Yaw Bearing: How Wind Turbines Turn to Face the Wind

A horizontal-axis wind turbine works best when its rotor points straight into the wind. The wind does not stay put, so the whole nacelle, the housing at the top of the tower that carries the rotor, gearbox or generator and controls, has to turn. That turning motion is called yaw. The component that lets the nacelle rotate on top of the tower while carrying its enormous weight and the forces from the rotor is the yaw bearing.

Wikipedia describes the yaw bearing as a component of the yaw system of modern horizontal-axis wind turbines. It connects the nacelle to the tower, carries large static and dynamic loads, and allows the nacelle to turn to face the wind. The same source says it should resist corrosion and wear and should last the full service life of the turbine.

A yaw bearing is not glamorous. It moves slowly, occasionally, and through limited angles. Yet it sits in the load path for everything above the tower, and when it fails the usual fix involves a large crane. This article explains how yaw bearings developed from wooden windmill pivots, how the two main designs work, how they are controlled, how their life is calculated, what goes wrong, and why small pointing errors cost real energy.

Where the yaw bearing fits

The yaw system

The yaw bearing is one part of a larger yaw system. Wikipedia's article on yaw systems lists the main parts:

The yaw bearing must hold the weight of the nacelle and rotor and resist the bending moments created as the rotor extracts energy from the wind. Wikipedia notes that large modern turbines take several minutes to turn through a full 360 degrees, so yaw motion is deliberately slow.

Active and passive yaw

Wind turbines can be pointed into the wind in two broad ways.

• Active yaw. A drive system produces torque to turn the nacelle, based on signals from wind direction sensors or on manual commands. Wikipedia calls active yaw the state of the art for modern medium and large turbines.
• Passive yaw. The wind itself turns the turbine. On small turbines, the simplest form is a roller bearing combined with a tail fin. Wikipedia notes that passive designs must avoid fast yaw movements, which create gyroscopic loads on a spinning rotor, and that tail fins work for small turbines but cannot yaw large ones.

Some turbines with the rotor downwind of the tower can orient themselves without a tail fin, because the rotor acts like a weathervane. Wikipedia mentions the French company Vergnet as a maker of such downwind turbines.

Vertical-axis turbines, which accept wind from any direction, need no yaw system at all.

A short history: from windmill caps to slewing rings

Windmills

Turning a rotor to face the wind is centuries old. Wikipedia's yaw bearing and yaw system articles describe eighteenth-century windmills with rotatable tops, or caps, turned either by people or animals or by a small auxiliary wind rotor called a fantail. A fantail is mounted at right angles to the main sails. When the wind shifts, the fantail catches it and slowly turns the cap until the main sails face the wind again and the fantail stops.

The fantail is usually credited to Edmund Lee, a blacksmith at Brockmill Forge near Wigan in England, who patented it in 1745; the Mills Archive notes that his patent petition dates from 1744. According to Wikipedia's article on the windmill fantail, designs were refined on mills around Leeds and Hull towards the end of the eighteenth century. The principle is a neat example of passive feedback control. The fantail keeps turning until its own sails are edge-on to the wind and can no longer be driven, and at that point the main sails face the wind squarely and produce the most power. On tower mills, gearing from the fantail drives a toothed rack around the top of the tower to turn the cap. On post mills, where the whole body of the mill turns, the gearing drives wheels running on a track. Fantails became more common in England, Denmark and Germany than elsewhere in Europe, and were especially useful where the wind direction changes often.

That rack around the top of a tower mill is the direct ancestor of the gear ring cut into a modern yaw bearing, and the fantail's job is now done by a wind vane, a controller and electric yaw drives.

These mills sat on axial gliding bearings made of wooden blocks. According to Wikipedia, citing a Dutch book on mill building by A. Sipman, the blocks were lubricated with animal fat or lined with copper or brass. Worn blocks were simply replaced, and movable locks held the cap in position. A later "hybrid" arrangement combined removable radial gliding pads with an axial roller bearing.

Modern turbines

The modern yaw bearing descends directly from these ideas. Today's two main families, rolling element bearings and gliding (sliding) bearings, are high-precision versions of the roller and wooden block solutions on old mills. What has changed is scale and precision. A modern multi-megawatt turbine's yaw bearing is several metres across, carries the full weight of the nacelle and rotor at the top of a tall tower, and has to survive a design life of decades under constantly varying loads.

