A water wheel is a rotating wheel fitted with paddles or buckets around its rim that converts the kinetic and gravitational energy of moving water into rotational shaft power. It solves the problem of getting steady, low-RPM mechanical work out of a stream without burning fuel or running an engine. Water enters at the rim — either by falling into buckets at the top (overshot) or pushing paddles at the bottom (undershot) — and the resulting torque drives a shaft for grinding, sawing, pumping or generating. A well-built overshot wheel reaches 60-75% hydraulic efficiency, which is why working examples like the 22 m Laxey Wheel on the Isle of Man have run reliably since 1854.
Water Wheel Interactive Calculator
Vary water mass, wheel radius, efficiency, and RPM to see overshot wheel torque and shaft power.
Equation Used
The article gives the overshot water wheel torque as tau = m g r, where water mass in the buckets creates a weight imbalance at the wheel radius. This calculator also applies hydraulic efficiency to estimate useful shaft torque and multiplies by angular speed to estimate shaft power.
- Overshot wheel modeled as an equivalent net water mass acting at the wheel radius.
- Gravity is fixed at 9.81 m/s^2.
- Efficiency scales gross water torque to useful shaft torque.
- Does not include bearing drag, splash losses, or bucket spill timing.
How the Water Wheel (form 1) Works
A water wheel works on one of two physical principles, sometimes both at once. An overshot wheel uses gravity — water enters buckets near the top of the wheel via a flume, and the weight of the filled buckets on the descending side rotates the shaft. An undershot wheel uses kinetic energy — moving water strikes flat paddles at the bottom and pushes them along. A breastshot wheel sits between the two, with water entering at axle height and using a mix of weight and momentum. The choice depends almost entirely on the head you have available. Less than 1.5 m of head, you build undershot. 1.5 to 4 m, breastshot. Above 4 m and you go overshot every time, because gravity gives you the highest efficiency.
The geometry matters more than people expect. On an overshot wheel the buckets must hold water through the descending arc — too shallow and they spill before bottom-dead-centre, too deep and they trap water past the dump point and fight the rotation. The bucket entry angle is typically set so the incoming jet hits the inside of the bucket lip without splashing, which means the flume exit velocity must roughly match the rim speed. Get that wrong and you lose 15-20% of available energy to turbulence at the entry. Tailrace clearance is the other killer — if the bottom of the wheel sits in backed-up water, you're dragging the lower buckets through their own discharge and the wheel slows to a crawl. Most working millwrights leave 100-150 mm of air gap between the bucket bottom and the tailrace water level.
Failure modes are mechanical and predictable. The shaft trunnions wear out of round if the bearings aren't lubricated — old mills used tallow-soaked oak journals, modern ones use bronze bushings. Spoke joints loosen as the wheel goes through wet-dry cycles and the timber swells, so traditional wheels use wedged mortise joints that self-tighten. Ice is the silent killer in cold climates — a stalled wheel exposed to freezing water builds up bucket ice that can split the rim overnight.
Key Components
- Rim and Shroud: The outer ring that carries the buckets or paddles. On an overshot wheel the shroud is a closed sidewall that retains water inside the buckets through the descending arc. Rim runout must stay within about 5 mm on a 4 m wheel or the bucket entries become uneven and you get slap-and-spill on the loading side.
- Buckets or Paddles: Buckets are the curved compartments on overshot and breastshot wheels that hold water by weight. Paddles are the flat boards on undershot wheels that catch flow. Bucket count is usually 36-48 on a wheel up to 5 m diameter — fewer and you get pulsing torque, more and the buckets run too shallow to fill cleanly.
- Spokes (Arms): Tension and compression members linking the hub to the rim. Traditional builds use 6-8 oak arms with wedged joints; modern microhydro builds use steel rod or bolted angle iron. Spoke alignment must keep the rim plane within 10 mm of true to avoid wobble at the bearings.
