Compression-based Facies Modelling

Abstract

AbstractSimple object- or pixel-based facies models use facies proportions as the constraining input parameter to be honored in the output model. The resultant interconnectivity of the facies bodies is an unconstrained output property of the modelling, and if the objects being modelled are geometrically representative in three dimensions, commonly-available methods will produce well-connected facies when the model net:gross ratio exceeds about 30%. Geological processes have more degrees of freedom, and facies in high net:gross natural systems often have much lower connectivity than can be achieved by object-based or common implementations of pixel-based forward modelling. The compression method decouples facies proportion from facies connectivity in the modelling process and allows systems to be generated in which both are defined independently at input. The two-step method first generates a model with the correct connectivity but incorrect facies proportions using a conventional method, and then applies a geometrical transform to scale the model to the correct facies proportions while retaining the connectivity of the original model. The method, and underlying parameters, are described and illustrated using examples representative of low and high connectivity geological systems.