Automated fault detection without seismic processing
For hydrocarbon exploration, large volumes of data are acquired and used in physical modeling-based workflows to identify geologic features of interest such as fault networks, salt bodies, or, in general, elements of petroleum systems. The adjoint modeling step, which transforms the data into the mo...
| Main Authors: | , , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Society of Exploration Geophysicists
2017
|
| Online Access: | http://hdl.handle.net/1721.1/110058 https://orcid.org/0000-0001-8467-1888 https://orcid.org/0000-0002-3944-0455 |
| _version_ | 1826206456300437504 |
|---|---|
| author | Araya-Polo, Mauricio Dahlke, Taylor Frogner, Charlie Hohl, Detlef Zhang, Chiyuan Poggio, Tomaso A |
| author2 | Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences |
| author_facet | Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Araya-Polo, Mauricio Dahlke, Taylor Frogner, Charlie Hohl, Detlef Zhang, Chiyuan Poggio, Tomaso A |
| author_sort | Araya-Polo, Mauricio |
| collection | MIT |
| description | For hydrocarbon exploration, large volumes of data are acquired and used in physical modeling-based workflows to identify geologic features of interest such as fault networks, salt bodies, or, in general, elements of petroleum systems. The adjoint modeling step, which transforms the data into the model space, and subsequent interpretation can be very expensive, both in terms of computing resources and domain-expert time. We propose and implement a unique approach that bypasses these demanding steps, directly assisting interpretation. We do this by training a deep neural network to learn a mapping relationship between the data space and the final output (particularly, spatial points indicating fault presence). The key to obtaining accurate predictions is the use of the Wasserstein loss function, which properly handles the structured output — in our case, by exploiting fault surface continuity. The promising results shown here for synthetic data demonstrate a new way of using seismic data and suggest more direct methods to identify key elements in the subsurface. |
| first_indexed | 2024-09-23T13:31:02Z |
| format | Article |
| id | mit-1721.1/110058 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T13:31:02Z |
| publishDate | 2017 |
| publisher | Society of Exploration Geophysicists |
| record_format | dspace |
| spelling | mit-1721.1/1100582022-09-28T14:37:45Z Automated fault detection without seismic processing Araya-Polo, Mauricio Dahlke, Taylor Frogner, Charlie Hohl, Detlef Zhang, Chiyuan Poggio, Tomaso A Massachusetts Institute of Technology. Department of Brain and Cognitive Sciences Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Zhang, Chiyuan Poggio, Tomaso A For hydrocarbon exploration, large volumes of data are acquired and used in physical modeling-based workflows to identify geologic features of interest such as fault networks, salt bodies, or, in general, elements of petroleum systems. The adjoint modeling step, which transforms the data into the model space, and subsequent interpretation can be very expensive, both in terms of computing resources and domain-expert time. We propose and implement a unique approach that bypasses these demanding steps, directly assisting interpretation. We do this by training a deep neural network to learn a mapping relationship between the data space and the final output (particularly, spatial points indicating fault presence). The key to obtaining accurate predictions is the use of the Wasserstein loss function, which properly handles the structured output — in our case, by exploiting fault surface continuity. The promising results shown here for synthetic data demonstrate a new way of using seismic data and suggest more direct methods to identify key elements in the subsurface. 2017-06-20T15:32:51Z 2017-06-20T15:32:51Z 2017-03 Article http://purl.org/eprint/type/JournalArticle 1070-485X 1938-3789 http://hdl.handle.net/1721.1/110058 Araya-Polo, Mauricio, Taylor Dahlke, Charlie Frogner, Chiyuan Zhang, Tomaso Poggio, and Detlef Hohl. “Automated Fault Detection Without Seismic Processing.” The Leading Edge 36, no. 3 (March 2017): 208–214 © 2017 Society of Exploration Geophysicists https://orcid.org/0000-0001-8467-1888 https://orcid.org/0000-0002-3944-0455 en_US http://dx.doi.org/10.1190/tle36030208.1 The Leading Edge Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Society of Exploration Geophysicists Society of Exploration Geophysicists |
| spellingShingle | Araya-Polo, Mauricio Dahlke, Taylor Frogner, Charlie Hohl, Detlef Zhang, Chiyuan Poggio, Tomaso A Automated fault detection without seismic processing |
| title | Automated fault detection without seismic processing |
| title_full | Automated fault detection without seismic processing |
| title_fullStr | Automated fault detection without seismic processing |
| title_full_unstemmed | Automated fault detection without seismic processing |
| title_short | Automated fault detection without seismic processing |
| title_sort | automated fault detection without seismic processing |
| url | http://hdl.handle.net/1721.1/110058 https://orcid.org/0000-0001-8467-1888 https://orcid.org/0000-0002-3944-0455 |
| work_keys_str_mv | AT arayapolomauricio automatedfaultdetectionwithoutseismicprocessing AT dahlketaylor automatedfaultdetectionwithoutseismicprocessing AT frognercharlie automatedfaultdetectionwithoutseismicprocessing AT hohldetlef automatedfaultdetectionwithoutseismicprocessing AT zhangchiyuan automatedfaultdetectionwithoutseismicprocessing AT poggiotomasoa automatedfaultdetectionwithoutseismicprocessing |