TY - THES

T1 - Simulation of a polymer core-flood experiment

AU - Shaymardanov, Marat

N1 - embargoed until null

PY - 2009

Y1 - 2009

N2 - Polymer flooding is the most commonly used Enhanced Oil Recovery (EOR) process. Precise prediction of an EOR project is very important. But sometimes there is not enough time for extensive simulation analysis. In these cases a simple 1D model can be an answer. One simulation run of a polymer core flood experiment takes only about 2-3 minutes. And the whole polymer flood project can be designed in a few days. Then only one real reservoir simulation run is needed for calculating technological and economical effect. But here the question is how reliable 1D simulation is. In this work I compare polymer flooding effects for 1D experiments and 3D models. This research was performed in 3 steps. The first one is history matching to obtain the right oil and reservoir properties. After that, different polymer core flood experiments were simulated to find optimal settings for polymer flooding projects - such as slug size, polymer concentration etc. In the final step, polymer flooding in a 3D model of the reservoir was simulated and corresponding results for 1D and 3D were compared. VIscous fingering effect was studied using 2D-model. Results have shown that a 1D model can be used for designing the polymer flooding process, but some corrections have to be made when transferring from the 1D to 3D model. But in general, 1D results predict 3D performance with a fair degree of correlation. Polymer flooding project for this particular field was not very successful due to relatively small oil viscosity and high permeability of the formation.

AB - Polymer flooding is the most commonly used Enhanced Oil Recovery (EOR) process. Precise prediction of an EOR project is very important. But sometimes there is not enough time for extensive simulation analysis. In these cases a simple 1D model can be an answer. One simulation run of a polymer core flood experiment takes only about 2-3 minutes. And the whole polymer flood project can be designed in a few days. Then only one real reservoir simulation run is needed for calculating technological and economical effect. But here the question is how reliable 1D simulation is. In this work I compare polymer flooding effects for 1D experiments and 3D models. This research was performed in 3 steps. The first one is history matching to obtain the right oil and reservoir properties. After that, different polymer core flood experiments were simulated to find optimal settings for polymer flooding projects - such as slug size, polymer concentration etc. In the final step, polymer flooding in a 3D model of the reservoir was simulated and corresponding results for 1D and 3D were compared. VIscous fingering effect was studied using 2D-model. Results have shown that a 1D model can be used for designing the polymer flooding process, but some corrections have to be made when transferring from the 1D to 3D model. But in general, 1D results predict 3D performance with a fair degree of correlation. Polymer flooding project for this particular field was not very successful due to relatively small oil viscosity and high permeability of the formation.

KW - 1D Simulation 2D Simulation 3D Simulation Polymer Core-Flood Viscous fingering Polymer injection

KW - 1D Simulation 2D Simulation 3D Simulation Polymer Core-Flood Kernflut

KW - Viscous fingering

KW - Polymer injection

KW - Polymer Injektion

M3 - Master's Thesis

ER -