Eternal Flame Falls: The Fire That Should Not Burn
The trail begins in unremarkable woods. A little south of Buffalo, New York, a footpath drops through the hemlocks of Chestnut Ridge Park, past the ferns and the wet black rock, and follows Shale Creek down toward a low cascade in a preserve that borrows the creek's name. It is not a grand waterfall. Barely thirty feet high, a modest sheet of water folding over a ledge of dark shale, it is the kind of place a hiker might pass without a second glance. Except for one thing. Push in close, crouch at the base, and peer into the shallow grotto that the water has scooped out of the rock behind the falling curtain, and you find something that has no business being there: a small golden flame, roughly the height of a hand, burning steadily inside the waterfall.
Mist beads on the stone around it. Spray runs down the rock and drips from the ledge overhead. The flame leans in the draught, recovers, and goes on burning in what may be the least hospitable spot on earth for a fire to keep going: wet, cold, drenched, walled in on every side by falling water. People have come to look at it for generations. They call it the Eternal Flame, and the postcards call it one of the most beautiful of all the world's burning springs.
Part of the story is not mysterious at all. Natural gas rises through fractures in the shale and pools in the little grotto, and at some forgotten moment a human hand set it alight. The flame is not even literally eternal. Wind and heavy spray snuff it out with some regularity, and it survives because hikers have made a quiet ritual of crouching down with a lighter to bring it back to life. So many of them now make the trip that in 2023 the county rebuilt the steep approach with 139 box steps and a long run of railing to handle the traffic. The fire, in other words, is ordinary combustion, tended by strangers.
The deeper question sits behind the spectacle: where does the gas come from, and how is a young, cold rock producing it at all? For most of the flame's known life that question simply sat there, unasked. Then, in 2013, it got its first proper interrogation.
That year the Italian geochemist Giuseppe Etiope, of the National Institute of Geophysics and Volcanology, joined forces with Arndt Schimmelmann and Agnieszka Drobniak of Indiana University to sample gas seeps across the northern Appalachian Basin. Their paper appeared in the journal Marine and Petroleum Geology under a flatly descriptive title: Natural seepage of shale gas and the origin of eternal flames in the Northern Appalachian Basin, USA. The framing question was almost bureaucratic. How much methane leaks out of natural seeps into the atmosphere is a number that matters for climate accounting, and the team set out to measure it. At the falls they estimated that the main seep releases on the order of a kilogram of methane a day, alongside a scatter of smaller micro-seeps in the surrounding rock. It was a tidy field study, the kind that fills the back pages of a specialist journal. Then the gas itself refused to behave.
Conventional petroleum geology teaches a strict recipe. Gas like this is thermogenic, which means an organic-rich source rock has to be buried deep and cooked, typically somewhere near 100 degrees Celsius, over immense spans of geological time before it matures into hydrocarbons. Temperature and time, time and temperature: that is the whole recipe, and it has accounted for gas fields the world over for a century.
But when the team compared the seep gas with gas drawn from wells across the region and traced its chemistry back to a source, the source turned out to be the Rhinestreet Shale, lying only about 400 meters down. Shallow. And by every reasonable estimate far too cool to have cooked anything at all. Schimmelmann put it without ornament: the source rock, he noted, is not very warm.
Then the composition twisted the knife. Most gas seeps in the world are almost pure methane, the lightest and simplest hydrocarbon. This one carried roughly 35 percent ethane and propane, heavier and wetter hydrocarbons, and the authors reported that this was possibly the highest such proportion ever measured at any natural gas seep on the planet. It is the wrong signature entirely. Wet, heavy gas in that abundance reads like the fingerprint of a mature, high-temperature reservoir, the sort of thing expected only from rock that has been deeply buried and thoroughly cooked. Here it was leaking out of a shallow, cool, comparatively young shale that, by the textbook, should not have been capable of making it.
The finding drew a flurry of press at the time, and the excitement was not really about a burning waterfall. If a cool, shallow shale can generate wet thermogenic gas, then the textbook map of where such gas can be found, and how deep a drill has to go to reach it, is missing something. Journalists reached for the phrase a potential new source of natural gas, and the researchers themselves suggested that the seep might point to a previously unrecognized geologic mechanism. The little flame was not just a curiosity. It was a small, bright anomaly poking a hole in a very large and very profitable body of received wisdom.
Conclusions and Open Questions
The flame itself is not the puzzle. Ordinary combustion, ordinary methane, a lighter in a stranger's hand. The puzzle is the rock, and there is only one serious attempt in the literature to explain it.
Etiope's team put forward an answer and were careful to label it a hypothesis rather than a finding. One theory holds that the gas is forming at unexpectedly low temperatures, with certain minerals in the shale acting as natural catalysts, chemical matchmakers that crack organic matter into hydrocarbons without the usual furnace heat. The strength of the idea is that it fits the evidence in front of them: it accounts for wet, heavy gas leaking out of a cold rock, and Etiope had argued for low-temperature catalytic gas generation in other settings across his career, so it was not conjured from nothing. Its weakness is equally plain, and the authors did not hide from it: no one has isolated the specific catalyst, reproduced the reaction in a laboratory, or caught it operating inside the Rhinestreet Shale. The hypothesis explains the data by proposing a mechanism nobody has yet observed in the act.
The conventional model has the mirror-image profile. Its strength is a century of success across the world's gas fields. Its weakness is that, applied to this spot, it fails the thermometer outright. To rescue it, some argue that the gas is not being manufactured at 400 meters at all, but migrating upward or sideways along fractures from a deeper, hotter, more mature stretch of the basin, and merely passing through the shallow Rhinestreet shale on its way to the surface. That keeps the textbook intact. It also runs straight into the team's own chemical tracing, which tied the gas to the Rhinestreet rock rather than to a deeper reservoir, and the plumbing beneath the falls has never been mapped deeply enough to settle the argument either way.
A third possibility is microbial gas, manufactured by living organisms in the shallow subsurface, which happens elsewhere and would sidestep the temperature problem completely. The heavy, thermogenic-looking mixture argues strongly against it: microbes overwhelmingly produce almost pure methane, not the ethane-and-propane-rich gas found here. The very feature that makes the seep strange, its wetness, is what rules out the tidiest low-temperature explanation available.
There is also a caveat about the headline number. The claim that this is possibly the highest ethane-and-propane proportion ever recorded at a natural seep rests on a comparison against the world catalogue of measured seeps, and that catalogue is patchy, unevenly sampled and heavily skewed toward places industry had reason to study. The measurement itself is real and repeatable. Its ranking as a world record is a quieter and shakier claim, because it is really a claim about everywhere else, and everywhere else has not been measured.
What remains unexplained is the mechanism, and the state of the evidence is thin. A single detailed study, now more than a decade old, framed the puzzle with care. It has not been overturned, but neither has it been closed, extended or nailed down by the sustained follow-up that a record-breaking measurement ought to attract. The catalyst remains hypothetical. The migration path remains unmapped.
The open questions are simple to state and hard to answer. How does a rock too cool to mature produce not merely gas but wet, heavy, thermogenic-looking gas? Why here, in this quiet creek bed, and in a proportion apparently without equal anywhere else? If the familiar recipe of temperature and time is not the one being followed, what recipe is? And how many other ordinary-looking seeps, in how many other quiet places, are breaking the same rule unnoticed, because nobody has yet knelt at the water's edge and thought to ask the rock what it is doing?