Data geo-Science Approach for Modelling Unconventional Petroleum Ecosystems and their Visual Analytics

Abstract

Storage, integration and interoperability are critical challenges in the unconventional exploration data management. With a quest to explore unconventional hydrocarbons, in particular, shale gas from fractured shales, we aim at investigating new petroleum data geoscience approaches. The data geo-science describes the integration of geoscience-domain expertise, collaborating mathematical concepts, computing algorithms, machine learning tools, including data and business analytics. Further, to strengthen data-science services among producing companies, we propose an integrated multidimensional repository system, for which factual instances are acquired on gas shales, to store, process and deliver fractured-data views in new knowledge domains. Data dimensions are categorized to examine their suitability in the integrated prototype articulations that use fracture-networks and attribute dimension model descriptions. The factual instances are typically from seismic attributes, seismically interpreted geological structures and reservoirs, well log, including production data entities. For designing and developing multidimensional repository systems, we create various artefacts, describing conceptual, logical and physical models. For exploring the connectivity between seismic and geology entities, multidimensional ontology models are construed using fracture network attribute dimensions and their instances. Different data warehousing and mining are added support to the management of ontologies that can bring the data instances of fractured shales, to unify and explore the associativity between high-dense fractured shales and their orientations. The models depicting collaboration of geology, geophysics, reservoir engineering and geo-mechanics entities and their dimensions can substantially reduce the risk and uncertainty involved in modelling and interpreting shale- and tight-gas reservoirs, including traps associated with Coal Bed Methane (CBM). Anisotropy, Poisson's ratio and Young's modulus properties corroborate the interpretation of stress images from the 3D acoustic characterization of shale reservoirs. The statistical analysis of data-views, their correlations and patterns further facilitate us to visualize and interpret geoscientific metadata meticulously. Data geo-science guided integrated methodology can be applied in any basin, including frontier basins.