Chapter 16: Water and Wind: Mills and Pre-Fuel Prime Movers

Era span: ~300 BCE → 18th c. CE · Difficulty: mid
Requires: Ch 3, Ch 14, Ch 15
Unlocks: Ch 18, Ch 22

Before fossil fuel, water and wind were major sources of non-muscular mechanical power alongside animal labour. A waterwheel's output depends on flow, usable head, wheel design, speed, losses, and maintenance; historical machines ranged from small workshop drives to substantial industrial power. Good mill sites can become fixed capital around which workshops cluster, but only after hydrology, flood risk, ecology, and local demand are assessed.

16.1 Water Power Fundamentals

Safety warning: mill ponds, dams, channels, wheel pits, and transmission systems can fail through overtopping, piping, foundation movement, debris, gate failure, drowning, or contact with rotating machinery. Dam and tailrace work requires qualified site investigation and civil/mechanical design, applicable dam-safety review, instrumentation, inspection, emergency action planning, controlled drawdown procedures, and exclusion zones. This manual does not supply universal spillway or structural dimensions.

Power available = water density × flow rate × g × head × efficiency (P = ηρQgh). Practical numbers:

Site engineering: a millpond can store part of a stream's flow for later use; a millrace/canal conveys water to the wheel; a tailrace returns it downstream. Available hydraulic power rises with ρQgh, but usable wheel output also depends on flow duration, wheel speed, losses, and downstream obligations. Extra head can raise output substantially, but not by a universal multiplier.

Key threshold: 1 m³/s falling 3 m has about 29 kW of hydraulic power before wheel and transmission losses. An overshot wheel at ~60 % and ~10 % gearing loss leaves roughly 15 kW at the line shaft — the order of two or three pairs of millstones at a few kilowatts each. Whether that supports a mill, fulling works, bloomery bellows, or paper stampers—and whether they can run together—depends on measured flow duration, peak demand, transmission, service factors, and local resource availability.

Survey doctrine: measure flow and usable head across seasons before a permanent investment. A notch or weir still needs a correct rating, stable installation, maintenance, and a safe measurement structure. Design loads, flood handling, environmental flows, downstream use, and drought operation from the local record; no universal “size to the minimum, spill the maximum” rule is safe.

16.2 Transmission from Wheel to Work

Wheel shafts turn slowly (~5–15 rpm); most work wants speed or reciprocation:

• Gear-up trains (small lantern pinion on driven shaft meshing large wheel) multiply speed at torque's expense.
• Cams and trip-hammers: lobes on a shaft lift hammer heads that drop under gravity — they full cloth, crush ore (Ch 13), and beat paper pulp.
• Crank-and-rod converts rotation to reciprocation for saws and pumps (Ch 15).
• Belt lines can distribute one water source to multiple workstations in one building—an early factory pattern, but not the only path to mechanised production.

Millwright's rules: gears meshed with consistent depth (noise means wear); crowned pulleys keep belts tracking, crossed belts reverse direction; trip-hammer helves of springy ash, cam faces greased; line-shaft hangers aligned within sighting tolerance or bearings eat themselves. Grease cups over every journal, checked each shift — friction is a tax collector who never sleeps.

16.3 Windmills

Where streams fail, wind serves — with caveats: it is intermittent, stormy, and needs orientation.

• Post mill: whole body pivots on a post to face the wind; simple and commonly small, with output set by sail area, wind, and duty.
• Smock/tower mill: rotating cap only—larger sails, heavier machinery, and usually greater output. Historical power figures vary widely.
• Sail physics: canvas-on-lattice sails, adjustable reefing; spring sails later automate gust response. Tip-speed ratios stay modest; torque is what you get.
• Dutch drainage duty: scoop wheels lift polder water — land reclamation as windmill economics.

Wind suits grinding and pumping where water is absent; don't run precision industry on gusts until storage exists.

