Resource and Well Performance Characterization in the Permian Basin with the Application of Artificial Intelligence Techniques

Abstract

Oil and gas resource quantification is at the center of asset valuation. Calculating recoverable resources across the company's Permian acreage is considerably difficult due to size and complexity. This work seeks to quantify resources and predict performance by using a reservoir engineering workflow powered by Artificial Intelligence (AI). This work is an extension of "Efficient Field Development Decisions Driven by Artificial Intelligence: A Permian Basin Example" Presented at URTEC in 2024 (Zhou et al. 2024). This workflow enables portfolio characterization efforts that inform guidelines to optimize portfolio size and value. A relevant portion of our Permian assets are Operated by Others. Therefore, valuation becomes increasingly complex as investment is driven by varied factors. These factors include geological and fluid attributes, engineering parameters, and varying operator philosophies. The workflow leverages Artificial Intelligence to combine geological, completion, and production data to generate data clusters that highlight well-performance differences across extensive target areas. Additionally, machine learning models are used for production forecasting based on historical data, geological characterization, and development strategies. Voronoi algorithms (or partitions) are also used to arrange wells depending on well spacing and development sequencing. This approach can highlight impacts on well performance at contrasting spacing. The workflow streamlines production forecasting, well-performance clustering, type curve generation and economics evaluation. This AI-inspired workflow allows probabilistic production profiles across multiple Type Curve Areas in a short time and the systematic process reduces human bias. The detailed characterization directs the ongoing portfolio optimization efforts that pinpoint areas of unrealized value. This workflow allows for structured analysis of resource and asset valuations, with robustness that makes the resulting recommendations easy to audit and document. The application of this AI inspired workflow improves the efficiency and efficacy of the team with resource quantification and investment decisions. It creates a framework that generates repeatable and consistent results that can be used across teams and basins. In using this methodology, petro-technical professionals can strategically prioritize their efforts and products. These may include, but are not limited to, subsurface characterization, land negotiations, asset valuations, and A&D activity.

Source: semanticscholar

Read the AI summary, key takeaways and discussion on WOBR Quant Research.