A 19.8 square mile sweep of northwestern Pennsylvania turned up more than 250 likely well sites from the air, and the state has paperwork for barely a tenth of the wells it is thought to hold

A wooded hillside in the northwest corner of Pennsylvania, steep, second growth, no road within half a mile.

Somewhere under the leaf litter is a steel pipe driven into the ground more than a century ago.

Nobody alive knows it is there. No map shows it and no agency holds a file on it.

Three national laboratories and a commercial drone operator sit alongside them, which is an unusual amount of institutional weight for a survey of rural woodland.

A drone crosses overhead at low altitude carrying an instrument that measures the magnetic field.

The pipe gives itself away.

So do hundreds more.

How a rotting pipe betrays itself to something flying over it

Steel is ferromagnetic. A length of it standing vertically in the ground behaves like a weak bar magnet aligned with the earth’s own field.

That distorts the local field in a shape a magnetometer can recognize. The disturbance is small, on the order of a few nanotesla against a background field of around fifty thousand.

Flying low is what makes it readable. Magnetic anomaly strength falls off steeply with distance, so an instrument at rooftop height sees what one at survey aircraft height would miss entirely.

The signature also survives conditions that defeat everything else. It works under leaf cover, under snow, under a parking lot, and it does not care whether anything is visible at the surface.

What it cannot do is tell a well casing from a buried pipeline, a length of fence or an old car. Every anomaly is a candidate rather than a find.

Steel bends the field. Leaves hide nothing.

What the ground crews have to do afterward

The aerial pass produces a map of anomalies, and that map is the beginning of the work rather than the end of it.

Crews then walk the flagged points carrying backpack magnetometers, which one participant described as a fancy metal detector.

Confirmation means finding the casing, which often means clearing brush and digging. Only then does a vehicle with a methane analyzer come through to see whether the thing is venting.

The northwest survey covered 19.8 square miles across three counties, split as roughly 8.5, 7 and 4.3, and flagged more than 250 likely well sites.

A second phase moved to six communities just outside Pittsburgh, in Allegheny and Washington counties, where the same wells sit under yards and streets.

Flying is the cheap part. Confirming is the slow part.

The arithmetic that makes this a very long project

Pennsylvania has roughly 30,000 orphaned and abandoned wells on its books. That is the number with a file attached.

The estimated true population runs between 300,000 and 700,000, which means the documented share sits somewhere around a tenth and possibly nearer a twentieth.

Drilling here started in the 1850s, decades before anyone was required to record a well, let alone plug one when it stopped producing.

The state has about 400 million dollars in federal money for this through the end of the decade. Divided across even the documented 30,000, that is well short of what a plugging job costs.

So the survey work is producing a queue faster than the queue is being cleared, which is an uncomfortable but honest description of where this stands.

Records cover 30,000. Reality may hold twenty times that.

Who is actually doing this and what they published

The fieldwork is a coalition rather than an agency program. A university group leads the science, an environmental organization funds and coordinates, and the state regulator supplies records and access.

Three national laboratories and a commercial drone operator sit alongside them, which is an unusual amount of institutional weight for a survey of rural woodland.

The results went into a peer reviewed journal this year under a title that promises exactly what it did, a large scale drone and ground based survey.

The team, the technologies and what the authors say the method cannot replace are set out by an advocacy blog.

A different route to the same wells reads old survey sheets instead of the ground, using historical maps that predate computers.

Five institutions, one hillside, one pipe.

Why a leaking well matters more than a quiet one

Not every buried casing is an emissions problem. A well that was properly plugged with cement decades ago can sit there indefinitely doing nothing.

The ones that matter are the unplugged and badly plugged, where gas migrates up the annulus and out into the air or into a basement.

Methane is the headline risk and it is not the only one. Surveys in the state have found benzene coming out of a large share of the abandoned wells sampled.

The national scale of that leakage is laid out in this account of silent leaks nobody can locate.

The field phase in southwestern Pennsylvania and the tools used on it are described by a regional outlet.

Finding these well sites is now routine, and paying to close them is not.