Monitoring Fracture Saturation With Internal Seismic Sources and Twin Neural Networks

• Published in: Journal of geophysical research. Solid earth 2022-02, Vol.127 (2), p.n/a
• Main Authors: Nolte, D. D., Pyrak‐Nolte, L. J.
• Format: Article
• Language: English
• Subjects: Acoustic emission; Atmospheric particulates; chattering dust; Chemical reactions; Classifiers; Complex media; Dust; Dust storms; Earthquakes; Fields; Fractures; Geophysics; GEOSCIENCES; induced seismicity; Learning algorithms; Machine learning; Monitoring; Neural networks; P-waves; Reduction; Reflected waves; Safety engineering; Saturation; Seismic activity; Seismic waves; Seismicity; twin neural network; uncontrolled sources; Wave analysis; Wave trains
• ISSN: 2169-9313; 2169-9356
• DOI: 10.1029/2021JB023005

Seismic coda‐wave analysis is a well‐developed method for detecting subtle physical changes in complex media by measuring arrival times in the late‐arriving energy from multiply scattered or reflected waves. However, a challenge arises when multiply scattered waves are not sufficiently separated in time from the direct arrivals to provide a clear coda wave train. Additional complications for monitoring changes in fracture systems arise when the signals originate from unsynchronized internal sources, such as natural or induced seismicity, from acoustic emission, or from transportable intra‐fracture sources (chattering dust), that generate uncontrolled signals that vary in arrival time, amplitude and frequency content. Here, we use a twin neural network (TNN also known as a Siamese neural network) for dimensionality reduction to analyze signals from chattering dust to classify the fluid saturation state of a synthetic fracture system. The TNN with shared weights generates a low‐dimensional representation of the data input by minimizing contrastive loss, serving as the input to a multiclass classifier that accurately classifies whether multiple fractures in a fracture system are fully saturated or partially saturated, or whether a change in saturation has occurred in different fractures in the system. These results show that information buried in unresolved codas from uncontrolled sources can be extracted using machine learning to monitor the evolution of fracture systems caused by physical and chemical processes even when the scattered and direct wave fields overlap. Plain Language Summary Earthquakes and human activities in the Earth's subsurface generate internal seismic waves that travel from wave sources to receivers. Along their paths, the waves interact with fractures that scatter the waves and cause them to partially interfere, producing complicated signals that carry information on the geometry and the physical conditions of the fractures. Extracting this information could give important insight into environmental safety of underground engineering projects affected by fracture sets. In this work, we use the unique properties of a twin neural network (TNN) to disentangle fracture information buried in complicated signals. The TNN is a dimensionality reduction technique and multiclass classifier that decides whether a fracture system is fully saturated with water or partially saturated, or whether a change in saturation has occurred in different fractu