Rolling yaw bearings

Design

The first main type uses rolling elements, balls or rollers, running between inner and outer rings, like a scaled-up version of an ordinary ball bearing. Wikipedia describes it as a large-diameter bearing, usually of the four-point type. A four-point contact ball bearing has raceways shaped so that each ball touches the rings at up to four points, which lets a single row of balls carry axial force, radial force and tilting moment together. For large loads, designers use two rows.

The German bearing maker thyssenkrupp rothe erde says that single-row or double-row four-point contact bearings with an integrated internal or external gear are particularly suitable for nacelle yaw applications. The gear teeth are machined directly into one of the rings, so the yaw drive pinions can mesh with the bearing itself.

Size

Manufacturers rarely publish yaw-bearing dimensions for specific turbines, but catalogues give an idea of scale. Liebherr's catalogue lists double-row four-point slewing bearings with external diameters from 400 to 6,000 millimetres, a general industrial range not limited to wind. For comparison, thyssenkrupp rothe erde states that its wind turbine rotor bearings reach diameters of up to 7 metres and its blade bearings up to 5 metres. Yaw bearings for large turbines are in the same multi-metre class, sized to match the diameter of the tower top.

Drives and brakes

Because rolling bearings have low friction, the nacelle turns easily. Wikipedia points out two consequences. First, smaller yaw drives are enough. Second, a separate yaw brake is needed, because without one the nacelle would be pushed around by the wind and rotor forces.

Wikipedia gives a typical arrangement as about six smaller yaw drives spaced around the gear ring. Spreading the torque over several drives makes each one smaller and easier to replace. Wikipedia also describes a clever control trick: half of the drives push one way and half the other. This removes backlash, the small play between gear teeth, and lets the drives hold the nacelle in place using their motor brakes.

According to the yaw system article, each electric yaw drive consists of an AC motor and a large gearbox. The largest have a maximum static torque of about 200,000 newton metres and gear ratios of about 2,000 to 1. Hydraulic alternatives use hydraulic motors or ratcheting cylinders. For braking, a hydraulic disc brake is most common, with calipers gripping a brake disc ring. Electric calipers are used as a compromise, and pneumatic brakes working at 6 to 10 bar are described as experimental.

Gliding (sliding) yaw bearings

Design

The second type replaces rolling elements with sliding pads. Wikipedia describes gliding yaw bearings as dry or lubricated pads on axial and radial surfaces that slide against a large steel disc, often combined with a gear rim. Some pads sit on top of the disc and carry the weight of the nacelle and rotor. Others sit underneath to resist the overturning moment that tries to lift one side. Radial pads around the edge keep the nacelle centred.

The pads are held in assemblies often called calipers or clamps. The plastics maker Mitsubishi Chemical Advanced Materials (MCAM) describes the arrangement as a yawing ring and several calipers that together act as a large slewing bearing.

Built-in braking

Gliding bearings have a natural advantage: friction. Wikipedia explains that pre-tensioning elements, which may be springs, pneumatic units or hydraulic units, press the pads against the disc. If the pressing force can be controlled, as with pneumatic or hydraulic elements, the same system can act as a brake. The friction damps unwanted yaw motion, and a designer can trade friction against the torque the yaw drives must provide.

Pad materials

Wikipedia lists polymers such as polyoxymethylene (POM) and polyamide (PA) as common pad materials, and mentions that some designs use self-lubricating materials such as PTFE. A European patent application on yaw bearing units notes that pads may be made of thermoplastics including polyester, polyurethane, polyamide or polyethylene terephthalate (PETP), which avoids metal-to-metal contact with the flange.

MCAM markets a PET-P grade, Ertalyte SLP, which it says was developed mainly for passive yaw slide pads in wind turbines. The company claims cost savings of up to 20 percent compared with roller bearing concepts, and states that its pads typically carry a minimum life expectation of 10 years, preferably the whole turbine life of 25 to 40 years. These are manufacturer's claims, not independent test results.

Adjustment and maintenance

Even wear depends on even pressure. Wikipedia explains that pad pressure is set with adjustment screws, pressure plates and springs, so load is spread evenly and uneven wear is avoided.

