Horizontal axis wind turbines (HAWTs) dominate utility-scale power because a three-bladed, lift-driven rotor on a tall tower reaches a power coefficient of about 0.45 to 0.50 and sits in the stronger winds higher up. Vertical axis wind turbines (VAWTs) peak lower, about 0.30 to 0.40 for a Darrieus and 0.15 to 0.25 for a Savonius, so they are kept for niche uses: small, urban, rooftop or turbulent sites where accepting wind from any direction and ground-level machinery matter more than yield.
Turbine parts (nacelle, gearbox, yaw drive, pitch system) are explained on our page on the basic components of wind energy, and lift, drag, angle of attack and tip speed ratio on the page on aerodynamic forces acting on windmill blades. This page sticks to how the two types compare and what drives their design.
Vertical axis vs horizontal axis wind turbine: comparison table
| Point | HAWT (three-bladed) | VAWT (Darrieus / Savonius) |
|---|---|---|
| Rotor axis | Horizontal, parallel to the wind | Vertical, perpendicular to the wind |
| Main force driving the rotor | Lift | Darrieus: lift. Savonius: drag |
| Typical peak Cp | 0.45 to 0.50 | Darrieus 0.30 to 0.40; Savonius 0.15 to 0.25 |
| Optimum tip speed ratio | About 6 to 8 | Darrieus about 3 to 6; Savonius about 0.8 to 1 |
| Rotor solidity | Low, about 0.05 to 0.10 | Darrieus 0.1 to 0.3; Savonius close to 1 |
| Facing the wind | Needs a yaw system or tail vane | Accepts wind from any direction |
| Self-starting | Yes, with blade pitch | Savonius yes; fixed-pitch Darrieus usually no |
| Torque | Nearly steady over a revolution | Pulsating in a Darrieus (torque ripple) |
| Gearbox and generator | In the nacelle at hub height | At or near ground level |
| Height above ground | Tall tower, rotor in stronger, smoother wind | Rotor usually low, in slower, turbulent wind |
| Largest sizes built | Offshore machines above 15 MW | A few MW at most, experimental |
| Best suited to | Wind farms, onshore and offshore | Small, urban, rooftop and research use |
Power available and the Betz limit
Both types extract power from the same stream of air, so the starting equation is shared:
P = 0.5 ρ A V³ Cp
where ρ is air density (1.225 kg/m³ at sea level, 15 °C), A is the swept area in m², V is the free wind speed in m/s, and Cp is the power coefficient. No rotor can exceed the Betz limit, Cp,max = 16/27 = 0.593, whatever its axis. The swept area is where the geometry differs:
- HAWT: A = πD²/4, a circle of rotor diameter D.
- Straight-bladed (H-type) Darrieus or Savonius: A = D x H, the rectangle of rotor diameter times blade height.
- Curved “eggbeater” Darrieus: roughly (2/3) x D x H for the usual troposkein shape.
Worked example: HAWT vs VAWT of similar swept area
Problem. Compare a small HAWT of 4.0 m rotor diameter with an H-type Darrieus of 2.5 m diameter and 5.0 m blade height, at a wind speed of 8 m/s and ρ = 1.225 kg/m³. Take Cp = 0.40 for the HAWT and 0.30 for the Darrieus, realistic figures for small machines.
Step 1, swept areas.
- HAWT: A = π x 4.0² / 4 = π x 16 / 4 = 12.57 m²
- Darrieus: A = 2.5 x 5.0 = 12.50 m²
Step 2, power in the wind. V³ = 8³ = 512 m³/s³.
- HAWT: 0.5 x 1.225 x 12.57 x 512 = 3941 W
- Darrieus: 0.5 x 1.225 x 12.50 x 512 = 3920 W
Units: kg/m³ x m² x m³/s³ = kg m²/s³ = W.
Step 3, rotor output.
- HAWT: 3941 x 0.40 = 1576 W, about 1.58 kW
- Darrieus: 3920 x 0.30 = 1176 W, about 1.18 kW
Same wind, nearly the same area, and the HAWT delivers about 34 per cent more shaft power. A Savonius of the same 12.5 m² at Cp = 0.18 would give only 3920 x 0.18 = 706 W. Generator and gearbox losses come off all three figures.
Design considerations for horizontal axis wind turbines
Rotor solidity and number of blades
Solidity is total blade area divided by swept area. A low-solidity, three-bladed rotor runs fast at low torque, which suits a generator. Three blades are the usual compromise: two-bladed rotors are cheaper but suffer uneven loads and wobble as they yaw, and adding a fourth blade adds cost with little extra Cp. High-solidity, many-bladed rotors belong to water-pumping windmills, which need high starting torque rather than high speed.
Tip speed ratio
A three-bladed HAWT is designed around a tip speed ratio of about 6 to 8. At 8 m/s and a ratio of 7, the tip moves at 56 m/s. Utility turbines limit tip speed to roughly 80 to 90 m/s onshore, mainly because blade noise and leading-edge erosion rise quickly with tip speed.
