AI Revolutionizes Search for Undocumented Orphaned Oil and Gas Wells

AI Tackles Environmental Challenge of Forgotten Oil Wells

Researchers at the Department of Energy's Lawrence Berkeley National Laboratory have developed an innovative approach to address a long-standing environmental issue: undocumented orphaned wells (UOWs). These forgotten relics of nearly 170 years of commercial drilling pose significant risks to the environment and contribute to climate change 1

1

2

3

4

The Hidden Danger of Undocumented Orphaned Wells

UOWs, which are not listed in formal records and have no known operators, can leak oil and chemicals into water sources, release toxic substances into the air, and emit methane—a potent greenhouse gas. With estimates ranging from 130,000 to 740,000 unplugged old wells scattered across the United States, the scale of the problem is immense 2

2

Merging Modern AI with Historical Data

To tackle this challenge, the Berkeley Lab team, led by postdoctoral fellow Fabio Ciulla, has combined artificial intelligence with historical topographic maps. The United States Geological Survey has digitized 190,000 historical maps dating from 1884 to 2006, providing a rich dataset for analysis 1

1

3

4

Training AI to Identify Well Symbols

The researchers trained an AI model to recognize the hollow black circle symbol used to denote oil and gas wells on maps between 1947 and 1992. This process involved:

1. Manually marking wells on nearly 100 California maps to create a training set
2. Teaching the AI to distinguish well symbols from similar circular patterns
3. Applying the algorithm to geotagged maps to compare identified wells with known records 1

   1

   3

   3

   4

   4

AI's Needle-in-a-Haystack Discovery

The AI algorithm scoured maps from four oil-rich counties in California and Oklahoma, identifying 1,301 potential undocumented orphaned wells. Of these, 29 have been verified using satellite imagery, and 15 more through field surveys 1

1

2

3

4

Related Stories

AI's Growing Energy Demands Spur Innovation in Sustainable Computing

Zanskar uses AI to discover hidden geothermal energy site, first industry find in decades

AI Revolutionizes Oil and Gas Industry: Faster, Cheaper Production Reported at CERAWeek

AI's Energy Paradox: How Data Centers Are Reshaping Global Power Demand and Climate Goals

Verification Process and Future Implications

Researchers employ a two-step verification process:

1. Remote analysis using satellite images and historical aerial photos
2. Field surveys using magnetometers to detect buried well casings 1

   1

   3

   3

   4

   4

The team's conservative approach likely underestimates the total number of UOWs, suggesting that refinement of their methods could uncover even more wells 1

1

3

4

Expanding the Toolset for Well Detection

Future efforts may incorporate additional technologies to enhance well detection and assessment:

• Drones equipped with magnetometers for rapid deployment in areas inaccessible to aerial detection
• Methane-sensing drones to measure air leakage
• Hyperspectral cameras to detect otherwise invisible methane plumes 2

  2

This AI-driven approach, combined with modern field technologies, represents a significant step forward in addressing the environmental and climate risks posed by undocumented orphaned wells. As the research progresses, it may pave the way for more effective policies and actions to mitigate these long-forgotten hazards of the fossil fuel industry.