Gliding bearings are often designed with maintenance in mind. Wikipedia notes that segmented designs let individual pad sections be replaced without taking the whole bearing apart. The top axial surface, which carries the weight of the nacelle and rotor, is harder: replacing those pads usually requires lifting the nacelle slightly with a crane or jacks. A United States patent, US 10954997, describes a method of replacing worn pads by yawing the nacelle so that the worn pads are unloaded. Another patent record notes that spreading high loads over many pads, and moving the nacelle so the same pads are not always the most heavily loaded, reduces wear.

Rolling versus gliding: a comparison

Neither design is universally better. The choice depends on the turbine maker's design philosophy, the loads, maintenance strategy and cost.

Loads and life: why yaw bearings are hard to calculate

Unusual operating conditions

Standard methods for calculating bearing life, such as ISO 281, assume a bearing that rotates continuously. Yaw and pitch bearings do not. They sit still most of the time under heavy load, then move slowly through small angles. They also carry a combination of axial force, radial force and large tilting moments that vary with the wind.

The US National Renewable Energy Laboratory (NREL) addressed this in its Wind Turbine Design Guideline DG03, "Yaw and Pitch Rolling Bearing Life," by T. Harris, J. H. Rumbarger and C. P. Butterfield, published in December 2009 as report NREL/TP-500-42362. The report gives design criteria, calculation methods and applicable standards for the performance and life of the ball and roller bearings that support yaw and pitch motion. NREL later published an updated version, report NREL/TP-5000-89161, incorporating research and field experience from 2009 to 2023. A 2023 NREL conference paper noted that these oscillating bearings do not have an applicable standardised approach for rating and life analysis.

How the methods compare

A 2020 paper in Wind Energy Science by Oliver Menck, Matthias Stammler and Florian Schleich of the Fraunhofer Institute for Wind Energy Systems (IWES) compared methods for the closely related problem of blade (pitch) bearings. They studied a double-row four-point contact ball bearing on a 7.5 megawatt reference turbine. The simplest method from DG03 uses only the applied tilting moment, radial force and axial force, combined into an equivalent load:

P = 0.75 × Fr + Fa + 2 × M / dm

where Fr is the radial force, Fa is the axial force, M is the tilting moment and dm is the pitch diameter of the bearing, the diameter of the circle through the centres of the balls.

The researchers found that this simple method predicted a life about 1.7 times longer than two more detailed methods based on individual ball loads. All the calculated lives were well below the 20-year design life of the turbine, which the authors attributed partly to standards built on much smaller bearings and to contact angles that can shift by 20 degrees or more under high load. Their study was on blade bearings, not yaw bearings, but the lesson carries over: simple formulas are useful for rough sizing, and detailed methods are needed for reliable life estimates of large, oscillating bearings.

Worked example: the tilting moment dominates

This example uses hypothetical round numbers to show the structure of the DG03 simplified formula. It is not data for any real turbine.

Assume a yaw bearing with:

• Pitch diameter dm = 4 metres
• Radial force Fr = 500 kilonewtons
• Axial force Fa = 3,000 kilonewtons (mostly weight)
• Tilting moment M = 20,000 kilonewton metres (20 meganewton metres)

Then:

• 0.75 × Fr = 0.75 × 500 = 375 kN
• Fa = 3,000 kN
• 2 × M / dm = 2 × 20,000 ÷ 4 = 10,000 kN

Equivalent load P = 375 + 3,000 + 10,000 = 13,375 kN.

In this hypothetical case, the tilting moment contributes about three quarters of the equivalent load (10,000 ÷ 13,375 ≈ 0.75). That matches the Fraunhofer finding, for blade bearings, that the resulting moment had the strongest influence on bearing load. It also shows why bearing diameter matters so much: doubling dm to 8 metres in this example would halve the moment term to 5,000 kN.

How a yaw system is controlled

Measuring the wind

Most turbines measure wind direction with a wind vane or similar sensor on the nacelle, behind the rotor. That location is convenient but imperfect. The rotor disturbs the flow. The industry inspection firm SkySpecs explains that rotating blades create a helix-shaped flow field in front of and behind the rotor, and that wakes from neighbouring turbines, topography, turbulence, gusts, wind shear and veer all make perfect alignment under all conditions extremely unlikely.

Deciding when to move

A yaw controller does not chase every gust. Moving too often wears the drives, brakes, gears and bearing and uses energy. Moving too rarely leaves the rotor misaligned. Controllers usually average the measured direction over a time window and only yaw when the average error exceeds a threshold, sometimes called a dead band.

