Knowledge-driven applications for geological modeling

Oil and gas exploration relies for a good part on 3D earth modeling operated from raw seismic data and from information issued from drillings. At present, the resulting models have to be shared by various potential users, who must be able to possibly extend, update, revise or rebuild them in view of additional geological data or according to new geological interpretations. The present paper proposes an knowledge-driven approach for "Shared Earth Models"[SEM, 1998, Shared Earth Model project web site http://www.posc.org/workprgm/sem.shtml.] building which ambitions to share throughout the workflow not only raw data and the various representations of the geometrical objects included in a definite model but also the geological interpretation related to this model. This approach rests on a Geo-Ontology that identifies and formalises the structural geologists' expert knowledge and on a derived abstract descriptor (Geological Evolution Scheme), which enables full sharing of the user's geological interpretation between the various applications. Geological assemblages result from a definite history made of various successive events, which create geological objects. For this reason, the arrangements of these objects and, in consequence, the arrangements of the various geological surfaces present in 3D models verify specific rules, which define a "geological syntax" [Perrin, M.,1998. Geological consistency: an opportunity for safe surface assembly and quick model exploration. 3D Modeling of Natural Objects, A Challenge for the 2000's, 3 (4-5) 1998 June.]. The proposed Geo-Ontology takes into account this particular chrono-spatial structure of geological models. The paper describes the main concepts that are used and defines the syntactic rules to which they must obey. Moreover, it is possible to deduce from the Geo-Ontology that we have defined, a standard geological descriptor (Geological Evolution Scheme=GES), which records the geologist's interpretation of any given geological assemblage and which can be used for automatically building the related 3D model. The resulting methodology ("geological pilot") is exposed and illustrated through an example, showing how a GES can be used all along the workflow that goes from raw data to shared structural and stratigraphical models fully exportable for reuse by other geologists. The knowledge-driven approach presented and the resulting "geological pilot" methodology appear very promising for producing a new generation of earth models, able to be fully shared by multiple, possibly distant users. 3D-model updating or revision will be much easier and the proposed methodology is also promising for kinematics (4D) modeling. (c) 2005 Elsevier B.V. All rights reserved.