Steam Power: Energy for an Industrial Age

A stream is a good engine until you need it somewhere else

A waterwheel is cheap to run once the leat is dug, and it has been turning stones since long before this century. It sits where the fall is. A dry August starves it. You cannot pick it up and set it at the bottom of a pit that is flooding. That is the hole steam filled first. Not “power anywhere.” Power where you can haul fuel, and where the water is the problem rather than the fuel.

Early engines ate coal. At a coal pit the fuel is under your feet, so a hungry engine can still be the right tool. In Cornwall, where the mines were metal and the coal came by sea, every bushel counted. That is why a more economical engine had a market there before it had one in every town. Fuel, freight, and a man who could keep the boiler tight still decided where the machine went. Geography did not step aside.

Savery had a patent. Newcomen had a pump that worked

Savery’s 1698 patent is for raising water by fire. Steam pushed water up a pipe. Then the steam was condensed, and the vacuum drew more water up from below. There is no piston. The lift was limited, the pressures were dangerous, and a deep mine was more than it could honestly claim. The patent was written broadly enough that later builders, Newcomen included, had to work in its shadow.

Newcomen’s engine is a beam, a cylinder, and a piston. Steam fills the cylinder. A jet of cold water collapses that steam. A partial vacuum is left, and the weight of the air on the piston pushes it down. The beam rocks. The other end lifts the pump rods. Then the cycle starts again. It is called atmospheric because the air does the pushing, not a high-pressure blast of steam. It is slow, huge, and wasteful. It also kept mines open that were drowning. A drawing made in 1719 of the engine at Dudley Castle dates the machine to 1712. That drawing is what the usual date rests on. I will not pretend we have a shop log for every engine that followed.

The waste is easy to see once someone points at it. The same cylinder is blasted hot with steam and then sprayed cold, every stroke. You spend fuel reheating the metal you just chilled. For a coal owner that bill could still be worth paying. It was not a factory engine. It was a mine pump that happened to prove the cycle.

Watt kept the cylinder hot. That is the invention

In the 1760s Watt was asked to repair a model Newcomen at Glasgow. He saw where the heat was going. His answer was to condense the steam in a second vessel, kept cold, and to leave the working cylinder hot. He patented that in 1769, under a title that says what it is: a method of lessening the consumption of steam and fuel in fire-engines. Not a new kind of universe. A cheaper stroke.

Matthew Boulton’s works at Soho, and the partnership they formed, turned the patent into a business. They built engines and they licensed the right to build them. Until the patent ran out in 1800 they charged a share of the coal saved against a plain Newcomen, commonly a third of that saving. Read that twice. The product they sold was the fuel you did not burn. Cornwall bought it because coal there was a ship away. A coal pit that was already burning its own slack had less reason to pay.

John Kanefsky and John Robey counted eighteenth-century engines and buried an old claim: that Boulton and Watt built most of them, or held anything like a monopoly. The atmospheric engine, and builders who were not Soho, did the bulk of the installations. Watt’s name sits on the improvement. It does not sit on every beam in Britain.

A pump goes up and down. A mill wants a shaft that turns

A beam is a natural pump. Spinning a line of machines wants rotation. A crank and a flywheel will do that. James Pickard patented a crank-and-flywheel arrangement in 1780, and Boulton and Watt did not want to pay him. The workaround, usually credited to their man William Murdoch and patented by Watt in 1781, is the sun-and-planet gear: the piston’s rod drives a gear around a fixed gear, and the shaft turns without a crank pin of the patented kind. When that patent was no longer in the way, cranks came back. The gear is clever. It is also a way around a lawyer.

Watt’s later engines were double-acting. Steam pushed the piston one way and then the other, instead of leaving the return to the atmosphere and the weight of the pump rods. That is a smoother push into a shaft. From the shaft, belts and drums could take the turn to several machines. A line shaft running the length of a shop is one arrangement, common later, not a law of every mill. Water mills had been driving several stones from one wheel for a long time. Steam did not invent the idea of sharing a prime mover. It changed what the prime mover ate.

