Scenario Testing: Exploring Uncertainty Through Sketch-Based Modeling

Abstract

Conventional reservoir modeling workflows are poorly suited to rapid prototyping of alternative geologic concepts, and testing of how these concepts might impact on reservoir behavior. An alternative and complementary approach to traditional workflows is to use sketch-based interfaces and modeling (SBIM) to allow rapid creation of prototype reservoir models. SBIM is a simple and intuitive approach for creating complex 3D geometries that has been adopted in a number of engineering and design fields but is only now being applied to reservoir modeling. In SBIM, the user sketches lines in map or cross-section view that are automatically extrapolated to create parameterized 3D surfaces. The extrapolation can be driven by an underlying conceptual model and/or be constrained by data or further sketched lines. We demonstrate here the prototype geologic sketching software Rapid Reservoir Modeling (RRM). In RRM, geologic heterogeneities are sketched to form a 3D geological model. Sketched surfaces bound volumes which can be meshed to perform rapid calculation of key reservoir properties. The speed and ease of use of RRM enables qualitative and quantitative testing of multiple working interpretations, developed from limited data, in order of minutes to hours. The sketched models allow uncertainty to be readily appraised, and models and figures to be easily updated to incorporate new data or concepts. We demonstrate application of RRM to scenario testing in a meandering fluvial setting. Starting with an initial dataset of facies logs, we present a range of potential interpretations for the continuity and distribution of mud-drape packages within meander-belt deposits. We then create models of these interpretations in RRM, mesh and flow simulate them to estimate the variation in upscaled permeabilities and STOIIP across the range of interpretations. The models are simple but quick to make and analyse, and it is easy to vary the properties assigned to different reservoir facies. We can therefore test both the effect of different conceptual models of reservoir architecture as well as the impact of varied reservoir quality.

AAPG Datapages/Search and Discovery Article #90350 © 2019 AAPG Annual Convention and Exhibition, San Antonio, Texas, May 19-22, 2019