A drainage wheel is a large-diameter paddle or scoop wheel that lifts water through a small head — typically 0.3 to 2 m — by rotating partially submerged paddles inside a close-fitting masonry or steel trough. The trough is the critical component; its curved bottom and side walls keep clearance to the paddle tips at 5 to 15 mm so water cannot slip back past the blades. Drainage wheels exist to move very large flows against very low heads cheaply, which is exactly the regime where centrifugal pumps run badly. They drained more than 600,000 hectares of English Fenland and Dutch polder by the late 19th century, much of it still below sea level today.
Drainage Wheel Interactive Calculator
Vary flow, lift, wheel speed, and efficiency to see power and shaft torque for a low-head drainage wheel.
Equation Used
The calculator estimates the hydraulic water power needed to lift a given flow through a low head, then divides by efficiency to find shaft power. Shaft torque is found from power divided by angular speed, which is why slow drainage wheels need high torque even at modest power.
- Water density is 1000 kg/m3.
- Head is the vertical lift between sump and higher channel.
- Efficiency represents hydraulic and mechanical losses together.
- Steady average flow with no allowance for leakage variation from tip clearance.
Inside the Drainage Wheel
A drainage wheel works by trapping water between paddles and the curved trough wall, then carrying that water around the rim and dumping it over a low weir on the discharge side. The wheel rotates slowly — usually 4 to 8 RPM at the rim — and each paddle scoops a near-fixed volume of water determined by paddle width, paddle depth into the trough, and trough clearance. You can think of it as a positive-displacement pump made of timber and brick instead of pistons and cylinders. Because the lift is small and the volume per revolution is large, the rim speed stays low and the water leaves the discharge with almost no kinetic energy, which is why scoop wheels hit 60 to 75% hydraulic efficiency on heads under 1.5 m where a centrifugal would barely break 40%.
The geometry has to be right or the wheel pumps backwards. Tip clearance between paddle and trough is the number one variable — open it up to 25 mm and you lose roughly half your discharge to slip, because water flows back under the paddle faster than the next paddle picks it up. Submergence matters too: the paddle must enter the water cleanly without slapping (which wastes energy as splash and damages the paddle roots), and it must leave the water at the discharge weir without lifting more than a thin film over the top. If you size the wheel diameter wrong relative to the head, the paddles either run too deep on the discharge side and churn the discharge pool, or run too shallow on the intake and lose prime when the sump level drops.
Common failure modes are timber paddle splitting at the bolted root, trough wear scoring out the brick or cast-iron lining and widening tip clearance over decades, and shaft journal wear on the wooden bearings used in older mills. The Stretham Old Engine in Cambridgeshire ran from 1831 to 1925 with a 11.3 m diameter scoop wheel, and the main rebuild work over that 94 year life was replacing oak paddles every 15 to 20 years and re-lining the trough twice.
Key Components
- Paddle (or float): Flat or slightly curved blade that pushes the water around the trough. Typically oak or steel plate 25 to 50 mm thick, bolted into iron arms radiating from the hub. Width sets the flow per revolution — a 1.2 m wide paddle on a 9 m wheel moves roughly 0.8 m³ per paddle pass.
- Trough (or race): Curved channel that mates with the paddle tips on a 5 to 15 mm clearance. Lined with brick, cast iron segments, or steel plate. Wear here is the dominant long-term failure mode — every extra millimetre of clearance costs about 4% discharge.
- Hub and arms: Cast iron hub keyed to the main shaft, with 8 to 16 wrought iron or steel arms supporting the paddle ring. Arms must take both bending from paddle reaction and torsion from out-of-balance loading when one side is in water and the other is dry.
- Main shaft and journals: Heavy forged or cast shaft, 200 to 400 mm diameter on large mills, running in bronze or lignum vitae bearings. Slow rotation (4 to 8 RPM) means hydrodynamic lubrication is marginal, so bearings rely on grease or water lubrication and need oversize projected bearing area.
- Drive (gearing or direct): Reduces prime mover speed to wheel speed. Wind-driven Fen mills used wooden compass-arm gearing at roughly 8:1 reduction; later steam engines used cast iron spur gears at 30:1 or higher. The drive must handle the full stall torque the wheel can generate when the trough silts up.
- Discharge weir: Low sill at the top of the trough over which lifted water spills into the higher channel. Weir crest height sets the actual head against which the wheel works — getting this 50 mm wrong on a 0.6 m lift changes power demand by 8%.
Real-World Applications of the Drainage Wheel
Drainage wheels dominate the low-head, high-flow corner of pump duty — typically 0.3 to 2 m of lift at flows of 0.5 to 5 m³/s. They are the right answer where the water source is a wide shallow drain, the destination is a slightly higher river or sea wall, and electricity is either expensive or absent. Modern installations are rare but heritage restorations and developing-world land reclamation schemes still build them.
