Title of Research Output

Deep Autoencoder Clustering of Anthropogenic Seismicity

Faculty

Engineering

Supervisor Name

Bing Li

Keywords

Machine Learning, Seismicity, Hydraulic Fracturing, Civil Engineering, Data Science

Description

Hydraulic fracturing is one of many methods used around the world to aid in the extraction of oil. Even considering recent advancements in the space, hydraulic fracturing can still be inefficient and even dangerous at times due to volatile seismic activity. This study used data taken from a fracturing site in Western Alberta to train and cluster similar seismic events using a machine learning algorithm. By the end of the study, the algorithm was able to successfully reconstruct seismic waves and cluster them based on characteristics chosen by the algorithm. Potential implications of this study in the hydraulic fracturing process include early detection of dangerous seismic events, as well as faster pinpointing of relevant seismic events to increase efficiency in oil extraction.

Acknowledgements

I would like to thank my supervisor, Dr. Bing Li, the Western USRI program, and the Faculty of Engineering for their support in this endeavour.

Comments

Data taken from this research paper.

Please email me at lwrigh89@uwo.ca for any inquiries relating to this research.

Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 4.0 License.

Document Type

Video

Event Website

https://github.com/lwrigh89/Deep-Autoencoder-Clustering-of-Anthropogenic-Seismicity

Share

COinS
 

Deep Autoencoder Clustering of Anthropogenic Seismicity

Hydraulic fracturing is one of many methods used around the world to aid in the extraction of oil. Even considering recent advancements in the space, hydraulic fracturing can still be inefficient and even dangerous at times due to volatile seismic activity. This study used data taken from a fracturing site in Western Alberta to train and cluster similar seismic events using a machine learning algorithm. By the end of the study, the algorithm was able to successfully reconstruct seismic waves and cluster them based on characteristics chosen by the algorithm. Potential implications of this study in the hydraulic fracturing process include early detection of dangerous seismic events, as well as faster pinpointing of relevant seismic events to increase efficiency in oil extraction.

https://ir.lib.uwo.ca/usri/usri2022/ReOS/189