Building Reservoir and Wellbore-based Earth Models in the Presence of Sparse and Missing Data in Unconventional Resource Plays; An Integration of Data and Physics Driven Workflows

Abstract

Economic success in Oil and Gas business is generated at the bit with successful well placement and completion engineering. An understanding of the rock and fluid physics along with the related data-driven models dictate the degree of uncertainty in resulting models. However, such models are dependent on the availability of reliable local data which is often sparse or missing. Under sparse and missing data conditions, model uncertainty is increased at the very least, and often local model construction becomes difficult if not impossible. This common problem is discussed in the context of current data analytics, machine learning, and imputation practices in combination with well-honed classical geoscience methods.

AAPG Datapages/Search and Discovery Article #90333©2018 AAPG Middle East Region, Shale Gas Evolution Symposium, Manama, Bahrain, December 11-13, 2018