Abstract
Reservoirs are dynamic, not only in the sense that fluid pressures and saturations change during development life, but also in the sense that geomechanical states and therefore absolute permeabilities alter with time. Evidence for this is in the strong correlation between the directionality of reservoir flow in waterfloods and the local orientation of horizontal earth stresses. In addition the correlation coefficients between flow rate histories at pairs of wells in a number of fields indicate that reservoir dynamics have at least some component linked to geomechanics. Coupled geomechanical-fluid flow numerical modelling is able to explain some of the observations and so offers an improved predictive tool for planning and managing waterfloods and determining optimal locations for infill wells. Such modelling is now at the stage when it can soon start to be applied to 3-dimensional problems.
Introduction
One of the most important parameters in designing the pattern of wells for waterflooding a reservoir is the directionality in horizontal flow; i.e. any preferred lateral direction for fluid flow across the reservoir. Recent studies have revealed a strong correlation between the directionality of reservoir fluid flow and the local orientation of modern-day maximum horizontal principal earth stress (Shmax). Even more surprisingly, this correlation holds equally well for the set of reservoirs which would not normally be described as "naturally fractured" as it does for those that obviously do contain open, conductive natural fractures. It has been conjectured that this correlation is explained by coupled processes m which the conductivities of natural fractures and faults (of which, generally speaking, there is a large population even in unfractured reservoirs) are altered by the geomechanical changes induced by the flooding process. According to the concepts of the metastability and self-organized criticality of the lithosphere, perturbation by even minor stress changes is likely. This conjecture was given credence by coupled numerical modelling of the geomechanical, fluid flow and heat flow processes involved in waterflooding; generic modelling of the progress of a flood front around a single injector well gave rise to similar patterns of directionality as observed in the aggregated field data. Coupled modelling therefore provides a potential new tool for improved design of waterfloods and infill drilling projects; this will be further demonstrated later in this paper.
