The Warren Central Discharge Turbine is a 19th-century inward-flow reaction turbine where water enters around the outer rim through fixed guide vanes, passes through curved runner buckets, and discharges axially down through a central opening in the wheel. Typical units ran at 80–250 RPM under 6–30 ft of head, delivering 70–78% efficiency at full gate. The design replaced overshot water wheels in textile and grist mills because it tolerated variable tailwater and packaged more horsepower into a smaller footprint — the Warren wheels installed at Lowell, Massachusetts mills in the 1850s typified the application.
Warren Central Discharge Turbine Interactive Calculator
Vary head, flow, gate opening, efficiency, and speed to see turbine power, torque, and animated inward-flow discharge.
Equation Used
The calculator applies the standard water-turbine power relation using net head and flow. For a Warren turbine, the gate setting adjusts the effective efficiency: the article notes about 55% efficiency at half gate and roughly 70-78% at full gate.
- Fresh water weight density is 62.4 lb/ft^3.
- Flow is in cubic feet per second and head is net available head.
- Gate efficiency is interpolated from about 55% at half gate to the selected full-gate efficiency.
- Mechanical losses are included in the selected efficiency.
Inside the Warren Central Discharge Turbine
Water enters the turbine case through a circular array of fixed guide vanes — sometimes called wicket gates in later designs — that direct the flow inward and tangentially onto the runner. The runner is a cast-iron wheel with curved buckets shaped to absorb both the pressure drop and the change in angular momentum as water spirals toward the centre. Once the water reaches the hub, it has nowhere to go but down through the central discharge opening, exiting axially along the vertical shaft into the tailrace. That central discharge is the defining feature — older outward-flow turbines like the Fourneyron threw water outward, which made them sensitive to backwater. The Warren design dodges that problem by dumping spent water straight down.
Geometry is what makes or breaks the unit. The bucket entry angle has to match the absolute velocity vector coming off the guide vanes within roughly ±2° — get it wrong and you generate shock losses at the leading edge, which you would notice as a hammering vibration in the shaft and a measured efficiency 10–15 points below spec. The runner-to-case radial clearance sits around 1/16 inch on a 36-inch wheel; open it up to 1/8 inch through wear or sloppy installation and you bleed leakage flow that bypasses the buckets entirely. Cast-iron buckets erode at the trailing edge first when grit-laden water passes through, and a runner that has lost 1/4 inch of trailing-edge metal will run at noticeably lower head before the gates even change position.
Flow control comes from a cylindrical gate or register ring that slides vertically to cover or expose the guide vane passages. Partial-gate operation is where the Warren design suffers — efficiency at half gate drops to roughly 55%, because the velocity triangles only line up cleanly at full flow. Practitioners running these turbines on variable streams learned to operate them either fully open or shut down, never throttled for long periods.
Key Components
- Fixed Guide Vanes: A ring of 12 to 24 stationary curved vanes around the runner that turn the incoming water from radial to tangential flow. The vane exit angle typically sits between 15° and 25° from tangential — a 20° design hits peak efficiency at one specific runner speed, so the angle is locked to the intended head and RPM.
- Inward-Flow Runner: The cast-iron wheel with 16 to 32 curved buckets that extracts work from the water. Bucket inlet angle must match the absolute flow vector within ±2°. Runners commonly measure 24 to 60 inches in outer diameter, with the central discharge opening sized at roughly 40% of outer diameter.
- Cylindrical Gate Ring: A vertically sliding sleeve that opens or closes the guide-vane passages to regulate flow. Travel is typically 6 to 12 inches on a mill-scale unit. Gate-leakage past the seal faces should stay below 2% of full flow, otherwise idle losses become unacceptable.
- Central Discharge Tube: The vertical passage below the runner hub where spent water exits into the tailrace. Diameter is sized so that exit velocity stays under 6 ft/s, otherwise you lose recoverable kinetic energy. A flared draft tube on later Warren installations recovered 3–5% additional head.
- Vertical Shaft and Footstep Bearing: The runner mounts on a vertical wrought-iron shaft supported by a footstep bearing below the discharge — usually lignum vitae or babbitt-lined bronze running submerged. Shaft runout above 0.010 inch on a 4-inch shaft will start the runner rubbing the case at the rim.
- Spiral Inlet Case: The cast-iron volute or scroll case that distributes water evenly around the guide-vane ring. Cross-sectional area decreases proportionally around the perimeter so each guide vane sees the same approach velocity within about 5%.
Who Uses the Warren Central Discharge Turbine
The Warren turbine found its niche in mid-19th-century North American industrial sites where streams provided moderate head and seasonal flow. It was the workhorse before Francis and Pelton designs displaced it, and you still find original Warren runners in restored mill museums and small heritage hydro projects. The reason builders chose it over a breast wheel or Fourneyron came down to footprint, tailwater tolerance, and the ability to run a vertical shaft straight up into a mill basement to drive line shafting through bevel gears.
- Textile Manufacturing: The Boott Cotton Mills at Lowell, Massachusetts ran multiple Warren-pattern central discharge turbines on the Pawtucket Canal in the 1850s, each driving picker and carding line shafts at roughly 120 RPM under 13 ft of head.
