In 1712, Thomas Newcomen installs the first operational steam engine in a coal mine at Dudley Castle, Staffordshire. The engine pumps water flooding the galleries. Its thermal efficiency is around 1%. It consumes quantities of coal that only immediate proximity to the mine makes bearable. Newcomen's engine functions because fuel is free. It cannot function elsewhere.
For sixty years, the steam engine is bound to geology. Manufactures requiring mechanical power locate near rivers for water wheels or near mines for Newcomen engines. The industrial geography of eighteenth-century England is a map of waterways and deposits.
In 1769, James Watt patents the separate condenser. Instead of cooling the cylinder each cycle to condense steam, steam is condensed in a separate chamber maintained permanently cold. The cylinder remains hot. Energy lost reheating it each cycle disappears.
Efficiency is multiplied by roughly three. Two thirds of coal that was burned for nothing is burned no longer.
This gain does not reduce total coal consumption. It increases it. Jevons (1865) formalizes this: increased efficiency makes steam usage viable in places and applications that did not consume it before. The condenser does not reduce coal demand. It extends it geographically.
The factory as organizational form existed before Watt. Arkwright opens Cromford in 1771, a modern water-powered spinning mill: factory discipline, continuous rotation, division of labor. The factory system is an organizational invention, not thermodynamic. What the condenser does is not invent this form. It is to detach it from the river.
The Bridgewater Canal opens in 1761, eight years before Watt's patent. It transports coal from Worsley to Manchester, some ten kilometers away. Infrastructure already existed. The condenser makes it profitable at a new scale. Manchester offers what mine and river do not offer together: humid climate favorable to cotton, available workforce, port access to Liverpool. The cotton industry settles there because Watt's efficiency makes viable a location chosen according to criteria other than energy.
Newcomen bound industry to geology. Watt partially detached it. This detachment has a residual cost: coal transport became an industry in itself. Canals then railways were built or made profitable to transport fuel toward machines that no longer needed to be beside it. British transport infrastructure of the nineteenth century is the logistical trace of this relocation.
Doctrine
A thermodynamic efficiency gain does not reduce consumption. It displaces the geographical constraints of production. What Watt liberated was not energy. It was territory. And each liberation is also a new attachment, displaced.
This displacement is not automatic. It requires three conditions: that the energy vector be transportable at low cost, that final conversion impose no irreducible local constraints, and that demand exist in places detached from the source. When one of the three is missing, the efficiency gain remains local.
Open vector
The internal combustion engine detached agriculture from animal traction. Wrigley estimates that in nineteenth-century England, a quarter of cultivated land served to feed draft horses. The tractor did not merely replace the horse. It rendered its lands available. The efficiency gain liberated territory in the literal sense.
Photovoltaics progressively detaches electricity production from hydrocarbon geology, but reattaches it to another geology: lithium, copper, rare earths, quality silicon. Liberation is displacement, not abolition.
Nuclear fission confirms that energy density gain does not automatically entail geographical restructuring. France could build its electrical capacity according to cooling water availability rather than according to the coal map, and naval nuclear propulsion detached submarines from fuel logistics. But restructuring was lesser than what energy density suggested, because conversion and distribution constraints persisted. Nuclear liberated from coal geology without liberating from water geography.
If fusion energy becomes exploitable one day, which of the three conditions will lift the constraint, and which will maintain it?
