This paper investigates and discusses connate water replacement during EOR chemical processes in micromodels that resemble porous media. Compared to experiments performed in plugs or cores, micromodels facilitate visual access to the displacement process, hence, enable more detailed process description.
The replacement of connate water during aqueous fluid phase injection is a miscible process. During connate water replacement different regions are observed. (1) A mixing zone (mixture of connate water and injected aqueous phase), (2) a pure connate water zone (3) and a zone composed of injected water, only. The different zones are visualized during flooding experiments by labeling the phases with different types of dye and by using a special experimental setup. To calculate concentration and saturations of the different zones listed above image analysis algorithms have been developed and are presented in this work. These algorithms enable qualitative and also quantitative analysis of connate water replacement.
In areas directly contacted by the displacing aqueous phase (e.g. polymer solution) connate water is rapidly replaced. In areas not directly contacted by the displacing aqueous phase the connate water is fully retained even after one pore volume injected. Only continuing aqueous phase injection leads to a full replacement of connate water after several PV of polymer solution injected. It is remarkable that in these areas the connate water replacement takes place without oil mobilization (increase in oil recovery), indicating that the exchange of connate water and injected water is a diffusive process. Further, the exchange of polymer product into stagnant zones can be seen as a polymer retention mechanism.
The results of this study provide new insights into the fundamental understanding of the impact of connate water replacement on oil recovery during chemical EOR. The use of multiple UV-tracers and a special experimental setup enable clear differentiation between different phases and concentration levels during flooding experiments. Development of image analysis algorithms enable quantitative process analysis.