A 2013 study by Brad Gaiser, published in Wind Systems magazine, used real data from an operating turbine fitted with an optical upwind measurement system and simulated different settings in MATLAB and Simulink. It modelled power below rated speed as varying with the cube of the cosine of the misalignment angle. With unlimited yaw activity, power increased by about 6 percent compared with legacy control. With yaw activity capped, the study recommended a moving average of 60 seconds and a threshold of 8 degrees, which gave an additional power gain of roughly 3 percent while the time spent yawing rose from about 5 percent to about 8 percent. It also found that averaging over too short a time made the controller react too early, before the measured wind actually reached the rotor, which reduced performance.

Worked example: how fast does a nacelle turn?

This is a hypothetical illustration based on the gear ratio described by Wikipedia.

Assume a yaw motor turning at 1,500 revolutions per minute and a yaw gearbox with a ratio of 2,000 to 1. The output pinion turns at 1,500 ÷ 2,000 = 0.75 revolutions per minute.

Now assume the pinion has one tenth as many teeth as the gear ring on the bearing, giving a further reduction of 10 to 1. The nacelle then turns at 0.75 ÷ 10 = 0.075 revolutions per minute.

One revolution is 360 degrees, so 0.075 × 360 = 27 degrees per minute, or 0.45 degrees per second. A full 360 degree turn would take 360 ÷ 27 ≈ 13.3 minutes. That is consistent with Wikipedia's statement that large turbines take several minutes to yaw all the way round, and it shows why yaw drives need such high gear ratios: the motor is small and fast, the nacelle is huge and slow.

Cable twist

Power and signal cables run from the nacelle down the tower. If the nacelle kept turning in the same direction, the cables would twist. Turbines therefore track how far the nacelle has turned from a neutral position and, past a limit, yaw it back to unwind the cables. A United States patent, US 10519929, titled "System and method for minimizing energy loss due to yaw untwist of a wind turbine," notes that power production is unnecessarily reduced during untwisting and describes a controller that checks the wind forecast before choosing when to untwist. Exact untwist limits are set by each manufacturer.

Why pointing accuracy matters

Yaw misalignment

Yaw misalignment is the angle between the direction the rotor faces and the direction the wind is actually coming from. SkySpecs distinguishes static misalignment, a constant offset often caused by a mis-set sensor or software, from dynamic misalignment, the lag between a wind shift and the turbine's response. Some dynamic misalignment is accepted deliberately to avoid excessive yaw drive loads.

How much energy is lost

Sources disagree on the exact mathematical relationship. SkySpecs states that power loss has been shown to be proportional to the square of the cosine of the misalignment angle. Gaiser's study used a cosine-cubed model. A 2020 paper in the journal Energies used cosine-cubed as a first approximation but described it as simplistic.

Worked example: cosine-squared versus cosine-cubed

Using a 10 degree misalignment, where cos(10°) ≈ 0.9848:

The cosine-squared figure matches SkySpecs' estimate that an extreme static misalignment of 10 degrees gives a power loss of about 3 percent. SkySpecs also points out that power loss does not translate directly to annual energy loss, because above a certain wind speed a turbine reaches its rated power even when slightly misaligned. It estimated the annual energy production loss at about 2.5 percent for 10 degrees. At 2 degrees, cos² gives about 0.9988, a loss of roughly 0.1 percent, which SkySpecs describes as probably not measurable.

The Energies paper gives a more pessimistic figure, saying an average yaw error of 10 degrees can cause a power loss of up to 10 percent, and that a zero-point shift of several degrees can affect yield by up to some percent of annual energy production. The spread between sources reflects real uncertainty about how rotors behave in yawed flow.

Finding misalignment in operating data

That Energies paper, "Wind Turbine Systematic Yaw Error: Operation Data Analysis Techniques for Detecting It and Assessing Its Performance Impact," by Davide Astolfi, Francesco Castellani, Matteo Becchetti, Andrea Lombardi and Ludovico Terzi, was published on 8 May 2020. Two of the authors worked for the Italian wind operator Renvico. The authors presented a method for detecting a systematic yaw error using only operational data that wind farms already collect, and applied it to identify a misaligned multi-megawatt turbine. After the error was corrected, they used operating data from the whole wind farm to assess the improvement at that turbine. They estimated that a systematic yaw error of 4 degrees had affected performance by up to about 1.5 percent of annual energy production.

