Abstract
This paper presents a new interpretation methodology of 4D seismic data to determine the steam-chamber distribution generated by the SAGD operation in the Athabasca oil sands. Thin mudstone layers and abrupt changes in facies present difficulties upon SAGD implementation. 3D seismic surveys were conducted in 2002 to aid in understanding of the facies distribution, and in 2006 to evaluate SAGD performance.
Two methods are demonstrated to estimate the steam-chamber areal distribution. The first approach is based on comparison of seismic traveltime maps from the two 3D surveys, and the second method uses the interval P-wave velocity (Vp) from the top to the bottom of the reservoir transformed from the seismic traveltime. We formulated a petrophysical model that expresses Vp as a function of temperature, pressure, and water saturation based on the previous experimental measurements of seismic velocity with oil-sands cores. Scaling factors for Vp reduction were first estimated to adjust the laboratory scale to field scale and distributions of Vp reduction and traveltime changes corresponding to the steam-chamber conditions were calculated.
Vp and traveltime maps that reflect the high pressure and high temperature zones generated by the SAGD process were obtained and distributions of Vp reduction and traveltime were calculated with the petrophysical model in order to decouple composite effects of temperature and pressure. Effects of pressure were assumed to be areally more extensive than temperature effects. By distinguishing high temperature and high pore-pressure zones from low temperature and high pore-pressure zones, the steam-chamber distribution was determined. The steam-chamber distribution obtained by the traveltime approach did not show a very good agreement with the well production performance, while the Vp approach presented consistent results. The bitumen volume in the steam-chamber zone estimated by the new approach was calculated, and compared with the actual bitumen production.
The methodology demonstrated here can be applied to other 4D seismic data at fields under thermal recovery processes.
