Maximizing Ultimate Recovery: AI/ML Advancements in Unconventional Fracturing and Well Diagnostics
Thursday, 22 October
Room 361 BECF
Technical Session
This session showcases how advanced AI/ML techniques are revolutionizing unconventional fracture design, characterization, and optimization to drastically improve stimulation efficiency. Attendees will learn about real-time multistage fracture optimization using reinforcement learning, fracture propagation modeling via graph neural networks, and AI-driven frac hit assessment and restoration. Join us to catch up on the latest data science and engineering analytics use cases driving the industry toward doubling ultimate recovery.
Session Chairpersons
-
0830-0855 234183Multi-modal Data-driven Fracture Response Surrogate: Integrating Development History And Prior Knowledge For Efficient Co2 Storage Assessment
-
0855-0920 234065Rollout-aware Graph Neural Network-based Simulator For Hydraulic Fracture Propagation
-
0920-0945 233938Real-time Optimization Of Multistage Hydraulic Fracturing Design Parameters Via Ai Surrogate-assisted Data Assimilation And Reinforcement Learning With Coupled Stress Field And Fracture Geometry Inversion
-
1015-1040 233950Conditional Variational Autoencoder For Rapid Prediction Of Cluster-level Fracture Geometry, Proppant, And Conductivity Maps In Hydraulic Fracturing
-
1040-1105 233910Physics-augmented Surrogate Model For Real-time Multi-cluster Fracture Propagation Monitoring And Optimization
-
1105-1130 234160Integrated Data-driven Hydraulic Fracture Modeling Based On Continuous Wavelet Transform (cwt)
-
Alternate 233903A Machine Learning Based Workflow For Frac Hit Assessment, Pattern Recognition, And Production Restoration
-
Alternate 234062A High-fidelity Digital Twin Framework For Proppant Transport: Integrating Gpu-accelerated Barracuda Virtual Reactor Mp-pic With Graph Neural Networks
-
Alternate 233922Rethinking Benchmarking For Unconventional Plays: A Machine Learning Paradigm At Global Scale