This kind of analysis matters because a static yaw error is invisible to casual inspection. The turbine runs, reports a wind direction, and produces power. Only careful comparison with neighbouring turbines, power curves or an independent measurement such as lidar reveals that the rotor has been pointing slightly off all along. Common causes include a wind vane mounted or calibrated with an offset, a controller setting that was never corrected, or flow distortion behind the rotor that the controller's correction does not fully remove.

Worked example: what does 1.5 percent mean?

Take a hypothetical turbine that would produce 10,000 megawatt hours in a year when perfectly aligned.

• A loss of 1.5 percent is 10,000 × 0.015 = 150 megawatt hours per year.
• Over a hypothetical 20 year life, that is 150 × 20 = 3,000 megawatt hours.
• Across a hypothetical wind farm of 40 such turbines with the same error, the lifetime loss would be 3,000 × 40 = 120,000 megawatt hours.

The numbers are illustrative, but they show why operators invest in finding and fixing errors of only a few degrees.

Deliberate misalignment: wake steering

The idea

Not all yaw offsets are mistakes. A wind turbine leaves a wake of slower, more turbulent air behind it. In a wind farm, turbines downwind sit in the wakes of those upwind and produce less power. Wake steering deliberately yaws an upstream turbine slightly away from the wind. That turbine loses a little power, but its wake is deflected sideways, so a downstream turbine sees faster wind. If the downstream gain is larger than the upstream loss, the farm as a whole produces more.

Wake steering turns the yaw system from a pointing device into a farm-level control tool. It puts more demands on the yaw bearing and drives, because offsets are commanded deliberately and change with wind direction, and it requires the yaw system to hold accurate, intentional angles.

Field evidence

Field tests show both promise and complications.

• A 2017 paper in Wind Energy Science by Fleming and colleagues reported a wake steering test at an offshore wind farm in China. They found that the controller increased power capture by amounts similar to those predicted by models, within the certainty allowed by the data.
• A 2021 field experiment led by Bart Doekemeijer, published in Wind Energy Science, tested open-loop yaw-based wake steering at a commercial onshore wind farm near Sedini in Sardinia, Italy, owned and operated by Enel Green Power. The farm has 43 GE turbines, and the experiment used eight of them, mostly GE 1.5s models, with two upstream turbines yawed. Power production increased by up to 35 percent for two-turbine interactions and up to 16 percent for three-turbine interactions, but there were losses for some ranges of wind direction. The authors concluded that more research is needed before wake steering consistently produces energy gains.

The gap between models and field results links directly to yaw hardware and control. A yaw controller acts on wind direction measured in the recent past, and by the time the nacelle has moved, the wind may have shifted again. A yaw system that moves slowly, with averaging windows and dead bands designed to limit wear, cannot track a constantly shifting optimum perfectly. Wake steering therefore sharpens the old trade-off between accurate pointing and mechanical wear.

Field evidence from lidar

Measured results give a practical sense of the stakes. In a US Department of Energy supported study at NREL, a nacelle-mounted Wind Iris lidar from Avent was installed on NREL's two-bladed CART2 research turbine. Press coverage of the 2014 results reported an estimated 2.4 percent increase in annual energy production from correcting a static misalignment of 7.5 degrees. In a commercial example reported in the trade press, First Wind measured an average yaw error of seven degrees over 30 days, and after correction the turbine's annual energy production increased by 1.8 percent.

The yaw bearing itself does not cause misalignment, but it is what makes correction possible. A yaw system that is stiff, worn or poorly maintained can also be slow to respond, adding to dynamic misalignment.

Reliability and failures

How often yaw systems cause downtime

Yaw systems are a significant source of lost availability. Sandia National Laboratories' 2012 Wind Turbine Reliability Benchmark, from its Continuous Reliability Enhancement for Wind (CREW) database of US wind plants, ranked forced outages of the yaw system third among the top drivers of unavailability, accounting for 8.2 percent. Only unscheduled and scheduled maintenance of the turbine as a whole ranked higher. The same benchmark reported that turbines in the dataset spent 76.5 percent of the time generating.

A 2023 study in the Journal of Physics: Conference Series by Anderson and colleagues at the University of Strathclyde, using an offshore dataset, found the yaw system among the next most frequently failing assemblies after the frequency converter, cooling system, gearbox and blades. Its main point was that estimated failure rates ranged from below 1 to more than 10 failures per turbine per year depending on how a "failure" is defined, which is a caution against comparing numbers across studies.

