Reservoir property modeling using multiple-point and advanced geostatistics
Where: University of Neuchâtel, Switzerland
When: 21-23 June 2017
Aim of the course: Aquifers and geothermal or oil reservoirs are highly heterogeneous at a broad range of scales. But measurements of hydraulic properties in reservoirs are often sparse and therefore modelers need to interpolate those properties and represent uncertainty. However, often, one can use additional information related to the type of rocks present in the reservoir, the kind of geological environment, and trends.
The aim of this course is to illustrate how one can integrate all the information to build a reservoir property model that is consistent and integrates geological concepts by using in particular the latest and most flexible generation of Multiple-Point Statistics methods, i.e. the DeeSse algorithm.
The course is organized around two practical cases. The first one considers a turbidite environment. Multiple-Point statistics will be used to integrate the knowledge available about the typical geometries of channels and lobes with orientation and proportion trends. Participants will learn how to construct all the required input information and generate an ensemble of stochastic realizations.
In the second case study, participants will deal with a quaternary deposit in which one needs to apply a hybrid methodology to define regions and apply different modeling techniques. In the last steps, participants will learn how to use advanced geostatistical techniques for the simulation of porosity and permeability fields constrained by the geological model. All those steps will be integrated in a workflow.