A new way of 4D interpretation of depleting reservoirs has been developed, which makes use of timeshifts measured in the overburden to evaluate the condition of the reservoir. This in contrast to conventional 4D interpretation in which overburden timeshifts are treated as a complication that needs to be corrected for.
The new interpretation method is based on the observation in geomechanical modelling that thinner, stiffer or undepleted parts of the reservoir act as stress attractors, which "arch away" overburden load from thicker, more compressible or strongly depleted parts of the reservoir. These stress redistribution effects are present over significant distances in the overburden and cause timeshifts in the overburden, which can be an order of magnitude larger than what is found in the reservoir. In fact, in depleting sandstone reservoirs without significant saturation changes/waterdrive the overburden timeshifts are often the only measurable effects.
The overburden timeshifts can be interpreted with the aid of geomechanical modelling. It appears that typical reservoir conditions like sealing faults, non depleted compartments, GWC/OWC, give very specific "timeshift fingerprints" in 4D. This new way of 4D interpretation by combining geomechanics and geophysics opens completely new avenues in reservoir surveillance.
1. INTRODUCTION
Since its introduction in the late eighties (Nur 1989, Wason et al. 1989) time-lapse seismic has had large impact on surveillance of reservoirs with strong waterdrive or efficient waterflooding: saturation changes were found to lead to significant density and velocity changes in the reservoir and to become clearly visible in 4D as areas of larger amplitudes or as areas with positive time shifts between base and monitor survey. As such time-lapse seismic appeared to be an excellent tool to track waterdrive or water flooding and to identify pockets of hydrocarbons bypassed by the water. Many positive reports can be found in publications, e.g. Koster et al. 2000.
In reservoirs with weak waterdrive or less efficient waterflooding however, it appeared difficult to discriminate saturation effects from depletion or injection effects, which affect velocities in the reservoir as well (Landrø et al. 2001). As the latter effects were governed by effective stress changes in the reservoir, coupled geomechanical and flow modelling were introduced in 4D seismic; first with one way coupling (e.g. Vidal et al. 2002 for gas storage), later also with two way staggered coupling, e.g. in the Belridge field by Minkoff et al. 2004. The stress effects on velocity appeared to be often large when measured on core, but almost negligible when measured in the field for depletion conditions - see Stammeijer et al. 2004. In injection conditions stress effects were far better visible (Par et al. 2000, Landrø 2001, Digranes et al. 2002). Nes et al. 2000 gave a possible explanation for the overrated core results by pointing at core damage effects and providing data from synthetic core representative for "virgin" reservoir conditions that showed much lower stress velocity dependence. Also dispersion effects and geometry upscaling effects are mentioned as cause of the discrepancy between field and core. The bottomline is that stress velocity effects are not only poorly understood, but also might be very small for many of the depleting reservoirs.
