Abstract - The principal purpose of the European Union SAIGUP project (completed in September 2003) was to use synthetic reservoir models to quantify objectively the sensitivity of geological complexity on production forecasts as a function of generic aspects of both the sedimentological architecture and faulted structure of shallow marine hydrocarbon reservoirs. The modelling suite comprises a range of progradational shallow marine reservoir types combined with structural models with different fault patterns, densities and permeability characteristics, upscaled using a number of different treatments. Production using a waterflood was simulated for up to five sets of well locations, resulting in production characteristics for over 35,000 full-field reservoir models. The reservoirs are typical of many North Sea examples, with reserves ranging from ca. 15 to 35 million m3, and recovery factors generally between 30% and 55%.
The approach taken, to build and simulate production in thousands of geologically distinct reservoirs drawn from a well-defined geological parameter-space, allows trends in production behaviour to be examined systematically, since a large number of models are needed for an interpretable signal-to-noise ratio. In the present study we examine the changes in reservoir performance between faulted and unfaulted versions of equivalent sedimentological models, focusing on both portions of the ratio. Unbiased correlations have been established for determining the reduction in discounted reservoir value as a function of a number of (mainly dimensionless) parameters that can be measured in the static flow models, or estimated as a function of the combined sedimentological and structural characteristics of the reservoirs. The “noise” component of the signal-noise-ratio reflects the variability in response arising as a function of reservoir-specific heterogeneities. We find that the uncertainty with which the changes in reservoir value as a function of fault characteristics can be estimated increases in proportion to the expected change.
The geometry of the fault system with respect to the principal flow directions in the reservoir is crucial for understanding the effects of faults on production. Although based on 2D idealisations, streamline theory, combined with a network model for determining 2D directional effective permeabilities as a function of characteristics of faults systems, has provided a reliable framework for interpreting the results qualitatively. The main restriction on applying these methods quantitatively has been including the effects of fault juxtaposition. Flow in a realistic faulted reservoir is a thoroughly 3D problem, and despite encouraging robust estimations of the juxtaposition function as a function of top-level geological characteristics of the reservoirs, inclusion of these estimations into an analytical framework has yet to be achieved. Despite this, we have been encouraged by how applicable existing analytical treatments, usually based on 2D idealisations of flow or fault systems, have been for interpreting the behaviour of the faulted reservoirs both quantitatively and qualitatively. The conceptualisations and tools described in this study should therefore be transportable outside our model parameter space to allow rapid estimation of the influence of faults on flow in clastic reservoirs as a function of well locations and basic sedimentological and structural reservoir characteristics.
Abstract of talk given to:
Future of Geological Modelling in Hydrocarbon Development, Geological Society of London, March 2005.