- Land drainage (heritage): Stretham Old Engine, Cambridgeshire — 11.3 m diameter scoop wheel driven by a 60 hp Butterley beam engine, lifted Fen drainage water 1.7 m into the Old West River from 1831 to 1925.
- Polder reclamation: Cruquius Pumping Station near Haarlem in the Netherlands originally used scoop wheels before its 1849 conversion to Cornish pumps, and many smaller Dutch polder mills (Kinderdijk windmill complex) still run scoop wheels under wind power.
- Mine dewatering (historical): Wheal Martyn china clay pit in Cornwall used a large overshot wheel coupled to scoop drainage on the secondary working face, lifting groundwater to the launder system.
- Tidal marsh management: Wicken Fen National Trust reserve operates a restored wind-driven scoop wheel to maintain water levels in the protected wetland, lifting water on a managed schedule instead of letting it drain naturally.
- Agricultural irrigation: Norfolk Broads grazing marsh systems and parts of the Somerset Levels still use small electrically-driven scoop wheels at 0.5 to 1.0 m lifts to manage seasonal water levels for cattle pasture.
- Developing-world drainage: Bangladesh and Nile Delta smallholder schemes occasionally build sub-3 m scoop wheels driven by diesel or animal power for paddy field drainage where centrifugals would cavitate or air-bind at low head.
The Formula Behind the Drainage Wheel
The shaft power a drainage wheel needs is set by the flow it must lift, the head it must lift against, and how efficiently the paddle-trough geometry converts mechanical work into useful pumping. At the low end of typical operating range — a 0.3 m lift moving 0.5 m³/s — power demand is tiny (under 3 kW theoretical) and a small wind drive or a horse gin handles it. At the high end — 2.0 m lift moving 5 m³/s — you are into 130 kW theoretical input, which is why the big Fen pumping stations needed beam engines and later diesel sets. The sweet spot for a scoop wheel sits around 0.6 to 1.2 m lift at 1 to 3 m³/s, where efficiency peaks near 70% and the wheel diameter stays under about 12 m so the structure remains buildable.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Pshaft | Shaft power required to drive the wheel | W | hp |
| ρ | Water density (≈1000 for fresh, ≈1025 for brackish) | kg/m³ | lb/ft³ |
| g | Gravitational acceleration (9.81) | m/s² | ft/s² |
| Q | Volume flow lifted by the wheel | m³/s | ft³/s (cusec) |
| H | Lift head from sump water level to discharge weir crest | m | ft |
| ηw | Hydraulic efficiency of the wheel (typically 0.55 to 0.75) | dimensionless | dimensionless |
Worked Example: Drainage Wheel in a restored heritage scoop wheel at a coastal grazing marsh
You are sizing the electric drive for a restored 8.4 m diameter timber-and-iron scoop wheel at the Elmley Marshes RSPB reserve on the Isle of Sheppey in Kent, lifting brackish drainage water from the grazing-marsh ditches over the sea wall into the Swale estuary. Trough clearance is 10 mm, paddle width 1.0 m, and the discharge weir sits 0.9 m above mean sump level. You want to know the motor size and what the wheel will actually do across the seasonal flow range.
Given
- D = 8.4 m
- Paddle width = 1.0 m
- H = 0.9 m
- ρ = 1025 kg/m³ (brackish)
- ηw = 0.68 dimensionless
- Qnom = 1.6 m³/s
Solution
Step 1 — at nominal winter drainage flow of 1.6 m³/s, compute the hydraulic (water) power being delivered:
Step 2 — divide by wheel efficiency to get shaft power demand at nominal flow:
Step 3 — at the low end of the typical operating range, summer trickle flow of about 0.4 m³/s, the same head still applies because the weir crest does not move:
This is light duty — a VFD-driven 22 kW motor would loaf along well below half load, and you would feel the wheel barely under torque if you put a hand on the gearbox casing. Step 4 — at the high end, a storm-event flush of 3.0 m³/s, the wheel is at the upper limit of what an 8.4 m diameter machine can pass without the paddles burying themselves on entry:
At that flow the wheel is moving water about as fast as it physically can — you will see the paddles entering the sump with visible splash, efficiency is already dropping below 0.65, and pushing harder buys you nothing because the trough cannot swallow more without overtopping the upstream sluice.
Result
Specify a 45 kW VFD-driven geared motor — that gives full headroom for the 39. 9 kW storm-event demand plus margin for trough silt loading and paddle wear over the next decade. At nominal 1.6 m³/s the wheel pulls about 21 kW and turns at roughly 5 RPM rim speed, which feels almost stationary to a visitor on the bank. The low-end summer 0.4 m³/s case at 5.3 kW versus the storm 3.0 m³/s case at 39.9 kW shows an 8x power swing across the operating range — the sweet spot for steady running sits between 1.0 and 2.0 m³/s where ηw stays above 0.65. If you measure shaft power 30% higher than these predictions, check trough tip clearance first (every extra mm above 10 mm costs roughly 4% efficiency and shows up as power demand), then look for paddle root cracking that lets water blow past the blade, and finally check the discharge weir crest level — a weir that has settled 100 mm low silently raises the head you actually pump against.
