Geological uncertainties associated with 3-D subsurface models

Subsurface models are generally built from both subjective interpretation and mathematical interpolation/extrapolation techniques. These models are therefore uncertain, but their uncertainty is rarely expressed in a geological forecast. In this paper, an evaluation method of geological uncertainties related to 3-D subsurface models is proposed and tested on a real case. This method is based on the subsurface model, which is considered the most probable prediction (best guess). The various geological interfaces are handled as Gaussian random fields to which a model of spatial variability describing possible fluctuations around the best guess is applied. Several structural constraints, such as the shape of folds and thickness of layers are accounted for in the model. At this point, the local variance can be estimated throughout the study area by application of the simple kriging technique. Finally, the variability is converted into probabilities of occurrence of the various rock masses present in the study area. The probabilities are calculated according to intersection rules governing the stratigraphic sequence of the subsurface model. They enable one to probabilistically model subsurface structures in the form of a three-dimensional (3-D) probability field.