Implications of rock criticality for reservoir characterization

There is mounting evidence that self-similarity (scale-invariance) characterizes many geological phenomena (and much of the physical world), particular those associated with the fluid-rock stress interactions of cracks and microcracks in rocks. This pervasive self-similarity is a result of the criticality of the distributions of stressed fluid-saturated crack, microcrack, and low aspect-ratio pores in reservoir rock. Criticality has profound implications for the response of hydrocarbon reservoirs to changes during production, and hence for the concept of reservoir characterization. The bad news is that these implications include: (1) the existence of spatial and temporal heterogeneities at all scale lengths; (2) the inappropriateness of Gaussian statistics (averages are no longer meaningful); (3) the inability to reliably extrapolate from place to plate and from time to time; and (4) the possibility of any known or measured reservoir characteristics degrading with time. These specifically limit the success of conventional reservoir characterization. The good news is that criticality means that: (1) the response of a known reservoir to known changes can be calculated by anisotropic pore-elasticity (APE); (2) the current configuration can be monitored by seismic shear-wave splitting; (3) the response to given changes (waterflood injection pressures, say) can be predicted by APE; and (4) the reservoir can be controlled via feedback by adjusting input parameters (injection pressures, say) to optimise the effects (water flooding, say) monitored by shear-wave splitting. Finally, the technology for monitoring producing reservoirs with high-frequency shear-waves along short ray paths within the reservoir itself is now being developed in the UNIWELL configuration of three-component geophones and source(s) in the same producing well. UNIWELL will allow detailed fluid-fluid and fluid-rock interactions to be monitored around the production well as frequently as necessary. (C) 1999 Elsevier Science B.V. All rights reserved.