An overshot water wheel is a vertical wheel with peripheral buckets that receives water near its top via a launder or flume, using the weight of trapped water in descending buckets to turn the shaft. It remains the workhorse of heritage gristmills and restored sawmills where head exceeds wheel diameter. The mechanism converts gravitational potential energy directly into rotational shaft power, bypassing the velocity-driven losses of undershot designs. Well-built overshot wheels reach 60–75% hydraulic efficiency, which is why 19th-century millwrights chose them over any alternative when site head allowed.
Overshot Water Wheel Interactive Calculator
Vary head, flow, efficiency, and wheel speed to see shaft power, torque, and an animated overshot wheel diagram.
Equation Used
The calculator uses the overshot wheel power relationship: water power is rho*g*Q*H, then usable shaft power is reduced by efficiency eta. Wheel rpm does not change power directly, but it sets the shaft torque required to transmit that power.
- Head is net vertical head from launder lip to tailwater.
- Flow is steady and fully captured by the buckets.
- Efficiency includes bucket spill, hydraulic, and mechanical losses.
- Water density is 1000 kg/m3 and g is 9.81 m/s2.
How the Overshot Water Wheel Actually Works
Water enters the wheel through a launder — the wooden or steel trough that delivers flow from the headrace — and drops into buckets just past top dead centre. From there, it's gravity doing the work. The mass of water held in the descending buckets creates an unbalanced moment about the shaft, and the wheel rotates because one side weighs more than the other. Speed stays low, typically 4–12 RPM on a 3–5 m wheel, which is why every traditional mill ran a gear train of pit wheel and wallower to step the speed up to something useful at the millstone.
Bucket geometry is where most builds go wrong. The bucket must fill close to its rated capacity at the top of the wheel and hold that water as long as possible before dumping near the bottom. If the bucket is too shallow, water spills out before the bucket reaches the 6 o'clock position and you lose the lower half of the moment arm. Too deep, and the bucket carries water past bottom dead centre, lifting it back up — pure parasitic loss. The classic Fitz steel overshot wheel solved this with a curved bucket profile and a vent slot that lets trapped air escape so the bucket fills cleanly. A bucket fill ratio of 0.33 to 0.5 is the sweet spot.
Tailwater clearance matters more than first-time builders expect. If the tailrace level rises within 50 mm of the bottom of the wheel, the wheel runs in backwater and efficiency collapses. You'll see it as sudden RPM drop after heavy rain, and the cure is either a deeper tailrace cut or a smaller-diameter wheel. Wheel diameter sizing is set by available head — the wheel must be slightly smaller than the vertical distance from launder lip to tailwater, otherwise you can't deliver water to the top.
Key Components
- Launder (Headrace Flume): Delivers water from the millpond or penstock to the top of the wheel. The lip should sit 100–150 mm above the wheel rim with a sluice gate to control flow. Cross-section is sized so velocity at the lip is 1.5–2.5 m/s — slow enough that water drops cleanly into the bucket rather than overshooting it.
- Buckets (Floats): Curved compartments around the wheel rim that catch and hold water. Bucket count is typically 36–48 on a 4 m wheel, depth 200–350 mm. The leading edge and curve angle determine fill quality — a 30° entry angle gives clean fill at design flow without splashback.
- Shroud Plates: The two vertical side plates that close off the buckets laterally. Without tight shrouds, water spills sideways out of the buckets and you lose 15–20% of the available torque. Gap between shroud and bucket end should not exceed 2 mm.
- Sole Plate (Inner Drum): The cylindrical inner surface that forms the back wall of every bucket. On Fitz-style steel wheels this is a continuous riveted drum; on traditional timber wheels it's individual sole boards between each pair of arms.
- Arms and Hub: Radial spokes connecting the rim to the central shaft. Cast-iron hubs with wrought-iron arms were standard on industrial wheels; wooden compass arms were used on smaller mill wheels. Arm count is usually 6–12 depending on diameter.
- Main Shaft and Bearings: Horizontal shaft transmitting torque to the pit wheel. Plain bronze or lignum vitae bearings are still preferred over rolling-element bearings on heritage restorations because they tolerate the slow speeds (4–12 RPM) and occasional water exposure without brinelling.
- Sluice Gate: Vertical gate at the launder controlling flow onto the wheel. A rack-and-pinion or screw lift gives proportional control. Used to throttle the wheel during low-load periods so RPM holds steady when the millstone is disengaged.
Who Uses the Overshot Water Wheel
Overshot wheels live wherever a small stream drops more than 3 m over a workable distance. They were never the right answer for big rivers — that's Francis and Kaplan country — but for heritage mills, off-grid micro-hydro, and demonstration installations on protected creeks, nothing else gives you that combination of high efficiency at low flow and the ability to run on debris-laden water that would shred a turbine.
