A pitchback water wheel is a gravity-driven wheel where water enters the buckets just behind top-dead-centre and falls on the upstream side, turning the wheel back toward the supply launder. Unlike an overshot wheel which throws water down the far side, the pitchback drops it back into the same channel — handy when tailrace flooding would otherwise drown the wheel. It captures roughly the full hydraulic head as gravitational potential energy and converts it at 60–75% efficiency. Small mills and craft workshops use it for steady 1–10 kW shaft output.
Pitchback Water Wheel Interactive Calculator
Vary water flow, head, efficiency, and wheel diameter to see shaft power, speed, and torque for a pitchback water wheel.
Equation Used
The available water power is rho g Q H. Multiplying by wheel efficiency eta estimates useful shaft power. The speed and torque estimates use a typical pitchback rim speed of 1.8 m/s, within the article's recommended slow-running range.
- Fresh water density is 1000 kg/m^3.
- Gravity is 9.81 m/s^2.
- Bucket and leakage losses are represented by the efficiency input.
- Wheel speed and torque use an assumed best-practice rim speed of 1.8 m/s.
The Water Wheel (form 2) in Action
The pitchback wheel sits under a launder that ends just past the top of the wheel, so water drops into the buckets a few degrees behind top-dead-centre. Gravity pulls the loaded buckets down the upstream side, the same side the water came from. The wheel turns the opposite direction to an overshot — and that's the whole point. If your tailrace backs up during winter floods, an overshot wheel drowns at the bottom and stalls, but a pitchback dumps its spent water back into the headrace channel where the level can rise without touching the wheel.
Bucket geometry decides whether you actually capture the head you paid for. The buckets need a vent slot — usually 6 to 10 mm wide along the inner rim — so trapped air can escape as water enters and re-enter as water spills out at the bottom. Skip the vents and the buckets fill maybe 40% by volume because air can't get out. Lip angle matters too: tilt the bucket lip 15–25° forward of radial so water enters cleanly without splashing back over the rim. Set it radial or backward and you lose 10–20% of the inflow as overspill before the bucket even passes the launder.
The wheel runs slow on purpose. Rim speed sits between 1.4 and 2.2 m/s for best efficiency — push it faster and centrifugal force throws water out of the buckets before they reach the bottom. Run it slower and water spills over the leading lip during the descent. Common failure modes: bearing seizure from a flooded shaft journal (raise the bearings above 100-year flood level), bucket corrosion at the welds when you use mild steel without galvanising, and ice loading in winter that doubles the rim weight and snaps undersized arms.
Key Components
- Launder (headrace flume): Channels water from the millpond or weir to the top of the wheel. The discharge lip must sit 50–150 mm behind top-dead-centre so water enters the bucket cleanly without back-splashing onto the rim. Cross-section sized for the design flow at 0.6–1.0 m/s — too fast and water shoots over the buckets, too slow and silt drops out and clogs the channel.
- Buckets (or compartments): Hold water on the descending arc. Depth typically 200–400 mm, count 32–48 around the rim. Vent slots 6–10 mm wide along the inner edge let air escape — without them the bucket fills less than half and you lose most of your power.
- Sole plate and shrouds: The sole is the cylindrical inner wall the buckets sit on; the shrouds are the side discs. Together they form the sealed pocket of each bucket. Shroud-to-bucket gap should be under 2 mm — wider and water leaks sideways during the descent, dropping efficiency by 5–15%.
- Cast iron or steel arms: Connect the rim to the central shaft. On a 4 m diameter wheel running 1 tonne of water in the loaded buckets, each arm carries 2–4 kN of bending load. Wood arms work for diameters under 3 m; steel or cast iron for anything bigger.
- Main shaft and journals: Carries the torque to the gearing or pulley. Typical shaft diameters 100–250 mm depending on output, supported on plummer-block bearings or older lignum-vitae bushes. Bearings sit above the high-water mark — flood the journal once and you're stripping the wheel.
- Sluice gate: Controls flow into the launder. Lets you start, stop, and throttle the wheel without diverting the whole stream. Manual rack-and-pinion type is standard; modern installations use a linear actuator on the gate stem for remote control.
Real-World Applications of the Water Wheel (form 2)
Pitchback wheels show up wherever the available head is high enough to act like an overshot but the tailrace can't be trusted to stay clear. They're the practical choice for restored heritage mills, small craft producers running off historic millponds, and micro-hydro sites where flood backwater would otherwise idle the plant for weeks every winter. Output ranges from a fraction of a kilowatt on garden-scale demonstration builds up to 15–20 kW on full-size restored industrial wheels.
- Heritage milling: The restored pitchback wheel at Mapledurham Watermill on the Thames, used for stoneground flour production, runs as pitchback specifically because Thames flood levels regularly back up into the tailrace.