A trade article in Energy Global on field experience in Japanese wind farms with severe wind conditions reported that yaw drive failures made up about 48 percent of yaw device failure occurrences and ring gear failures about 9 percent. That figure comes from a trade publication rather than a peer-reviewed study.

Typical problems

The sources point to several recurring issues:

• Yaw drive failures. Motors, gearboxes and pinions are the most frequently failing parts of yaw systems in the Japanese data.
• Gear ring damage. Teeth on the bearing's gear ring can wear, pit or break. Because the ring is part of the bearing, a damaged ring is expensive to replace.
• Raceway and pad wear. Rolling raceways can wear or fatigue under oscillating loads that standard methods do not model well. Gliding pads wear by design and must be adjusted and replaced.
• Uneven pad loading. Wikipedia emphasises correct pad pre-tension to avoid uneven wear.
• Sensor and controller faults. Wikipedia's yaw system article cites a condition monitoring case report titled "Yaw Bearing System Fault Detected," and SkySpecs notes that static misalignment can come from sensor or software settings.

Repairs and the crane problem

Replacing a yaw bearing is one of the largest jobs on a wind turbine, because the bearing sits under the whole nacelle. A patent application on yaw bearing replacement methods identifies the crane, labour and ground preparation as the main costs of the conventional approach, which is to lower the nacelle and rotor to the ground with a large crane.

Because of that, the industry has developed alternatives. The Danish company CNC Onsite repairs damaged yaw ring teeth in place, machining them up-tower without dismantling turbine parts, and a company executive has claimed this can save 30 to 50 percent of the cost of a repair. A Danish energy research programme has funded a project on new segments and service procedures for partial replacement of wear-damaged yaw rings. Gliding bearings with segmented, replaceable pads are themselves partly a response to the crane problem.

Design considerations

Pulling these threads together, designers choosing and sizing a yaw bearing must balance:

1. Load capacity. Weight, thrust and above all tilting moments, which grow quickly with rotor size.
2. Stiffness. The tower top and nacelle frame must support the bearing evenly, or loads concentrate on parts of the raceway or a few pads.
3. Friction. Low friction means smaller drives but a separate brake; higher friction gives built-in damping but larger drives.
4. Life under oscillation. Guidance such as NREL DG03 and more detailed methods are used, recognising that standard rotating-bearing formulas do not fit well.
5. Maintainability. Whether parts can be inspected, adjusted and replaced up-tower without a crane.
6. Corrosion. Especially offshore, where salt air attacks unprotected steel. Wikipedia names corrosion resistance as a basic requirement.
7. Control. Averaging times, thresholds and untwist strategies affect how often the bearing moves and how well the rotor tracks the wind.

Standard engineering texts listed by Wikipedia for further study include Wind Power Plants by R. Gasch and J. Twele and the Wind Energy Handbook by T. Burton and colleagues.

Common misconceptions

"The yaw bearing turns all the time"

It does not. The nacelle yaws only when the averaged wind direction has changed by more than a threshold, and slowly when it does. Most of the time the bearing is stationary under load, which is exactly why its life is hard to calculate.

"A few degrees off doesn't matter"

At 2 degrees the loss is tiny, but around 7 to 10 degrees, field studies and estimates put annual energy losses at roughly 2 to 2.5 percent, and some estimates higher. Over a fleet and a 20 year life, that is a lot of energy.

"Rolling bearings are always better than sliding ones"

Each has advantages. Rolling bearings turn easily and need smaller drives. Gliding bearings provide built-in damping and braking and have replaceable pads. Both are used in modern turbines.

"The yaw drive holds the nacelle still"

Sometimes it helps, for example with opposed drives using their motor brakes, but rolling-bearing designs generally rely on a dedicated yaw brake, and gliding designs rely on pad friction.

"Bearing life formulas from catalogues apply directly"

Standard formulas assume continuous rotation. Oscillating yaw and pitch bearings need specialised guidance, and researchers have found that different methods give quite different answers.

Frequently asked questions

What is a yaw bearing?

The large bearing between a wind turbine's tower and nacelle that carries the weight and forces of everything above it and lets the nacelle rotate to face the wind.

What are the two main types?

Rolling yaw bearings, usually large four-point contact ball bearings with an integrated gear ring, and gliding yaw bearings, in which polymer pads slide on a steel disc.

Why do wind turbines need to yaw at all?

