Geologic Modeling, Upscaling and Simulation of Faulted Reservoirs Using Faulted Stratigraphic Grids

Abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractA key difficulty associated with modeling faulted reservoirs is honoring both the structure and stratigraphy while simultaneously generating models that are acceptable in terms of practical simulation run times. In this paper, we describe our approach for designing finely gridded, faulted simulation models which we then scale up using a flow based nonuniform coarsening technique that takes into account the impact of the faults. Two field examples of faulted reservoirs are presented to demonstrate the new grid design and upscaling tools and, also, to investigate the impact of faults on flow behavior.An important characteristic of our work process to generate faulted simulation models is that all of the tasks, from horizon and fault mapping to geostatistical reservoir characterization to scale up, are performed within an integrated earth modeling environment. Our grid design and scale up calculations accommodate complex fault networks and non-vertical faults (including reverse faults), although in the current implementation we allow only "logically-vertical" faults; i.e., faults with slip along the w or nominally vertical coordinate in a u, v, w stratigraphic grid. The rationale for specifying logically vertical faults is that they represent a compromise between accurately modeling fault geometry while maintaining computational efficiency when simulating flow in faulted reservoirs.Although fault geometry, which we capture with our grid design and upscaling tools, is an important component of fault transmissibility, it is not the only one. Consequently, we present a technique for modifying fault transmissibility by introducing transmissibility multipliers at fault locations in the grid. We use this technique to generate sealing faults. We then compare reservoir performance predictions based on sealing, partially communicating, and no-fault formulations. Plots of individual well performance from the field examples reveal significant differences between the three formulations.

Building Complex Structural Frameworks

“…The primary differences between our technique to fault modeling and others (Belcher, 1994;Chambers et al, 1999) are not so much in the surface generation as in the fault/fault intersection calculations. Two of the approaches commonly used to calculate intersections rely on pillars or on a binary fault tree.…”

“…One method is to go directly to a grid [3], and the other is to create a set of faulted surfaces and then calculate a grid as an additional step [2]. The first method would appear to be faster, cutting out one of the steps in the workflow, but it requires that gridding parameters be chosen very early in the workflow.…”