- Grist and Flour Milling: The Wight Grist Mill in Sterling, Massachusetts used a Warren turbine to replace its earlier breast wheel, increasing throughput on French buhr stones from 8 to 14 bushels per hour.
- Sawmill Power: Small New England sawmills along the Merrimack and Connecticut rivers fitted Warren wheels in the 30-inch class to drive circular saws through belt-and-pulley reductions, replacing seasonal-water-limited overshot wheels.
- Heritage Micro-Hydro Restoration: The Slater Mill historic site in Pawtucket, Rhode Island restored its turbine pit installation using period-correct central-discharge wheel geometry to demonstrate working 19th-century mill power.
- Paper Mills: The Crane & Co. paper mill at Dalton, Massachusetts used Warren-type turbines in the late 1800s to drive Hollander beaters and Fourdrinier rolls before electrification.
- Iron Works and Forges: Trip-hammer forges in the Berkshires drove tilt hammers off Warren turbine output through cam shafts, benefiting from the steady torque available across moderate head variation.
The Formula Behind the Warren Central Discharge Turbine
Power output of a Warren central discharge turbine follows the standard hydraulic power equation modified by the unit's gate-position efficiency curve. What matters for a practitioner sizing or restoring one of these is how the result shifts across the head and flow range your site actually delivers. At the low end of typical mill-scale head (6 ft) the unit is barely earning its keep — you might pull 8 hp from a 36-inch wheel. At nominal head (13–15 ft) you hit the design sweet spot where efficiency peaks near 75%. Push the head past 25 ft and you start cavitating the bucket trailing edges because the design predates modern blade-loading analysis. Flow scales linearly, head scales linearly, and efficiency moves with gate position — so all three need to be evaluated together, not as a single nameplate number.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| P | Shaft power output | W | ft·lbf/s (÷550 for hp) |
| η | Overall turbine efficiency at the operating gate position | dimensionless (0–1) | dimensionless (0–1) |
| ρ | Water density | kg/m³ (≈1000) | lbm/ft³ (≈62.4) |
| g | Gravitational acceleration | 9.81 m/s² | 32.2 ft/s² |
| Q | Volumetric flow rate through the runner | m³/s | ft³/s (cfs) |
| H | Net hydraulic head across the turbine | m | ft |
Worked Example: Warren Central Discharge Turbine in a restored heritage textile mill turbine
You are sizing a restored 36-inch Warren central discharge turbine for the rebuilt power pit at the Harmony Mills heritage textile complex on the Mohawk River near Cohoes, New York. The site provides a measured net head of 14 ft after penstock losses, with flow varying seasonally from 18 cfs in late summer up to 42 cfs during spring runoff. The runner is rated at 75% efficiency at full gate, dropping to 58% at half gate. You need to predict shaft horsepower across the operating range to size the bevel-gear takeoff to a re-installed line shaft.
Given
- H = 14 ft
- Qnom = 30 ft³/s
- Qlow = 18 ft³/s
- Qhigh = 42 ft³/s
- ηfull = 0.75 —
- ηhalf = 0.58 —
- ρ × g (water, imperial) = 62.4 lbf/ft³
Solution
Step 1 — at nominal flow of 30 cfs with full gate at 14 ft head, compute the gross hydraulic power and convert to shaft horsepower:
That 35.7 hp at roughly 120 RPM is exactly what a mill of this size would expect — enough to drive a full picker line plus carding through a bevel gear and line shaft, with reasonable margin.
Step 2 — at the low end of the typical operating range, late-summer flow of 18 cfs. The gate has to throttle to half open, dropping efficiency to 0.58:
Less than half the nominal output. In practice this means the mill operator either runs a reduced shift on fewer machines, or shuts down entirely until flow recovers — the half-gate efficiency penalty makes long-duration partial operation economically painful, and you would feel it in the shaft as a slightly rougher cadence because the velocity triangles no longer match the bucket geometry cleanly.
Step 3 — at the high end, spring runoff of 42 cfs at full gate:
Theoretically 50 hp is on the table, but in reality the runner starts to cavitate at flows above roughly 38 cfs on this geometry because the bucket trailing edge wasn't designed for the higher relative velocity. You would hear a distinct gravelly rattle through the case and see pitting on the runner trailing edges within a season. Most heritage operators cap throughput at 35 cfs and let the excess spill over the bypass weir.
Result
Nominal shaft output is 35. 7 hp at 30 cfs and 14 ft head, which translates to a comfortable 120 RPM drive for the bevel-geared line shaft and matches the original Harmony Mills nameplate within 2 hp. Across the operating range the unit delivers 16.6 hp at 18 cfs (low summer), 35.7 hp at nominal, and a theoretical 50 hp at 42 cfs runoff — but the practical sweet spot is 28–35 cfs at full gate, because efficiency collapses below that and cavitation kicks in above it. If your measured output is 20% below predicted, check three things in order: (1) gate-ring leakage past worn bronze seal faces, which can bypass 5–10% of flow without you noticing, (2) bucket trailing-edge erosion measured with a profile gauge — anything beyond 1/8 inch of metal loss noticeably shifts the velocity triangle, and (3) penstock head loss greater than your assumed value, often caused by partially blocked trash racks adding 1–2 ft of unaccounted-for friction head.
