Three-bladed turbines

Why do wind turbines almost all have exactly three blades when more blades would catch more wind?

Look at a wind farm and the uniformity is almost suspicious. Different manufacturers, countries, decades and sizes — and three blades, nearly every time. Meanwhile the old farm windpump on the same landscape has twenty.

The intuition says a rotor is a net: more blade, more wind caught. If that were true, three would be an odd place to stop. So either the intuition is wrong, or something else is being paid for.

Begin by fixing what a rotor can possibly do, because there is a ceiling and it is not a matter of blade count. A turbine extracts energy by slowing the air passing through the circle its blades sweep. Slow it too little and you take little. Slow it too much and the air refuses to come — a nearly stopped column downstream blocks the flow, which diverts around the rotor instead of through it. Between those failures there is an optimum, and the classical result is Betz's limit: no open-flow rotor of any design can take more than 16/27, about 59.3 per cent, of the kinetic energy in the wind crossing its swept area. Modern machines reach 45 to 50 per cent in practice. That cap applies to a one-bladed rotor and a hundred-bladed rotor alike — it is a statement about the air, not the hardware.

That already breaks the net intuition. The rotor is not catching wind with its blade area; it is imposing a retardation on a column of air of fixed diameter. So what does a blade do?

It sweeps. What matters is how much of the passing air gets worked on, which depends on how fast the blades move relative to the wind — the tip-speed ratio. A rotor with few blades can still interact with the whole flow provided each blade comes round often enough, so fewer blades means running faster and more blades running slower for the same effect. The twenty-bladed windpump sits at the slow, high-solidity end, developing large torque at low speed, which is what a piston pump wants. Generating electricity wants the opposite, since a slow shaft demands a heavier drivetrain.

Now the marginal question. Going from two blades to three does raise energy captured — typically quoted as a few per cent of annual output — and the fourth buys distinctly less again, because each added blade works in air already slowed by the one ahead of it. Meanwhile every blade costs money and adds mass at the top of a very tall tower. The curve of gains crosses the curve of costs early.

So why not two, where costs are lower still? Here is the part genuinely about three specifically, and it is not aerodynamic. Consider the rotor's resistance to being turned about the vertical tower axis as the machine yaws to follow a shift in wind direction. For a two-bladed rotor that resistance depends on where the blades happen to be: with them horizontal their mass is spread far from the yaw axis, and with them vertical it is not. So the rotor's inertia about that axis rises and falls twice per revolution, and yawing while spinning produces cyclic loading at that rhythm, felt through hub, shaft and tower.

Three equally spaced blades do not have this property. Three is the smallest number of identical, evenly spaced blades for which the rotor's mass distribution in its own plane is the same in every direction, so the inertia the yaw motion sees no longer depends on which way the rotor points — the cyclic excitation disappears rather than merely shrinking. Two-bladed machines are built, but they need a mitigation: usually a teetering hub, a hinge letting the rotor rock so the loads are absorbed rather than transmitted, which is a real mechanism with real cost and wear added to buy back what a third blade gives for free. Smaller penalties point the same way: fewer blades means higher tip speed, and aerodynamic noise rises very steeply with it.

The most common misstatement is to attribute the 59.3 per cent to blade count — as though three blades were the configuration closest to Betz, or as though more blades would push against it. Betz's limit is derived from momentum in the flow with no reference to blades at all, and bounds every design equally. Blade count is a decision made underneath that ceiling, about how cheaply and smoothly you approach it.

It is worth being precise, too, about what "more blades catch more wind" gets right. Adding a blade does raise the power coefficient slightly, and if energy capture were the only term more blades would win. The reason it does not is that the gain per blade shrinks quickly while the cost per blade does not shrink at all — and that the third blade uniquely also removes a dynamic problem. Three is where two different arguments agree, which is why the answer is so stable across manufacturers.

An honest caveat: three is an optimum given a set of prices and constraints, not a law. Two-bladed designs are periodically revisited offshore, where noise matters less and a rotor that can be raised as one piece is worth real money.

• How a teetering hub works, and why two-bladed offshore designs keep being reconsidered despite it.
• Why blade length has grown so much faster than tower count.

Every term the collection defines is gathered in the glossary.