Abstract
Sand production poses significant challenges in oil and gas extraction, affecting reservoir integrity and production efficiency.In this study, a novel approach was developed that integrates X-ray computed tomography (CT) imaging and convolutional neural networks (CNNs) to predict both porosity and sand production behavior in sandpacks subjected to varying flow rates and stresses.Using a mixture of sand and sodium silicate solution, sand samples were prepared, consolidated under pressure, and analyzed for sand production behavior under multiphase flow conditions.The CNN was trained to predict porosity based on segmented images of the sand microstructure, achieving accurate predictions that align with observed experimental results.This study demonstrates the feasibility of image-based porosity prediction and provides insights into sand production behavior