Characterizing user-defined objects from outcrop and modern system interpretations for stochastic object-based reservoir modelling

Outcrops and modern depositional environments are important analogues for subsurface hydrocarbon-, water- or CO2-sequestration reservoirs, as they supplement limited well- and seismic- data and provide information on connectivity of sandbodies observed in subsurface datasets. Object based modelling is one of a series of methods that is widely used for modelling subsurface facies architecture. A common criticism of object based modelling is the simplistic shapes in the models and the limited dimensional data typically used to describe those objects. Advances in virtual outcrop mapping and availability of remotely sensed data have the potential to generate quantitative geometrical information of sandbody geometries from large volumes of geospatially-constrained sources; however, there are no systematic ways to describe and incorporate bodies from outcrop or satellite imagery for reservoir modelling. Here we present an objective approach to derive geometric attribute and shape information from outcrop and modern elements, by measuring multiple width and centreline deviation parameters in relation to an object’s centreline and integrating those objects directly into existing object-based reservoir modelling packages. We demonstrate the technique on a fluvial outcrop model from the Beckwith Plateau in Utah and an equivalent modern environment from the Mitchell Delta in Australia to produce an object-based reservoir model.