, different models may cause extensive influences on the carbon dioxide distributions at the region that is close to the injection well and at the leading edges of carbon dioxide plume in the saline aquifer. Hence, the results indicate that for the evaluation of the storage capacity and efficiency of a saline aquifer to trap carbon dioxide, the storage capacity of the saline aquifer could be overestimated by using the BCB model, whereas it could be underestimated by using the VGM model. On the other hand, for evaluations of potential leakage of carbon dioxide in saline aquifers, where the movement of the leading edge of carbon dioxide plume is monitored, the VGM model is a safe choice in simulations. These results have potentially significant implications for estimation of carbon dioxide storage capacity of saline aquifers and assessments on potential risks of leakage.