- Hub and Shaft: The central axle that takes the output torque off the wheel. Cast iron or forged steel for any serious application. Shaft diameter scales with wheel diameter and load — a 4 m grain mill wheel typically uses a 100-130 mm shaft running in bronze or oak journal bearings.
- Flume or Sluice: The channel that delivers water to the wheel. On an overshot the flume terminates above the wheel with a controlled jet matching rim speed. The sluice gate sets flow rate and acts as the off switch — drop the gate and the wheel stops within a minute or two.
- Tailrace: The discharge channel below the wheel. Must stay 100-150 mm clear of the lowest bucket point on an overshot, otherwise back-pressure drag drops efficiency hard. On undershot wheels, tailrace depth determines how much paddle stays submerged on the discharge side and directly affects net torque.
- Pit Wheel and Gear Train: The first stage of speed multiplication. A working water wheel turns at 4-12 RPM, far too slow for grinding stones or generators, so a large pit wheel meshes with a smaller pinion (the wallower) to step the speed up by a factor of 5-10x at the first stage.
Who Uses the Water Wheel (form 1)
Water wheels are not just heritage — they still earn their keep on small streams where the head and flow are too modest for a turbine and the budget rules out grid-tie electronics. Anywhere you have steady year-round flow under 4 m of head and you need shaft work rather than high-voltage electricity, a water wheel is often the cheapest mechanism that will last a century. The selection comes down to head and flow: low head and high flow favours undershot, moderate head favours breastshot, and high head with low flow favours overshot.
- Grain milling: The Daniels Mill in Gloucestershire, England has used an 8 m overshot wheel since 1648 to grind flour, currently still operating commercially.
- Industrial pumping: The Laxey Wheel (Lady Isabella) on the Isle of Man, built 1854, used a 22 m overshot wheel to pump water out of the Great Laxey lead mines at 1,140 litres per minute.
- Microhydro generation: The Settle Hydro scheme in North Yorkshire uses an Archimedean-screw-style wheel to feed 50 kW into the grid from the River Ribble at 1.8 m head.
- Sawmilling: Sturbridge Village's restored 1830s sash sawmill in Massachusetts runs a breastshot wheel to drive a vertical reciprocating saw at roughly 60 strokes per minute.
- Textile processing: The Quarry Bank Mill at Styal, Cheshire, uses an 8.4 m breastshot wheel, restored 1985, to drive cotton spinning machinery at the National Trust site.
- Cider and oil pressing: Traditional French moulin à huile installations in Provence used undershot wheels to drive walnut-oil edge runners through the 19th and 20th centuries.
The Formula Behind the Water Wheel (form 1)
The useful shaft power from a water wheel comes from the product of flow rate, head, water density, gravity and overall efficiency. The formula matters because it tells you immediately whether your stream can do the job. At the low end of typical microhydro flows — say 30 litres per second on 2 m of head — you're looking at a few hundred watts at best. At the nominal middle of the range used by historical mills, around 100 L/s and 4 m head, you get into the kilowatts. At the high end, a major industrial wheel like Laxey at 22 m head and high flow approaches 200 kW. The sweet spot for a backyard or small-farm install sits around 50-150 L/s at 2-4 m head — enough power to do real work, low enough flow that the civils stay affordable.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| P | Useful shaft power output | W | hp |
| ρ | Density of water (≈ 1000 kg/m³) | kg/m³ | lb/ft³ |
| g | Gravitational acceleration (9.81 m/s²) | m/s² | ft/s² |
| Q | Volumetric flow rate of water onto the wheel | m³/s | ft³/s (cfs) |
| H | Effective head — vertical drop from flume to tailrace | m | ft |
| η | Overall hydraulic efficiency (0.35 undershot, 0.55 breastshot, 0.70 overshot typical) | dimensionless | dimensionless |
Worked Example: Water Wheel (form 1) in a small craft brewery in Vermont
You are sizing an overshot Water Wheel for a small craft brewery in Vermont that wants to mechanically drive a malt mill and a recirculation pump from a hillside spring-fed stream. Surveyed head from flume to tailrace is 3.5 m, measured flow at the proposed sluice is 80 L/s in summer (the lean season), and the wheel will run at an overall efficiency of 0.70 because it's a properly built closed-bucket overshot.
