Many E&P operations- and business-improvement opportunities exist through real-time production surveillance and optimization, drilling and completions surveillance and optimization, and oilfield-equipment reliability and maintainability. In addition to smart-well technology, three data-management capabilities are crucial to modern E&P operations:

1)     Integrating and harmonizing data of different timescales,

2)     Integrating and harmonizing data across multiple disciplines and diverse subject areas, and

3)     Augmenting the industry’s traditional model-driven methods with data-driven methods, including AI statistical and stochastic approaches.

Drilling, reservoir, and production engineering would be further enabled by combining historical (accumulated over years), tactical (weeks to months), and high-frequency data from historians (sub-seconds to days) with data from shared-earth-modeling and discipline-oriented source systems (geology, petrophysics, reservoir, production, facilities, geographical, financial, ERP, supply chain, procurement, etc.).  The industry’s core deterministic, physics-based methods can be augmented on the fringe with AI empirical methods, by employing a “Modeling plus Mining” strategy.

Mike will explore how integrated asset modeling (IAM) can be combined with historians and massively-parallel-processing (MPP) data-warehousing and data-mining technology to create real-time architectures for E&P operations optimization and performance management. He will also review case studies outside the E&P industry to assess the feasibility and viability of such continuous “Modeling plus Mining” systems.