Abstract

Abstract The often stated benefits of time-lapse seismic are the identification of bypassed oil, detection of untapped compartments, and avoidance of surprise through recognition of pattern imbalance. Visual images of the time lapse change may be sufficient to identify such features in simple cases, but most reservoir production involves complex changes to fluid saturation and pressure, both of which generally change the seismic response of the reservoir rock. Reservoir engineers, using a numerical model of the reservoir, can generate predictions of fluid flow and production, and distinguish between pressure and saturation changes in the reservoir. Through a history matching process that incorporates both production and time-lapse seismic data, an improved reservoir model can be generated that is conditioned to fluid movement between the wells. The improved reservoir model will increase the predictive capability of reservoir simulations, and hence increase the effectiveness of one of the reservoir manager's tools. After describing the integration of time-lapse seismic data and reservoir simulation, this paper concludes with a brief case study in which field production and seismic data were integrated to produce an improved reservoir model. Introduction Time-lapse 3D, or 4D, seismic involves the comparison of 3D images to identify seismic differences that can be related to reservoir fluid changes. Successful projects are ones where the repeatability and resolution of the seismic surveys are sufficient to clearly relate observed seismic differences to reservoir fluid movements, while minimizing seismic differences outside the reservoir where changes are not expected1. While the concept is simple, the application is fraught with difficulties as the seismic data is never as repeatable or as well resolved as one would like. That said, it is important to note that great strides are being made and that several projects around the world have shown that time-lapse seismic can work in a variety of seismic and reservoir conditions encompassing nearly all known producing fields. When time-lapse seismic does work, there are at least three distinct, yet complementary, ways to use time-lapse seismic data to improve reservoir simulation, all of which seek to improve the reservoir model used for reservoir simulation. By improving the reservoir model, the simulations should be more accurate in predicting future reservoir performance and allow the asset to be better managed. The three methods are:Qualitative observation of time-lapse differences, followed by manual reservoir property changes;Qualitative use of time-lapse differences to identify regions that have changed saturation, and then use that information to improve the static characterization of the reservoir; and,Quantitative use of time-lapse differences within an optimization scheme to 'history match' both production history and seismic differences simultaneously. These three methods will be first be described briefly, and then demonstrated on a time-lapse case study from the Gulf of Mexico.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call