Simulation study on the Steam-Assisted-Gravity-Drainage (SAGD) preheat phase strategy: Consequences for the SAGD performance

Abstract

The Steam-Assisted-Gravity-Drainage (SAGD) recovery technique is a rather new technology for extracting the extra heavy oil and bitumen deposits in Canada and a lot of research is currently being done on this topic. This master thesis presents a quick introduction to the topic heavy oil in terms of geological setting, different recovery techniques and ways of simulating the SAGD process with the fully coupled thermal reservoir simulator STARS provided by CMG.

Statoil has just recently been granted a preview of the newest option in the STARS simulator called FlexWell. It allows arbitrary well trajectories and different completion designs making it one of the more advanced options out on the market to simulate the SAGD process. It has been tested for both single and dual tubing completion designs and for different values for the circulation rate, maximum allowable injection pressure and pressure differential between the wells. The single wellbore design seemed however to be a bit more robust than the dual tubing design, yielding more reasonable results.

To try and understand more of the different phenomenon that may occur on the way down from the surface and in to the reservoir, there were constructed two different models. One was a simplified reservoir model while the other was extended all the way up to the surface. The surface model was after several runs put aside due to liquid holdup problems resulting in zero steam quality at reservoir depths.

Based on the observed behaviour from varying the different well constraints for the circulation phase, two cases were selected for further studies. The two cases, where the temperature distribution was the most even at the time of steam breakthrough, were chosen to include a SAGD phase as well as the circulation phase.

Determining when to transform the wells into a traditional SAGD well-pair were determined based on several switching criteria. First indication of increased production in the upper wellbore, estimates for when the average midblock temperature was 100 degrees Celsius, 30% loss of circulation and a new plateau rate in the upper wellbore, were the ones investigated in this study. The runs were very unstable and the result not conclusive. Still, it was possible to recognize some indication of which switching criterion had the best influence on the SAGD performance. If the switch is done too early, like for instance at first indication, parts of the reservoir will cool down before the SAGD phase is activated resulting in an uneven development of the steam chamber.