The factory was not a steam engine with a roof

Cotton machinery, machine tools, a wage labor force, and a market for the cloth were already an argument before a rotative engine was common. Early mills sat on rivers on purpose. Arkwright’s Cromford works, in 1771, was a water mill. Steam arrived beside that world and then, slowly, gave a builder a second choice: a site near coal, a canal, and hands, rather than only near a fall. Richard Hills’s history of the stationary engine is the book that keeps this sequence in order. David Landes’s Unbound Prometheus tells the wider European story. Read it as a big frame, not as a count of engines.

Water did not leave when Watt’s patent was sealed. Through the end of the eighteenth century a great deal of industrial turning was still done by wheels. I will not put a horsepower total on this page. The point of the counts is the direction: steam was growing, and it was not yet the whole of the power. A district with coal and a port took the engine sooner than a valley that already had a good stream and a mill race.

The governor is a feedback device. It is not a brain

If a rotative engine drives a shaft, and the machines on that shaft come on and off, the speed wanders. Too fast and you throw a belt or tear the work. Too slow and the work stalls. The centrifugal governor is a pair of weights spun by the engine. As the speed rises, the weights fly outward. A linkage follows them and throttles the steam. Speed falls, the weights drop, the steam opens again. The engine is using its own motion to correct its own motion.

Watt’s firm fitted this to rotative engines around 1788. He did not invent the device. A patent the year before had already put a centrifugal regulator on windmills, to mind the gap of the stones. What changed was the job: the weights now pinched a steam valve. It works inside a band. It can hunt, swinging past the speed you wanted and then swinging back. It does not know what the cloth should weigh, and it does not start the fire. An engineman still tends the boiler, the packing, and the water. James Clerk Maxwell’s paper “On Governors,” in 1868, is the later mathematics of how such a regulator behaves. It is not a description of the Soho floor.

Rails and hulls are a second machine, not a footnote

A locomotive and a steamship take the same fire and put it on a vehicle. They need track, or a hull, and a network of coal and water stops. They matter to the mills because they move the coal, the cotton, and the finished cloth. They are not a gift that fell out of the mine pump by itself. Trevithick’s high-pressure experiments, the locomotives of the 1810s and 1820s, and the long argument about boilers that burst are their own chapter. This page does not pretend the 1712 beam was already a railway.

High pressure is the break to notice. Newcomen and the early Watt engines used steam near atmospheric pressure and let the air, or a mild push, do the work. A locomotive needs a smaller, hotter, more dangerous boiler. The men who maintained those boilers were not optional. A burst is a shop failure, not a metaphor.

The engine moved the hard push. It hired a different crew

Somebody stoked. Somebody watched the gauge and the packing. Somebody leathered the valves, bored the cylinder, and set the beams. Inside the mill, people fed the machines, pieced the ends, and carried the product. The engine took the waterwheel’s job, or the horse’s, at the shaft. It did not take the jobs clustered around the shaft. Hours, danger, and who got paid are real, and they belong to the factory page. I will not settle them with a sentence about progress.

What the engine did add is a trade. Boilermakers, erectors, and enginemen are new skills beside the old millwright. The millwright did not vanish on the day the condenser was patented. A bad erection still shook itself apart.

Where this sits in the history of automation

A modern plant takes power from a wire, puts a motor on the machine that needs it, and uses a sensor and a controller when a quantity has to stay inside a limit. That layout is an electrical one. It arrived over decades, and it made the line shaft less necessary. It did not arrive inside Watt’s patent.

The kinship is three questions the engineman already had. Where does the push come from? How do you get it to the tool? How do you stop the speed running away? Steam answered the first with a fire you could locate, within the limits of coal. Shafts and belts answered the second. The governor answered a narrow version of the third. The break is everything the governor cannot see: the product, the fault, the next machine. I am describing equipment a plant can have now, not a rule that every shop has it.

The long use, without the straight line

Steam stayed in the mines, the mills, and the ships well after the first electric motors. It answered a limit of the waterwheel: you could put a large, steady push where the coal could be delivered, not only where the river fell. Electricity then answered a limit of the line shaft: you could put a motor where the machine was, and stop sharing one belt. Digital control answered a limit of the governor: you could measure something other than speed, and change more than a throttle. Each step showed up when somebody could build it and somebody would pay for it. None of them was hidden in the 1712 cylinder.

The engine earns its place because it was the first widely used way to turn a fire into a repeated industrial stroke, and then into a shaft. It does not earn the claim that it automated the factory or erased the map. The fire does not care what we call it. It only pushes what the boiler will stand.