AI discovers thousands of abandoned oil wells hidden in old maps

By searching old maps, an Artificial Intelligence application found thousands of abandoned oil wells — which makes it possible to eliminate methane emissions from these silent polluters that no one notices, authentic environmental time bombs.

Abandoned oil wells pose a significant risk to the environment and health as they can release methane — a potent greenhouse gas — into the atmosphere and contaminating groundwater.

The estimated total number of undocumented orphan wells — which are not formally owned or documented — is between 310,000 and 800,000.

These wells are often “out of sight and out of mind“, which makes them a great danger, which is almost impossible to locate.

But, according to a new report, published this Wednesday in the magazine Environmental Science & Technology, AI can help.

During the course of the study, a state-of-the-art Machine Learning program was trained to recognize oil well symbols and gas in historical topographic maps spanning decades.

Using this approach, researchers have identified more than 1000 wells undocumented and later confirmed their presence within a radius of 10 meters.

To validate the findings, researchers used recent satellite images and field surveys. Magnetic field detectors identified the presence of buried metal pipes, confirming the existence of wells in previously undocumented locations.

Zombie wells

For decades, oil and gas wells have marked the North American landscape — after all, the USA was one of the first countries to have a large-scale oil industry, says .

But this industry’s least visible legacy It’s an environmental disaster. These wells are silent polluters that no one notices, environmental time bombs.

Methane emissions from abandoned wells (also known as “zombie wells”) contribute significantly to climate change and, in addition, these wells can release brine and hydrocarbons, threatening aquifers and ecosystems.

Plugging these abandoned wells would cost billions of dollars. But before covering them, you need to find them.

At some point, the location of these wells was known: were marked on some maps. But abandoned wells have gotten lost in the shuffle and are not mentioned in any database — they simply exist as dots on old maps.

But there are many old maps. Since 2011, the United States Geological Survey (USGS) has uploaded USGS histories taken between 1884 and 2006.

“This problem is equivalent to find a needle in a haystackas we are trying to find some unknown wells that are scattered among many more documented wells”, he said Charuleka Varadharajana Berkeley Lab scientist and lead author of the study.

The process used by the study authors involves two steps.

In a first step, AI analyzes georeferenced historical mapsusing a neural network trained to recognize well symbols. These symbols represent wells marked by cartographers on maps created decades ago.

Then, the detected wells are crossed with databases state and federal documented wells.

If a particular well that is detected by the AI ​​is more than 100 meters away from any well that is documented, it is marked as a potential zombie oil well — which is then marked for deletion.

Generally speaking, the method was successful and researchers are confident the method can be used to find more orphan wells.

“With our method, we were conservative in relation to what would be considered a potential undocumented orphan well,” said Varadharajan.

“We intentionally chose to have more false negatives than false positives, as we wanted to be careful with the individual locations of the wells identified through our approach”, concludes the researcher.