- Heritage Milling: The 18 ft Fitz steel overshot wheel at Mabry Mill on the Blue Ridge Parkway, still grinding corn for park visitors using a head of roughly 5 m off Mabry Creek.
- Off-Grid Micro-Hydro: Custom 3 m steel overshot wheel installations supplied by Hydrowatt and similar specialists for remote farms in Wales and the Lake District where head is 4–6 m and design flow is 30–80 L/s.
- Living-History Sawmills: The reconstructed overshot wheel at the Hanford Mills Museum in East Meredith, New York, driving a vertical sash sawmill on Kortright Creek.
- Industrial Heritage Restoration: The 22 m Laxey Wheel (Lady Isabella) on the Isle of Man — a giant overshot wheel originally built in 1854 to pump water out of the Great Laxey lead mines.
- Iron Works Demonstration: The overshot wheels at Saugus Iron Works National Historic Site in Massachusetts, driving the bellows and trip hammer on a reconstructed 17th-century ironworks.
- Small Hydropower Generation: Grid-tied micro-hydro setups using a refurbished overshot wheel coupled to a permanent-magnet generator through a 50:1 step-up gearbox, typical of installations registered under the UK Feed-in Tariff before 2019.
The Formula Behind the Overshot Water Wheel
The shaft power output of an overshot wheel is set by the gravitational potential energy of water arriving at the top of the wheel, multiplied by the wheel's hydraulic efficiency. At the low end of the typical operating range — say a 30% bucket fill on a half-open sluice → efficiency drops to 45–55% because incomplete buckets carry a smaller moment arm. At nominal design flow with buckets running 40–50% full, a well-built wheel hits 65–75%. Push beyond rated flow and water overshoots the buckets, splashing past the rim, and efficiency falls back to 50% as you essentially throw water away. The sweet spot is design flow ±15%.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| P | Shaft power output | W | hp |
| η | Hydraulic efficiency (0.60–0.75 typical) | dimensionless | dimensionless |
| ρ | Water density | kg/m³ (1000) | lb/ft³ (62.4) |
| g | Gravitational acceleration | m/s² (9.81) | ft/s² (32.2) |
| Q | Volumetric flow rate at the launder | m³/s | ft³/s (cfs) |
| H | Effective head (launder lip to wheel discharge) | m | ft |
Worked Example: Overshot Water Wheel in a restored Cornish tin-streaming mill
You are sizing a refurbished 4.2 m diameter steel overshot wheel for a restored Cornish tin-streaming mill at Wheal Martyn near St Austell, drawing from a leat that delivers a design flow of 75 L/s with an effective head of 4.0 m from launder lip to tailwater. The wheel will drive a small Cornish stamps demonstration via a 1:8 step-up gear train. You want shaft power at nominal flow plus the swing across realistic seasonal variation.
Given
- undefined = 0.075 m³/s
- undefined = 4.0 m
- undefined = 1000 kg/m³
- undefined = 9.81 m/s²
- undefined = 0.70 dimensionless
Solution
Step 1 — compute the gross hydraulic power available at nominal design flow before applying efficiency:
Step 2 — apply nominal hydraulic efficiency for a well-built steel overshot wheel running at design flow:
That's enough to drive a small set of Cornish stamps at demonstration speed, or to feed a 1.5 kW grid-tied generator after gearbox and generator losses. At the low end of the operating range — late summer when the leat drops to 40 L/s and the buckets only fill to about 30% — efficiency falls to roughly 0.55 because incomplete buckets shorten the effective moment arm:
You'd feel this as the wheel still turning steadily but the stamps lifting noticeably slower — the gear train would need a load reduction or the demonstration runs at half cadence. At the high end after winter rain, the leat can deliver 110 L/s, but the launder is sized for 80 L/s so excess water overshoots the buckets entirely:
Less than nominal, even though more water is arriving. The overshoot losses combined with bucket splash-out push efficiency back down to 0.55. This is why the sluice gate matters — throttling back to 75 L/s in flood conditions actually gives you more shaft power than letting the full 110 L/s blast across the top of the wheel.
Result
Nominal shaft power is 2. 06 kW at 75 L/s and 4.0 m head — the wheel will turn at roughly 6 RPM and feel like a slow, deliberate, quiet machine. Compared to the 0.86 kW you get at summer low flow and the 1.73 kW you get when flood flow overshoots an unthrottled launder, the design flow really is the sweet spot — operate within ±15% of it and you stay above 65% efficiency. If you measure shaft power 25% below the predicted 2.06 kW, the most common causes are: (1) shroud-to-bucket gaps over 2 mm letting water spill sideways before the buckets reach 9 o'clock, (2) tailwater rising within 50 mm of the wheel bottom and dragging the lower buckets through standing water, or (3) launder lip set too high above the rim, causing water to shoot past the buckets rather than dropping cleanly inside.