Given
- ρ = 1000 kg/m³
- g = 9.81 m/s²
- Q = 0.080 m³/s
- H = 3.5 m
- η = 0.70 dimensionless
Solution
Step 1 — compute the gross hydraulic power available in the flow at nominal summer conditions, before efficiency losses:
Step 2 — apply the overshot efficiency factor to get useful shaft power at the nominal operating point:
That's enough to run a 2 hp malt mill and have a bit left for the pump. Now check the operating range. In spring snowmelt the same stream will hit roughly 200 L/s — but you can't use all of it because the bucket fill capacity caps out. Sized for 80 L/s, a wheel will only swallow about 100 L/s before the buckets overflow on entry and the surplus spills into the bypass. So the practical high end is:
At the low end — late summer drought — flow can drop to 30 L/s. Efficiency also drops because the buckets only fill about a third full and tip out early:
So your operating range is roughly 0.6 kW in drought to 2.4 kW in spring, with 1.9 kW as your design-point summer number. The malt mill draws about 1.2 kW under load, so you stay above the threshold for 9-10 months a year — which is the realistic sweet spot for this site.
Result
Useful shaft power at the nominal 80 L/s summer flow is approximately 1. 9 kW, enough to drive the malt mill with margin to spare. The full operating range runs from 0.57 kW in drought (when the buckets only partly fill and efficiency collapses to around 55%) up to 2.4 kW in spring snowmelt (capped by bucket swallow capacity, not by stream flow), with the 1.9 kW summer point sitting in the sweet spot where efficiency is highest and the malt mill stays above its load threshold. If you measure significantly less than 1.9 kW at the shaft, check three things in order: (1) flume jet velocity mismatched to rim speed — if the jet enters faster than the bucket lip is moving, you splash 15-20% of the energy at the entry; (2) tailrace water level too high — bucket drag in the discharge pool can knock 25% off output overnight after a rainstorm; (3) bucket profile worn or misaligned, which causes early spill and is visible as a continuous water curtain on the descending side instead of clean water-in-bucket transport.
Water Wheel (form 1) vs Alternatives
A water wheel is not your only option for converting stream energy to shaft work. Once you know the head and flow you have, the choice between a wheel, a low-head turbine and an Archimedean screw comes down to efficiency, capital cost, fish-passage rules, and how much speed multiplication your downstream load needs.
| Property | Water Wheel (overshot) | Crossflow Turbine | Archimedean Screw |
|---|---|---|---|
| Operating head range | 1.5-15 m | 2-200 m | 1-10 m |
| Peak hydraulic efficiency | 60-75% | 80-85% | 75-85% |
| Output shaft RPM (direct) | 4-12 RPM | 300-1500 RPM | 20-50 RPM |
| Capital cost (small site, USD) | $8k-$30k DIY-buildable | $15k-$60k off-the-shelf | $40k-$150k installed |
| Fish-friendly passage | Yes (slow, open) | No (high RPM, blades) | Yes (slow, smooth) |
| Typical service life | 80-150 years (timber 30-50) | 25-40 years | 30-50 years |
| Best application fit | Mechanical shaft work, heritage sites | Grid-tie electricity generation | Low-head sites with debris and fish concerns |
Frequently Asked Questions About Water Wheel (form 1)
Two causes you can check in 10 minutes. First, the tailrace is almost certainly running deeper in winter because of higher water-table base flow downstream — even if the flow you're delivering to the wheel hasn't changed, the discharge pool is higher and your lowest buckets are dragging through it. You want at least 100 mm of clear air between the bottom bucket and the tailrace water surface; if you can't see that gap, dredge or extend the tailrace.