Storm doctrine: reef early (canvas shortened before the gust, not during), tail-pole or fantail holds heading, brake locks rotor for maintenance only — never trust the brake in a gale; turn out of wind instead. Storm overspeed and the fires it starts in dry timber wrecked many historical mills.

16.4 What Mills Make Possible

Priority deployment order for rebuilding civilization:

1. Grain milling — can replace substantial hand-grinding labour and produce more consistent meal; the social effect depends on labour organisation, access, and food demand.
2. Sawmills — powered saws can increase throughput and standardise dimensions for construction (Ch 6) and shipbuilding; site-specific gains can be large but are not a universal multiplier.
3. Fulling and spinning machinery — increases textile throughput (Ch 3); it was an important early mechanised industry, though not the single trigger of every Industrial Revolution.
4. Ironworks air blast and hammers — can support furnace and forge operations (Ch 14); yield still depends on charge, furnace design, and skill.
5. Ore crushing, paper pounding, and oil pressing — batch-processing industries can cluster where power and transport are economical.
6. Mine drainage — water-driven pumps can extend workable depth while increasing both capacity and the consequences of failure (Ch 13).

Mill-site map = proto-industrial geography. Record flow gauges seasonally; drought years decide which sites deserve capital upgrades.

16.5 Limits — and Why Steam Became Important

Water binds production to particular catchments and seasons; wind binds it to weather and site. Both can deliver rotary power but cannot be carried conveniently to every ship, field, or workshop. Burning fuel created a more portable rotary-power option (Ch 23). Where water and wind are suitable, mills remain useful; historically, millwright and machine-shop experience helped train engineers who later developed steam, but that transfer was not automatic or universal.

Dead end avoided: over-investing in exotic perpetual-motion schemes to "beat" intermittency — history burned real talent there. Intermittency is managed by storage ponds, mixed sources, and task scheduling (grind when water flows, saw when wind holds), not defeated by cleverness.

Scheduling doctrine (pre-steam grid): millpond gates meter day power; heavy hammering at night flow; sawing to windy days; grinding to calm water days. The miller is a dispatcher — log flows, post the week's run order, never promise all machines at once.

16.6 The Mill Record

• Barbegal (Gaul, 2nd c. CE): sixteen overshot wheels stacked in two parallel flights down a hillside channel — a Roman industrial flour complex; capacity estimates range from a few tonnes of flour daily upward (scholars argue), but nobody disputes the plant's industrial character.
• Rome's Janiculum Hill hosted state mills; when Gothic sieges cut the aqueducts (537 CE), Belisarius installed floating mills on the Tiber — power infrastructure defended like the walls it fed.
• Domesday Book (1086) records ~5,624 watermills across surveyed England — roughly one for every 50 recorded households — on the order of one mill per 300–400 people in the total population; milling had saturated rural geography three centuries before anyone wrote a treatise about it.
• Wind arrives late and eastward: vertical-axis Persian panemone mills recorded by 10th-century geographers in Sistan; horizontal-sail postmills appear in northwestern Europe by the late 12th century; the great Dutch drainage program (polder pumping, §16.3) turned mill engineering into national survival infrastructure.
• Historiographical flag: Eleanora Carus-Wilson's 1941 essay called fulling-mill diffusion "an Industrial Revolution of the thirteenth century" — wool processing fleeing towns for cheap rural waterpower, reorganizing regional economies. Later scholars moderated the thesis (transport costs, guild politics mattered too), but the pattern — power sites restructuring industry's map — stands and repeats with steam (Ch 23) and electricity (Ch 26).

16.7 Choosing Sites Like Cities

Record water rights, usable head, flow duration, flood history, environmental obligations, road or wharf access, materials, hinterland demand, and safety. Mark each condition with its evidence and uncertainty (Ch 11). Permanent construction should follow competent site investigation and risk review; no voting-style “seven of eight” rule is a safe engineering criterion.