forecasting

The US Energy Information Administration said it expects US crude oil production to stay close to 2025 levels in 2026 before dipping in 2027.

The authors of this paper apply a deep-learning model for multivariate forecasting of oil production and carbon-dioxide-sequestration efficiency across a range of water-alternating-gas scenarios using field data from six legacy carbon-dioxide enhanced-oil-recovery projects.

The leaders of US oil and gas companies in Texas and neighboring states cite regulatory uncertainty, tariffs, and volatile prices as drags on activity.

The agency’s July outlook also showed crude prices holding below $60/bbl for most of 2026.

Best practices are not static; they evolve alongside advancements that redefine what is achievable.

This article presents a comparative study evaluating four machine-learning approaches, including three deep-learning methods, for forecasting gas and condensate production over a 5-year horizon.

This paper presents a workflow that combines probabilistic modeling and deep-learning models trained on an ensemble of physics models to improve scalability and reliability for shale and tight-reservoir forecasting.

The authors of this paper review the advantages of machine learning in complex compositional reservoir simulations to determine fluid properties such as critical temperature and saturation pressure.

Machine learning is refining gas lift production optimization with scalable automated workflow.

Supervised learning was used to develop an ensemble of models that account for historical production data, geolocation parameters, and completion parameters to forecast production behavior of oil and gas wells.