Reservoir simulation

This paper analyzes several configurations of convolutional neural networks suited for predicting upscaled fracture permeabilities and shape factors required to close a dual porosity/dual permeability model.
The authors write that simple and straightforward observations on outcrops can be used to build 3D models that mimic geological relationships accurately.
This paper presents an integrated work flow to model mechanical properties at sufficiently high resolution to honor accurately rock fabric and its effects on height and complexity and, thus, production.

Page 1 of 9