Ontology based multidimensional data warehousing and mining of heterogeneous unconventional-reservoir ecosystems

Abstract

A full understanding of many unconventional hydrocarbon resources is not possible because either there are no datasets or only incomplete or unevaluated. Some resources do not even have datasets from wells that have been drilled for exploration purposes. Specifically, unevaluated information on coal, tight gas, shale gas and gas hydrates, is delaying use of technologies that are in place in the market on a commercial scale. In addition, lack of knowledge makes the environmental impact of exploiting an unconventional resource, unpredictable. As a result of the unknowns involving exploration and development risks, productibility and recovery costs, the development of these global resources is being delayed. Evaluation and organization of data on these unconventional resources are needed for any analysis of petroleum ecosystems. As a solution, we propose a robust data-warehousing and mining approach, supported by ontology. Data from unconventional data need to be gathered in a proactive and systematic way. These multidimensional heterogeneous data can be integrated to explore unknown multiple connections among attributes of multiple dimensions of unconventional resources (from different geographic, geological and production regimes). This paper presents an attempt to make use of ontologies written for multiple dimensions to facilitate connections among unconventional petroleum ecosystems. Fine-grained data assist the data-mining procedures for forecasting, in competent and turbulent markets. Sweet spots may have been hidden in databases. The proposed methodology is robust and may be able to resolve issues associated with mining of sweet spots and uncover them from unconventional resource data warehouses and to help adapt technologies for tapping these sweet spots. If the proposed methodology is successful, it can be applied in any basin for all unconventional reservoir ecosystems present.