Ontological Foundations for Petroleum Application Modeling

PNEC 18th International Conference on Petroleum Data, Integration and Data Management Conference - 2014.


Before designing a software application, a conceptual model is defined in order to simplify the geological reality and to help understanding it. Such a model does not represent the reality itself but the geoscientist’s conceptualization. For a geologist, the reality that lies behind the words “horizon” or “fault” corresponds to material objects (a layer of small thickness, an interruption of a strata set) observed in the field or on well bore samples. Conversely, for a geophysicist, the same words refer to specific arrangements of pixels on a seismic image. These distinct views hinder of merging or interoperating conceptualizations operated by various types of geoscientists, who intend to refer to the same reality. In order to face this interoperability problem, many efforts have been made to create file format standards to transfer of information from one application to another. For keeping the original intended meanings, ontologies were used for expliciting the semantic of the models and for integrating the data and files generated in the various stages of the exploration chain. In this article, we apply methodologies based on foundational ontologies for analyzing the communication standard formats (LAS, WITSML, PRODML, and RESQML) most used in the construction of reservoir models. We show how the notions of identity, rigidity, essentiality and unity applied to ontological concepts lead the modeler to more precisely define the geological objects in the model. By making explicit the identity properties of the modeled objects, the modeler who applies data standards, can overcome the ambiguities of the geological terminology. Our analysis has helped us in producing an adequate conceptual model of the objects entering into the geological models (e.g., well, wellbore, horizon, fault, lithological unit, porosity properties). The nature and properties of the objects are made explicit in each format and this guides the mapping of these objects from one application to another.

Authors:  Ricardo Werlang(1,2), Mara Abel(2), Michel Perrin(3), Joel Luis Carbonera(2), Sandro Rama Fiorini(2)

(1) ENDEEPER – UFRGS Scientific and Technological Park - CEI - Institute of Informatics – Porto Alegre – RS – Brazil
(2) Institute of Informatics – Universidade Federal do Rio Grande do Sul (UFRGS) Porto Alegre – RS – Brazil
(3) Geosiris SA – 78112, Fourqueux, France