Yaw system
The rotor must face the wind. Small machines use a passive tail vane. Large machines use an active yaw drive, a slewing bearing with motors and brakes driven by a wind vane on the nacelle. A yaw error of angle θ cuts power roughly in proportion to cos²θ to cos³θ, so poor yaw tracking in shifting winds is a direct loss.
Tower height and wind shear
Wind speed rises with height. A common estimate over open ground is the power law V2 = V1(h2/h1)α with α about 1/7. Going from 10 m to 100 m multiplies wind speed by 101/7 = 1.39 and power by 1.39³ = 2.68. That single effect is a large part of why HAWTs have grown so tall, and it is a structural advantage a VAWT cannot easily match because its heavy parts sit at the bottom and tall VAWT rotors need guy wires.
Nacelle mass, gearbox and maintenance
The gearbox (or a direct-drive generator) and main bearing sit at hub height, so the tower carries a heavy top mass and major repairs need a crane. Pitch and yaw control add parts, but they let the turbine start itself, limit power in high winds and shut down safely.
Design considerations for vertical axis wind turbines
Darrieus (lift type) vs Savonius (drag type)
The Darrieus rotor uses aerofoil blades, either curved (eggbeater) or straight (H-rotor), and is driven by lift. It is the efficient VAWT, with a peak Cp of about 0.30 to 0.40 at a tip speed ratio of roughly 3 to 6. The Savonius rotor uses two or three scoops and is driven by drag. The blades can never move faster than the wind, so its tip speed ratio stays near 1 and its peak Cp is only about 0.15 to 0.25. What it offers in return is high starting torque, simplicity and quiet running.
Self-starting
A fixed-pitch Darrieus usually cannot start on its own. At low rotor speed the blades meet the air at angles well past stall and produce little net torque. Designers solve this with a small Savonius mounted on the same shaft, helical blades, variable pitch, or by motoring the generator to bring the rotor up to speed.
Torque ripple and cyclic fatigue
Each Darrieus blade sweeps through changing angles of attack as it goes round, producing most of its torque on the upwind and downwind passes and little at the sides. An N-bladed rotor therefore delivers torque that pulses N times per revolution, twice per revolution for the classic two-bladed machine. This torque ripple loads the shaft, gearbox and generator cyclically, and the blades bend back and forth every revolution. Over 108 or more cycles in a turbine’s life, fatigue of blade joints and struts has been a leading failure mode. Three blades or helical blades smooth the torque considerably.
Omnidirectional operation
No yaw system is needed, which saves parts and helps where wind direction swings quickly, such as around buildings. The catch is that wind near rooftops is slow and turbulent for any rotor, and the cube law makes low wind speed the bigger penalty.
Ground-level drivetrain and access
The generator and any gearbox can sit at ground level, which makes them lighter to support and easier to service. The catch is that the main bottom bearing carries the full rotor weight and the overturning load, and a tall rotor needs guy cables that take up land.
Noise and vibration
Lower tip speeds generally mean less aerodynamic noise than a HAWT of similar output, but vibration carried into a building can still be a problem on rooftops.
Horizontal and vertical axis wind turbines: which to choose
- Grid-scale wind farm, onshore or offshore: HAWT, without real competition. Higher Cp, taller towers and decades of refinement give the lowest cost per unit of energy.
- Farm, telecom tower or off-grid home in open country: a small HAWT on a tall mast usually yields more.
- Rooftop, street furniture or sites with rapidly changing wind direction: a VAWT, accepting lower output for no yaw, lower noise and easier access.
- Very low-cost, low-speed mechanical work such as ventilation or small pumps: a Savonius.
For current data on how wind power is being deployed, see the U.S. Department of Energy wind energy basics.
FAQs
Which is more efficient, HAWT or VAWT?
HAWTs are more efficient. A modern three-bladed HAWT reaches a peak power coefficient of about 0.45 to 0.50, against about 0.30 to 0.40 for a Darrieus VAWT and 0.15 to 0.25 for a Savonius. All are below the Betz limit of 16/27, or 0.593.
Why are vertical axis wind turbines not used in wind farms?
They produce less power from the same swept area, their rotors usually sit low in slower wind, and the Darrieus type suffers torque ripple and cyclic blade fatigue. They also do not scale up well, so HAWTs give a lower cost per unit of energy.
What is the main advantage of a VAWT over a HAWT?
A VAWT accepts wind from any direction, so it needs no yaw system, and its generator and gearbox can sit at ground level for easy maintenance. It is also generally quieter because of lower tip speeds.
Is a Savonius turbine lift or drag type?
A Savonius is a drag-type VAWT. Its scoops are pushed by the wind, so the tip speed ratio stays near 1 and the peak power coefficient is about 0.15 to 0.25. A Darrieus is the lift-type VAWT.
How much power does a 4 m HAWT produce at 8 m/s?
Swept area = pi x 4^2 / 4 = 12.57 m2. Power in the wind = 0.5 x 1.225 x 12.57 x 8^3 = 3941 W. At a power coefficient of 0.40, the rotor delivers about 1576 W, or 1.58 kW, before generator losses.