- Craft distilling: Small Highland whisky distilleries with hillside burns use 3–5 m diameter pitchback wheels to drive malt mills and grain elevators when the tailrace runs into a peat-bog drainage area prone to flooding.
- Micro-hydro electricity: Off-grid homesteads in upstate New York and Vermont fit pitchback wheels coupled to 5–15 kW permanent magnet generators where stream tailwater rises 1–2 m in spring melt.
- Working museum demonstrations: The Wheal Martyn clay-works museum in Cornwall and similar sites run pitchback wheels for public demonstration where the tailrace doubles as a visitor pond with fluctuating level.
- Sawmill restoration: Restored 19th-century sawmills in the German Black Forest use pitchback wheels of 4–6 m diameter to drive reciprocating frame saws via a pitman arm at 80–120 strokes per minute.
- Small textile workshops: Heritage fulling mills and woollen workshops in the Yorkshire Dales drive stocks and tucking hammers off pitchback wheels where the original mill leat empties back into the same beck.
The Formula Behind the Water Wheel (form 2)
The shaft power a pitchback wheel produces depends on three things: how much water you have, how far it falls through the wheel, and how cleanly the buckets capture and release it. At the low end of your typical flow range you get less than half the nominal output because the buckets only fill partially and bearing friction takes a fixed bite regardless of load. At the high end, overflow at the launder dumps water past the wheel entirely and efficiency collapses. The sweet spot sits where the launder discharge matches the bucket fill rate at design rim speed — usually within ±20% of nominal flow.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Pshaft | Useful shaft power output | W (watts) | hp (horsepower) |
| η | Overall hydraulic-to-shaft efficiency, typically 0.60–0.75 for a well-built pitchback | 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 through the launder | m³/s | ft³/s (cfs) |
| H | Effective hydraulic head — vertical distance from launder lip to bucket exit point | m | ft |
Worked Example: Water Wheel (form 2) in a restored Black Forest sawmill
You are restoring a 19th-century frame-saw operation in the Black Forest. The site has a hillside leat delivering a nominal 0.18 m³/s with seasonal range from 0.10 m³/s in late summer to 0.28 m³/s in spring melt. Effective head from launder lip to bucket exit is 3.2 m. The wheel is 3.6 m diameter, steel, with 40 buckets, a 25° lip angle and 8 mm vent slots. You need to know whether the wheel can reliably drive the reciprocating frame saw, which needs about 4 kW at the shaft to cut softwood at 100 strokes per minute.
Given
- Qnom = 0.18 m³/s
- Qlow = 0.10 m³/s
- Qhigh = 0.28 m³/s
- H = 3.2 m
- η = 0.68 dimensionless
- ρ = 1000 kg/m³
- g = 9.81 m/s²
Solution
Step 1 — compute the available hydraulic power at nominal flow before efficiency losses:
Step 2 — apply the 0.68 efficiency factor for a well-built pitchback wheel with vented buckets and tight shrouds:
That clears the 4 kW saw demand only barely — and barely is not good enough on a working sawmill where chip load varies. You'll either need to run the saw a touch slower or accept that on hard knots the wheel will drop rim speed under the load. Now check the seasonal extremes.
Step 3 — at the low end of the typical operating range, late-summer flow of 0.10 m³/s:
At 2.1 kW you cannot drive the frame saw at all — it will stall on the first cut. In late summer the mill must either run a smaller secondary load like a grain mill, or shut down. This is the reality of seasonal hydro and you need to plan for it.
Step 4 — at the high end, spring melt of 0.28 m³/s:
In theory you'd see 6 kW, but in practice 0.28 m³/s overruns the launder cross-section sized for nominal flow. Excess water spills over the front lip and bypasses the wheel entirely, so you measure perhaps 4.5–5.0 kW with the rest pouring uselessly down the bypass. The sluice gate should be partially closed in spring to hold flow nearer 0.20 m³/s and avoid wasting head on a flooded launder.
Result
Nominal shaft output is 3. 8 kW at 0.18 m³/s and 3.2 m head. That is just under the saw's 4 kW demand — workable for steady cuts in clean softwood but marginal whenever the blade hits a knot, and you'll feel the wheel labour and slow audibly under heavy load. The seasonal range runs from 2.1 kW in late summer (saw cannot run at all) up to a theoretical 6.0 kW in spring melt (in practice closer to 4.5–5.0 kW because excess water overshoots the launder), with the design sweet spot sitting at the nominal flow where launder discharge matches bucket fill rate. If you measure 2.5 kW instead of the predicted 3.8 kW at nominal flow, look first at bucket fill — clogged or undersized vent slots cap fill at 40–60% volume; second at the shroud-to-bucket gap, where wear above 3 mm bleeds water sideways during the descent; and third at the launder lip position, which should sit 50–150 mm behind top-dead-centre — anything further back and water enters too late to ride the bucket all the way down.
