#850 1835 · Zigong salt-well drillers, Sichuan (自贡盐商) · Salt production (brine mining)
Sichuan's salt boilers were running out of wood, so drillers went deep enough for their own wells to start burning their own fuel
the problem
Turning brine into salt requires continuous boiling, and centuries of drilling deep inland in Sichuan had stripped the region of the firewood that boiling depended on, driving fuel costs up as fast as the well count grew
background
Unlike coastal salt makers who could rely on sun to evaporate seawater, Zigong's brine came from wells drilled far inland, so every drop had to be boiled off over a fire, and Sichuan's salt industry had been consuming wood at industrial scale for two thousand years by the time forests within reach of the wellheads began to run out. The obvious response was to economize on the fuel itself: spreading brine on tree branches in the sun to pre-concentrate it before boiling, and routing several boiling pans off a single chimney to reuse residual heat. Both bought time, neither solved the underlying problem, and the cost of wood kept climbing as haulage distances grew.
Drillers pushing their percussion wells deeper in search of richer brine had been striking pockets of flammable gas in the rock below the brine layer since at least the 16th century — at first a hazard to route around or vent off, since a well only a few hundred meters deep rarely produced enough of it to matter. It took centuries of incremental advances in bamboo-cable drilling technology before wells could reliably reach the 700-800 meter Triassic formation where gas and brine occurred together in commercially useful quantity.
what everyone would do
Economize on the wood you still have to buy: pre-concentrate the brine in the sun before boiling it, share a chimney's residual heat across several pans. Both were real improvements and neither stopped the fuel bill from rising, because they still depended on an input whose local supply was shrinking every year the industry kept growing.
what they saw
The rock that held the brine, once a well went deep enough, also held pockets of combustible gas the drillers had been hitting and discarding as a hazard for centuries — the fuel shortage and the wasted byproduct were sitting in the same borehole, and nobody needed to look outside the well to solve the problem inside it.
the move
Rather than treat the gas as a hazard to manage or vent, Zigong's drillers kept extending well depth specifically to reach that gas-bearing layer, culminating in 1835 when the Shenhai well became the first well in the world to pass 1,000 meters (the deepest wells in the US at the time were roughly 500 meters). At the wellhead, a device called the Kang Pen drum, introduced in the late 18th century, controlled pressure to separate the co-mingled gas and brine coming up the same borehole into two streams, each routed into its own dedicated network of sealed bamboo pipeline running straight to the evaporation sheds — so the same hole that produced the raw brine also produced the fuel to boil it.
why it works
Wells driven past roughly 700-800 meters into the Triassic gas-bearing formation beneath the brine aquifer began striking both fluids from the same hole. The Kang Pen drum at the wellhead used controlled pressure to separate the co-mingled gas and brine into two distinct streams without requiring two separate wells, and a dedicated network of tung-oil-sealed bamboo pipeline carried the gas straight to the adjacent evaporation pans, eliminating both the purchase cost of wood and the labor of hauling it. Because the fuel supply and the raw material came from the identical wellbore, a richer or deeper well produced more of both at once — the fuel scaled automatically with exactly the resource it existed to boil, which is precisely the opposite of a purchased-wood supply that got scarcer the more the industry expanded.
the payoff
By 1821 natural gas fueled roughly a tenth of Zigong's evaporating pans, with the rest still burning wood or charcoal; by 1857 gas had overtaken wood as the more common fuel. Zigong grew into what became known as China's Salt Capital, with annual production around 150,000 tons by the 1850s; over the industry's full history it drilled an estimated 130,000 brine and gas wells and produced a cumulative 30 billion cubic meters of natural gas, transported through hundreds of kilometers of bamboo pipeline — some of it still in operation as late as the 1950s.
where it breaks
It only works where the same extraction process genuinely co-locates a usable byproduct with the primary resource — a process whose waste stream has no relation to what it needs for fuel or input has nothing to capture. It also requires reaching the depth or process stage at which the byproduct becomes available in commercially useful quantity, which for centuries of Sichuan's shallower wells simply wasn't an option; the fix only became available once drilling technology itself advanced enough to go deep enough. And the separation and pipeline infrastructure (the Kang Pen drum, hundreds of kilometers of sealed bamboo line) is a fixed investment that only pays off at the scale of an entire wellfield, not a single isolated well.
what came after
Zigong's co-production of brine and gas from a single wellbore, with dedicated separation and pipeline infrastructure to route the byproduct straight into the production process it fueled, predates the Western oil and gas industry's own gas-gathering systems by well over a century and is cited in petroleum-industry retrospectives as the earliest known example of an extraction operation engineering itself to run on its own byproduct at industrial scale.
references
- [1]Ancient Chinese DrillingCSEG Recorder (Canadian Society of Exploration Geophysicists), 2004cseg.ca
- [2]With Bamboo Pipes, China Was Extracting Gas a Thousand Years Before the First Modern WellClick Petróleo e Gás, 2025en.clickpetroleoegas.com.br