Automatic Corrosion Detection Using High-Resolution Magnetic Flux Data for Casing Inspection of Oil & Gas Wells

Abstract

Abstract Well-casing integrity management is critical for the safe and reliable operation of oil and gas wells. High-resolution Magnetic Flux Leakage (MFL) inspection systems generate large volumes of circumferential casing inspection data that must be analyzed to identify corrosion defects and assess well integrity. However, these data are often dominated by strong responses from casing collars, centralizers, scratchers, and other hardware artifacts, while corrosion signatures typically appear as small, localized anomalies that are difficult to identify. This paper presents an artificial intelligence (AI)-based workflow that automatically aligns high-resolution MFL data and detects corrosion defects using Dynamic Time Warping (DTW), a patch-based denoising convolutional autoencoder (DCAE), and object localization techniques. The high-resolution MFL inspection tool used in this study consists of upper and lower inspection modules, each containing 48 tri-axial Flux Leakage (FL) sensors and 48 Discriminator (DIS) sensors, providing 100% circumferential coverage of the casing wall. Due to mechanical motion of the tool and the physical separation between the inspection modules, measurements acquired by different sensor arrays are inherently misaligned, complicating corrosion identification and historical inspection comparison. To address this challenge, an automated preprocessing workflow was developed to identify casing collars, estimate module separation, and align MFL measurements using DTW. After MFL alignment, a patch-based denoising convolutional autoencoder was applied to the circumferential-depth images to enhance weak corrosion signatures and suppress dominant hardware responses. Reconstruction errors were combined with a statistical anomaly score to generate corrosion probability maps, enabling automatic localization of candidate defects through image-based anomaly detection. The proposed workflow transforms high-resolution MFL interpretation from a labor-intensive expert-driven process into a scalable and automated analytics framework. By combining DTW-based alignment with deep-learning-based anomaly detection, the methodology provides improved consistency, enhanced corrosion identification, and more reliable casing integrity assessment for large well populations.