When to Use a Overshot Water Wheel and When Not To
An overshot wheel isn't always the right pick for a low-head site, and at high flows it loses badly to turbines. Here's how it stacks up against the two alternatives a heritage mill restorer or micro-hydro builder typically considers — the breastshot wheel for moderate-head sites and a Pelton turbine for true high-head, low-flow installations.
| Property | Overshot Wheel | Breastshot Wheel | Pelton Turbine |
|---|---|---|---|
| Hydraulic efficiency | 60–75% | 50–65% | 85–92% |
| Operating speed | 4–12 RPM | 6–15 RPM | 300–1,500 RPM |
| Required head | 3–10 m | 1.5–4 m | 20–600 m |
| Typical flow range | 10–500 L/s | 50–2,000 L/s | 5–500 L/s |
| Tolerance to debris/silt | High — passes leaves, twigs | High | Low — needs fine screening |
| Capital cost (similar power) | Medium | Medium | High |
| Step-up gearing required | Yes, ~50:1 to gen | Yes, ~40:1 to gen | Often direct-drive |
| Lifespan (steel construction) | 80–120 years | 80–120 years | 30–50 years |
| Best application fit | Heritage mills, off-grid micro | Slow streams with modest fall | Mountain streams, high head |
Frequently Asked Questions About Overshot Water Wheel
Tailwater backing up. When the stream below the mill rises, the bottom of the wheel starts dragging through standing water. Each ascending bucket has to fight buoyancy and viscous drag on the way back up, which directly subtracts from net torque.
Check the gap between the lowest bucket and the tailwater surface — you want a minimum of 50 mm clearance, ideally 100 mm. Cures are deepening the tailrace cut, installing a tailwater bypass weir, or accepting that flood-flow operation requires throttling the launder sluice to keep the wheel from being overwhelmed.
Look at flow variability and bucket geometry tolerance. An overshot wheel needs head greater than wheel diameter, so at 3 m head you're capped at roughly a 2.7 m wheel — small for any serious power. A breastshot at 3 m head can be 3.5–4 m diameter, which gives more torque per unit flow and tolerates flow fluctuations better.
Rule of thumb: if head is at least 1.2× the wheel diameter you want, go overshot. If head is less than the wheel diameter you want, go breastshot. The efficiency penalty of a breastshot is real but you recover it through the bigger moment arm.
Three things almost always account for it. First, bucket vent slots — without them, air trapped in the bucket prevents full water entry, and you typically lose 8–12% efficiency. Original Fitz wheels had vent slots cut into the sole plate; many reproductions skip them.
Second, launder geometry. If water enters the bucket with too much horizontal velocity, it bounces or overshoots. Aim for 1.5–2.5 m/s entry velocity, and set the launder lip 100–150 mm above the rim.
Third, bucket count and depth being wrong for your design flow. Too few buckets means each one overfills and spills; too many means each one underfills. The classic ratio is one bucket per 250–300 mm of circumference for a wheel running at typical 4–8 RPM.
Yes, but you size for the median, not the extremes. Pick a design flow around 75–90 L/s and add a sluice gate that throttles winter excess and a bypass weir that diverts everything above wheel capacity. The wheel will deliver rated output for maybe 7 months a year and run at reduced power the rest of the time.
Don't size for peak winter flow — you'll have a wheel that's idle or running at 20% capacity for most of the year, and the bigger wheel costs roughly with diameter squared.
You're hitting the torque ceiling, which means either the bucket fill is below design or the moment arm is short. Spin the wheel slowly by hand with the launder running and watch where buckets first start spilling. If they spill before reaching 5 o'clock, the buckets are overfull — reduce launder flow or deepen the buckets. If they're nearly empty by 4 o'clock, water is escaping early through shroud gaps or vent leakage.
Also check the gear train. A pit-wheel and wallower with worn teeth or excess backlash can absorb a surprising fraction of input torque before any reaches the load. On heritage restorations we routinely see 15% transmission losses through tired wooden gearing.
Worth it if you actually need electricity and have at least 1 kW of shaft power available. The challenge is the speed mismatch — an overshot turning at 6 RPM needs a 50:1 to 100:1 step-up to drive a typical PMG at 300–600 RPM. Two-stage chain or belt drives work better than a single huge gearbox because they're easier to service and tolerate the slow input speed without the bearing brinelling that plagues high-ratio worm boxes at low input RPM.
If you only need a few hundred watts for lighting, the gearbox losses eat too much of the available power. Below 1 kW shaft, stay mechanical or use a small induction motor running as a generator with a VFD-style controller.
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.