A horizontal-axis rotor captures the most energy when it faces directly into the wind. Wind direction changes, so the rotor must follow it.

How many yaw drives does a turbine have?

It depends on the design. Wikipedia gives about six as a typical number for designs with rolling bearings, using several smaller drives instead of one large one.

How fast does a large turbine yaw?

Slowly. Wikipedia says a full 360 degree turn takes several minutes on large modern turbines.

What is yaw misalignment?

The angle between where the rotor points and where the wind is coming from. It can be static, from a sensor offset, or dynamic, from the lag in following wind changes.

How much energy does misalignment cost?

Estimates vary. At 10 degrees, SkySpecs estimates about 3 percent power loss and about 2.5 percent annual energy loss. Lidar studies have measured gains of 1.8 to 2.4 percent after correcting misalignments of around 7 to 7.5 degrees.

Why are yaw bearings hard to replace?

The bearing sits under the nacelle and rotor, so conventional replacement means removing them with a large crane. That is why up-tower repair methods and replaceable pad designs exist.

What are gliding pads made of?

Polymers such as POM, polyamide, PTFE-based materials and PET-P are used, according to Wikipedia, patents and material suppliers.

Do small wind turbines have yaw bearings?

Yes, but they are usually simple roller bearings combined with a tail fin for passive yaw, as Wikipedia describes.

Do vertical-axis turbines need yaw systems?

No. They accept wind from any direction.

Summary

The yaw bearing is the hinge between a wind turbine's tower and everything that makes power. It descends from the wooden, fat-greased gliding blocks of eighteenth-century windmills and survives today in two main forms: large four-point contact rolling bearings with integrated gear rings, turned by several opposed yaw drives and held by a separate brake, and gliding bearings in which pre-tensioned polymer pads slide on a steel disc and provide their own damping. Both must carry very large tilting moments while moving only occasionally and slowly, a duty that standard bearing-life methods do not capture well, which is why NREL developed its DG03 guideline. Yaw systems are a significant source of turbine downtime, and replacing a yaw bearing usually means a crane. The payoff for getting them right is accurate pointing: misalignments of a few degrees can cost a few percent of a turbine's annual energy.

Sources

• Yaw bearing (Wikipedia)
• Yaw system (Wikipedia)
• Windmill fantail (Wikipedia)
• Mills Archive: Lee's patent windmills 1744-1747
• Astolfi et al., Wind Turbine Systematic Yaw Error, Energies 2020 (MDPI)
• Fleming et al., Field test of wake steering at an offshore wind farm, Wind Energy Science 2017
• Doekemeijer et al., Field experiment for open-loop yaw-based wake steering at a commercial onshore wind farm in Italy, Wind Energy Science 2021
• Harris, Rumbarger and Butterfield, Wind Turbine Design Guideline DG03: Yaw and Pitch Rolling Bearing Life, NREL 2009 (OSTI)
• Wind Turbine Design Guideline DG03: Yaw and Pitch Bearings, updated edition (OSTI)
• Menck, Stammler and Schleich, Fatigue lifetime calculation of wind turbine blade bearings, Wind Energy Science 2020
• Sandia National Laboratories, Wind Turbine Reliability Benchmark, SAND2012-2961P (OSTI)
• Anderson et al., On the sensitivity of wind turbine failure rate estimates to failure definitions, 2023 (Strathprints)
• Energy Global: Improving turbine downtime through fail-safes
• Gaiser, Yaw Control Optimization, Wind Systems Magazine 2013
• SkySpecs: Wind Turbine Yaw Misalignment: Setting Some Things Straight
• Energies 2020 paper on yaw misalignment (University of Brescia repository)
• Renewable Energy Magazine: R&D project in US shows nacelle lidar reduces yaw misalignment (2014)
• NS Energy: First Wind selects Wind Iris lidar to correct yaw error
• thyssenkrupp: Wind turbines, our key components
• Liebherr slewing bearing product listing
• MCAM: Ertalyte SLP PET-P yaw slide bearing material
• EP3594495A1 Wind turbine yaw bearing (Google Patents)
• US10954997B2 Wind turbine (Google Patents)
• US10519929 System and method for minimizing energy loss due to yaw untwist (USPTO)
• Wind power generator and yaw bearing replacement method (Justia Patents)
• CNC Onsite uptower machining (WeatherGuard Wind podcast)
• Development of new segments and service procedure for partial replacements in wear damaged yaw rings (Energiforskning.dk)