A Deep Learning Based Framework to Identify Undocumented Orphaned Oil and Gas Wells from Historical Maps: A Case Study for California and Oklahoma
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.
View abstract on PubMed
Summary
This summary is machine-generated.Researchers developed a computer vision model to find undocumented orphaned wells (UOWs). This method accurately located over 1300 potential UOWs, aiding environmental protection efforts.
Area Of Science
- Environmental Science
- Geospatial Analysis
- Artificial Intelligence
Background
- Undocumented Orphaned Wells (UOWs) pose environmental risks, including methane leaks and groundwater contamination.
- Estimates suggest 310,000 to 800,000 UOWs exist in the US, with largely unknown locations.
Purpose Of The Study
- To develop and validate a novel framework for precisely identifying the locations of potential UOWs.
- To create a scalable solution for UOW detection across the United States.
Main Methods
- Utilized a U-Net computer vision neural network model trained on historical topographic maps to detect oil and gas well symbols.
- Identified potential UOWs as symbols located more than 100 meters from documented wells.
- Developed a custom tool for rapid validation and confirmed findings using satellite imagery and magnetic surveys.
Main Results
- Discovered 1301 potential UOWs across >40,000 km² in California and Oklahoma.
- Confirmed the presence of 29 UOWs via satellite imagery and 15 via field magnetic surveys.
- Achieved spatial accuracy on the order of 10 meters for confirmed UOW locations.
Conclusions
- The developed framework effectively identifies potential UOWs using historical maps and AI.
- This scalable approach can be applied nationwide, significantly improving the detection of hazardous UOWs.
Contact us if these videos are not relevant.
Contact us if these videos are not relevant.