When to Use a Drainage Wheel and When Not To
Drainage wheels compete with Archimedes screws and low-head axial-flow propeller pumps in the same duty band. The choice between them comes down to head, flow, civil cost, and how much fish-friendliness or maintenance access you need.
| Property | Drainage Wheel | Archimedes Screw | Axial-Flow Propeller Pump |
|---|---|---|---|
| Useful head range | 0.3–2.0 m | 1.0–8.0 m | 0.5–6.0 m |
| Typical flow range | 0.5–5 m³/s | 0.05–4 m³/s | 0.2–20 m³/s |
| Peak hydraulic efficiency | 60–75% | 70–85% | 75–88% |
| Rotational speed | 4–8 RPM | 20–60 RPM | 300–1500 RPM |
| Civil works cost (relative) | High — large trough and pit | Medium — inclined trough | Low — sump and short pipe |
| Service life before major rebuild | 50–100 years (timber paddles every 15–20 yr) | 30–50 years | 15–25 years |
| Fish and debris tolerance | Excellent — slow paddles, no shear | Excellent — fish-friendly certified | Poor — high tip speed shreds fish |
| Best application fit | Heritage drainage, wetland reserves | Modern land drainage, fish passage sites | High-flow stormwater, irrigation |
Frequently Asked Questions About Drainage Wheel
The wheel is sensitive to paddle submergence on the intake side and almost insensitive to discharge submergence as long as the weir is not drowned. When sump level drops, the paddles enter the water shallower, each scoop carries less volume, and you lose flow roughly in proportion to the lost paddle immersion depth. Discharge level can rise quite a bit before it starts pushing back into the trough — only when the discharge channel actually backs up over the weir crest do you start to see drowned-flow losses.
Quick diagnostic: drop a stick in the sump and measure how deep the paddle tip goes below the surface. If immersion is under about 60% of paddle depth, your sump intake is silting up or the upstream sluice is throttled.
Modern Archimedes screw, almost every time. Both are fish-friendly because tip speeds are low, but the screw has a 10 to 15 percentage-point efficiency advantage in the 1 to 4 m head band, the civil works are cheaper because you only need an inclined trough rather than a deep pit and tailrace, and screws come as packaged units from suppliers like Spaans Babcock or Landustrie with documented eel and salmonid passage data.
The drainage wheel only wins if you are restoring an existing structure for heritage reasons, if you specifically want the visible mechanical character (wetland reserves, museum sites), or if your head is genuinely under about 0.6 m where screw efficiency starts to fall off because the screw needs slope to work.
Almost certainly the trough lining. Original cast-iron or brick trough wear over decades typically opens tip clearance from a build-spec 8–10 mm to 20+ mm. New paddles cut to original drawings now run in an oversized trough, so slip past the paddle tip is much higher than the historical machine ever saw — even when the historical machine was new the trough was probably tighter than the drawings suggest because it was hand-finished to fit the paddles.
Fix is to re-line the trough with bolted steel segments scribed to the actual paddle swept circle, set to 8 mm cold clearance. Expect to recover most of the lost discharge.
Two things usually happen together at low sump. First, paddle entry stops being clean — the paddle slaps the water surface instead of slicing in, and that splash energy shows up as shaft torque without producing any lift. Second, the wheel can start carrying air pockets around the trough on the underside, which then collapse noisily and load the bearings unevenly.
If you are seeing torque rise as flow falls, you are running below the wheel's designed minimum submergence. Either fit a sump-level cutout that stops the wheel below a set level, or step the motor down with the VFD so rim speed drops and entry stays clean.
Textbook 0.70 figures assume a wheel built and maintained to original Fen-engineering standards: tip clearance 8–12 mm, paddle root sealed and unsplit, trough geometry matching the paddle swept circle within 5 mm anywhere around the arc, and the discharge weir set so paddles leave the water with only a thin lifted film. Real surveyed wheels almost never hit all four conditions.
0.55 is what you get with 18 mm tip clearance, a couple of cracked paddles, and a discharge weir that has dropped 50 mm since installation. That's normal for a working heritage machine. To recover textbook efficiency you have to attack all four geometry sources, not just one.
You can, and most modern restorations do, but the useful turndown is narrower than the VFD's electrical range suggests. Below about 50% of design rim speed the paddle entry geometry stops working — at low rim speed the water surface has time to deflect away from the approaching paddle, the scoop fills incompletely, and discharge per revolution drops faster than rim speed does. Practical turndown is roughly 50 to 110% of design speed.
Above 110% you start to see paddle root stresses climb sharply because hydrodynamic load on the paddle scales with the square of rim speed. Original fixed-speed mills sat at one design point and never had to worry about this — your VFD-equipped restoration does.
References & Further Reading
- Wikipedia contributors. Scoop wheel. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