Choosing the Warren Central Discharge Turbine: Pros and Cons
The Warren central discharge turbine sits between the older Fourneyron outward-flow design and the later Francis mixed-flow turbine. If you are restoring a heritage site or designing a low-budget micro-hydro install, the comparison matters because each option has different penalties on partial-flow efficiency, footprint, and tailwater behaviour.
| Property | Warren Central Discharge Turbine | Fourneyron Outward-Flow Turbine | Francis Mixed-Flow Turbine |
|---|---|---|---|
| Peak efficiency at full gate | 70–78% | 65–80% | 88–94% |
| Efficiency at half gate | ~55% | ~50% | ~80% |
| Typical operating head | 6–30 ft | 3–20 ft | 30–2000 ft |
| Tailwater submergence tolerance | High — central discharge handles it | Low — backflow stalls runner | High with proper draft tube |
| Speed range typical for mill-scale | 80–250 RPM | 50–150 RPM | 100–600 RPM |
| Capital cost (mill-scale, period-correct) | Moderate | High — larger casting | Highest — precision blading |
| Best application fit | Restored 19th-century mill sites, low-mod head heritage hydro | Museum demonstration only | Modern micro and macro hydro plants |
| Maintenance interval (cast-iron runner, clean water) | 10–20 years between rebuilds | 8–15 years | 20–40 years |
Frequently Asked Questions About Warren Central Discharge Turbine
The central discharge tube relies on free or nearly free exit conditions to dump spent water without backpressure. When tailwater submerges the exit by more than roughly 12 inches, you start adding a column of water that the runner has to push against — effectively reducing your net head by that submergence depth. On a 14 ft head site, losing 1.5 ft to backwater is over 10% of your driving head, and power scales linearly with head.
The fix on heritage sites is either deepening the tailrace channel below the discharge or fitting a flared draft tube that recovers some of the exit kinetic energy. Original Warren installations rarely had proper draft tubes — that was a Francis-era refinement.
It comes down to permitting and output requirements. If the site is on a historic register and the wheel pit geometry is original, regulators often require period-correct equipment, which forces the Warren rebuild. If the goal is generating revenue from grid-tied micro-hydro, a Francis runner will deliver 15–20% more annual energy on the same flow because of its dramatically better part-load efficiency.
Rule of thumb: under 25 hp nominal and heritage-driven, rebuild the Warren. Over 50 hp nominal and revenue-driven, retrofit Francis and keep the original runner as a static display.
You are hitting a hydraulic resonance where the guide-vane passing frequency lines up with a structural mode of the runner or shaft. With 16 buckets and 18 guide vanes, the blade-passing frequencies create excitation pulses, and at one specific RPM those pulses match the natural frequency of the wrought-iron shaft.
The diagnostic is to measure shaft frequency response with a hammer test off-line and compare it to N × nvanes / 60. If they cross within your operating band, your options are: change runner speed by re-gearing the takeoff, add a damping mass to the shaft, or — the period-correct fix — replace one or two guide vanes to break the symmetry.
A 25% shortfall on a Warren turbine almost always traces to one of three places not covered in the worked example: spiral case maldistribution, footstep bearing drag, or air ingestion at the runner crown. Spiral case maldistribution happens when the volute hasn't been re-trued during rebuild — uneven approach velocity to the guide vanes throws off velocity triangles around 3–4 of the vanes and you lose 5–8% efficiency.
Footstep bearing drag on a worn lignum vitae bearing can eat 2–3 hp on its own. Check it by spinning the runner by hand with no water — it should rotate freely with under 50 lb-ft of breakaway torque on a 36-inch wheel.
Air ingestion shows up as a milky discharge stream and chuffing noise. It usually means the headwater level dropped below the minimum submergence for the case inlet, sucking air down through a vortex.
Mechanically yes, hydraulically no. The bucket profiles on a Warren runner are designed for inward-flow energy extraction — running them backward as a pump-as-turbine gives you maybe 30–40% efficiency because the bucket exit edges become unfavourable inlet edges with separation everywhere.
If you need pumped storage on a heritage-style installation, use a dedicated reversible Francis pump-turbine. The Warren design predates the analytical fluid mechanics needed to make a runner reversible.
This catches restorers regularly. Original Warren guide vanes were cast iron with a hard chilled surface from the casting process — surface hardness around 350–400 HB on the leading edge. Modern bronze replacements at maybe 120–150 HB look prettier but erode 3–4 times faster in grit-laden water.
If you are rebuilding for service rather than display, specify ductile iron guide vanes with a hard-chrome leading-edge overlay, or use a high-tin aluminium bronze (C95400+) with at least 180 HB hardness. The original metallurgy was chosen for a reason.
References & Further Reading
- Wikipedia contributors. Water turbine. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