Choosing the Water Wheel (form 2): Pros and Cons
The pitchback competes mainly with the overshot and the high breastshot wheel. All three sit in the same head range but differ in how they handle tailwater, splash losses, and direction of rotation. Pick wrong and you either drown the wheel every winter or pay for head you never use.
| Property | Pitchback wheel | Overshot wheel | High breastshot wheel |
|---|---|---|---|
| Typical hydraulic efficiency | 60–75% | 65–85% | 55–70% |
| Useful head range | 2.5–6 m | 3–10 m | 1.5–4 m |
| Tolerance to tailrace flooding | High — water exits upstream side | Low — backwater drowns the wheel | Medium — partial submersion tolerable |
| Typical rim speed for peak efficiency | 1.4–2.2 m/s | 1.5–2.5 m/s | 1.0–1.8 m/s |
| Direction of rotation | Toward the launder (upstream) | Away from launder (downstream) | Away from launder (downstream) |
| Construction complexity | Medium — vented buckets, careful launder placement | Medium — vented buckets, longer launder | High — curved buckets matched to breast race |
| Typical shaft power range | 1–15 kW | 1–25 kW | 0.5–10 kW |
| Best application fit | Sites with variable tailwater level | Steep sites with free-draining tailrace | Lower-head millponds with stable level |
Frequently Asked Questions About Water Wheel (form 2)
Because that's how the geometry works — and it's correct, not wrong. On a pitchback the launder ends just past top-dead-centre, so water enters the bucket and gravity pulls the loaded side back down toward the launder. The wheel rotates upstream, opposite to an overshot.
If you wanted overshot rotation you'd need to extend the launder past TDC by another 100–200 mm and accept that spent water then exits the downstream side — which puts you back at risk of tailrace drowning. The reverse rotation is the whole reason to choose pitchback in the first place.
Buckets that look full from the side often aren't. Without proper vent slots in the inner rim, trapped air keeps water out of the back third of each bucket. From outside the wheel you see water at the lip and assume the bucket is full, but it's a half-cup at best.
Drill or cut 6–10 mm vent slots along the inner edge of each bucket, parallel to the shaft. Most restoration jobs that come in at half power are missing or clogged vents. Second most common: the shroud-to-bucket gaps have opened up past 3 mm and water leaks sideways during the descent.
Size the wheel diameter and bucket capacity for nominal flow, then size the sluice gate and launder to throttle peak flow back to design. You cannot usefully build for peak — a wheel sized for 0.28 m³/s running at 0.10 m³/s in summer will fill its buckets only 35% and run inefficiently all summer.
Pick nominal, accept that summer output drops with flow, and add a bypass channel to dump excess spring flow around the wheel without flooding the launder. This is how every working heritage mill handles seasonal variation.
Use 0.65 for a careful new build with steel buckets, vented inner rims, and tight shroud gaps. Use 0.55 for a wood-bucket restoration where you can't guarantee the seal between buckets and shrouds. Use 0.70–0.75 only after you've measured actual output and confirmed the geometry is right.
Published figures of 80%+ refer to laboratory wheels with machined buckets and zero shroud leakage. A real wheel in a real stream rarely hits 75%.
Two likely causes. First, your launder is undersized for spring flow, so excess water shoots over the front lip and bypasses the wheel — you have less effective flow in spring than in autumn even though the stream is higher. Second, spring brings debris: leaves from autumn that washed downstream all winter pile up against the trash rack and starve the launder.
Check the trash rack first — clear it and watch the launder level. If it rises immediately, you had a debris problem. If the launder is overflowing freely with a clear rack, partially close the sluice to throttle flow back to design.
The buckets must empty before they reach the bottom-dead-centre position — typically by about 30–40° before BDC for a properly profiled bucket. If water is still in the bucket past BDC, you're carrying dead weight up the back side and bleeding power.
The lip angle (15–25° forward of radial) and bucket depth together control when each bucket spills. If yours are still loaded at BDC, the lip is too steep or the bucket too deep for your rim speed. Reduce bucket depth or shallow the lip angle and the spill point moves earlier.
You need a flywheel for any reciprocating load like a frame saw, pitman-driven hammer, or piston pump. The wheel itself has rotational inertia but not enough to smooth the torque pulses from a reciprocating cut — without a flywheel the wheel rim speed surges and dips with every stroke, and over a few seconds the bucket fill timing drifts off the launder discharge and you lose 10–15% efficiency.
A flywheel on the saw shaft sized at 5–10× the wheel's own moment of inertia smooths the load enough that the wheel sees a near-steady torque demand. Most 19th-century mills had this built in as the saw frame itself.
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.