Second, in cold conditions ice forms on the inside of the buckets between rotations and reduces effective bucket volume. A wheel running 24/7 stays warm enough from the water itself; one that's been stopped overnight in sub-zero air will need a few minutes of slow free-running to clear bucket ice before you load the shaft.
3 m sits in the overlap zone, so the deciding factor is wheel diameter and site geometry, not the head number itself. An overshot wheel needs to be slightly smaller than the head — figure on roughly 0.8 × H — so 3 m of head gives you a 2.4 m wheel maximum. A breastshot can be larger than the head because water enters at axle height and uses both kinetic and gravitational energy, so on the same site a breastshot can be 3.5-4 m diameter.
The bigger wheel gives you more torque per litre, which matters if your end load is grain milling or anything else that needs torque rather than RPM. If your site has the vertical room for an overshot but you can't fit a wider wheel pit, go overshot. If you have lateral room and want maximum torque, go breastshot.
The formula gives you useful hydraulic power on the wheel itself. By the time it reaches a usable output shaft, you've lost more to bearing friction, gear-train losses and shaft alignment. A traditional wooden journal bearing eats 5-8% on its own. A pit wheel and wallower gear stage takes another 5-10%. If you've got a belt drive after that to a generator or mill, knock off another 5-7%.
So a wheel rated 2 kW hydraulic typically delivers 1.4-1.6 kW at the working tool. The fix is bronze or roller bearings instead of timber journals (recovers ~5%), and tight gear-mesh setup with proper backlash (recovers another 3-5%). Don't chase the missing power by oversizing the wheel — fix the drivetrain first.
Probably yes, if your head allows it. Undershot efficiency depends on water velocity squared, so when summer flow drops by half, your power drops by roughly four times — not by half. That's the brutal scaling that catches people out. An overshot wheel scales linearly with flow because it's gravity-driven, so the same flow drop only halves your power.
If you have at least 2 m of head, build a flume to deliver the water over the top of a new wheel and convert to overshot. You'll keep useful output through dry months that completely shut down an undershot installation.
The rule of thumb traditional millwrights use is one bucket per 250-300 mm of rim circumference, which on a 3 m wheel (circumference ~9.4 m) gives you 32-38 buckets. Below that count and you get torque pulses you can feel through the gear train as the loaded buckets cross top-dead-centre one at a time. Above that count and the buckets become too shallow front-to-back to fill cleanly from the flume jet.
The bucket depth (radial dimension) typically sits around 200-250 mm on a wheel this size, with the inner shroud carrying water until roughly the 4-o'clock position before the dump angle releases it.
Almost certainly a flume-velocity mismatch. The old wheel was sized empirically — the flume exit velocity matched the rim speed, so water transferred into the bucket smoothly. On a new wheel you've usually got either a rim running faster than the jet (water slaps the back wall of the bucket) or slower than the jet (water overshoots and slaps the leading bucket lip). Either way, that slap is energy lost to noise and turbulence — typically 10-20% of your input.
Measure your jet velocity with a pitot or by timing a float over a known distance, then measure rim speed from RPM × π × D. Adjust the sluice gate height (which sets jet velocity) until the two match within 10%. The slap will disappear and your output will jump.
Not practically. A water wheel turns at 4-12 RPM. A standard 4-pole induction generator wants 1500-1800 RPM. Even a low-speed permanent-magnet generator designed for wind use typically wants 200+ RPM. So you need at least one stage of speed multiplication, usually two — pit wheel to wallower (5-8x) followed by a belt drive (3-5x) gets you into generator range.
The exception is a purpose-built direct-drive permanent-magnet alternator with a very high pole count, which can run usefully at 60-80 RPM. You'd still need a single 8-10x gear stage off the wheel shaft. Skipping the gear-up entirely and running a generator at 10 RPM produces almost no voltage and isn't a real option.
References & Further Reading
- Wikipedia contributors. Water wheel. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
