Based on Faraday's law of electromagnetic induction, electrical energy is generated through the mechanical process of passing a conductor through a magnetic field. This process creates a potential difference (voltage) between the ends of the conductor. From Ohm's law, it is known that an electric current will flow when the ends of this conductor are connected to an electrical circuit with a resistance. The flow of current will interact with the magnetic field and act to produce a force resisting the motion of the conductor. In order to continue moving the conductor at the same speed, a mechanical force of the same magnitude must be provided and applied in the direction of movement. The result is that mechanical power (force × speed) is converted into electrical power (voltage × current).

An electrical generator performs this process of converting mechanical power into electrical power writ large. In doing so the generator produces torque (the result of force acting at some radius from the axis of the rotating machine) opposing the mechanical torque from the wind turbine rotor and drive train. From the electrical point of view, the generator is a voltage source (with some impedance) which depends on the type and design of the generator and on the operating conditions of the wind turbine and the generator.

In most electrical generators, it is convenient to induce electromotive force (emf) (and hence current) in stationary conductors, and hence they are located in a part of the machine which does not rotate (called the stator, as shown in Fig. 8.2(a)). The magnetic field is moved instead, and its source is mounted on a rotating part of the generator (called the rotor). The rotor and stator are separated by a clearance termed the ‘air gap’. Fig. 8.2(b) shows a generator and its rotor and stator.

In a synchronous generator the rotor magnetic field is produced either by using permanent magnets or else direct current is used to excite an electromagnet. As the excited rotor rotates it produces a rotating magnetic field in the air gap between the rotor and the stator. The stator contains a number of conductors that cut the magnetic field that is rotating in the air gap, in turn generating a current and hence electrical power output. Section 8.2 provides a more detailed explanation of the operation of the synchronous generator as well as detailing the operation of asynchronous generators.

As with many engineering systems, a number of these qualities complement one another and in some cases run counter to one another. For example, a more reliable generator may have lower maintenance costs but might be less efficient and more expensive to build. As a result of this, a balance must be struck to ensure a generator with the required qualities is chosen for a particular purpose. One approach for the turbine generator designer to balance these qualities and gauge the relative importance of them is to consider how the generator design affects the levelised cost of energy [1]. This approach will be used in Section 8.4.

Fig. 8.3 shows some of the large wind turbines which are currently in development or just starting to be manufactured. Generally, a good generator should be efficient, with as few losses as possible over a wide speed and torque range. Permanent magnet synchronous generators (PMGs) are usually better in this area than electrically excited machines as no current is needed to produce the rotor magnetic field and hence there are no copper losses on the generator rotor. Generators should also be reliablem with as few failure modes as possible and PMGs and squirrel cage induction machines show greater reliability due to their lack of slip rings and brushes, which are prone to failure and need regular servicing. However, this improved generator reliability has a consequence for the reliability of the power converter: there is a higher failure rate for the fully rated converters required for variable speed operation of PMGs and squirrel cage induction machines than for those used in doubly fed induction generators [2]. Onshore, failures to power converters tend not to lead to long downtime as they are relatively quick to repair. Offshore – where access to the turbine is more challenging and expensive – the consequence of even small failures can be costly. Manufacturing cost effectiveness and cost planning is also an important aspect for generator and wind turbine manufacturers. In this area PMGs are not as favourable as the electrically excited machines because the rare earth materials required for the permanent magnets are more expensive and cost volatile [3].

The generators in conventional plants tend to operate at a fixed speed and the mechanical input torque can be controlled relatively easily. Conventional plants tend to use synchronous generators which are usually directly connected to the grid. One result is that the rotational inertia of the turbine and generator naturally contributes to the stability of the electrical frequency of the electrical network. Because the excitation current can be varied, it is possible to control the power factor of the generator and provide or absorb reactive power as necessary. Although some earlier wind turbines were directly grid-connected, they used induction machines. This meant that the wind turbines were fixed speed rather than variable speed (the latter being preferable from a mechanical loading point of view) and that they consumed reactive power from the grid, this often having to be compensated for.

Another difference between conventional plants and the wind turbine is the location of the electrical generator. In conventional plants the generator is placed in well-controlled conditions, the generator can be easily accessed and the total generator mass is relatively unimportant. In wind turbines, the generator is almost always placed in a nacelle on top of the wind turbine tower. Obtaining the required amount of power for driving the generator requires large wind turbine rotor blades. Due to this large blade requirement, wind turbine generators are usually located on towers that can be more than 100 m above sea level. The fact that these generators are elevated to such heights means that mass can become a constraint as additional tonnes of material can affect the tower, foundation and installation costs.

Offshore wind has a higher cost of energy than onshore wind, and one of the contributory factors to this is the higher operations and maintenance costs. Newer offshore wind farms also tend to be placed further offshore than the first wind farms of the 1990s and 2000s. Moving further offshore often means sites have higher mean wind speeds and are subject to higher mean wave heights. As maintenance and repair vessels have wind speed and wave height operational limits, sites further offshore are less accessible than near shore, and near-shore sites are less accessible than onshore. These offshore access problems are leading wind turbine manufacturers to develop different types of generators for the offshore market compared to the onshore market. Manufacturers are focusing on making offshore generators more reliable, introducing redundancy and extending time between services to ensure offshore generators do not have to be visited as often as onshore generators [4].

Another aspect influencing the design of offshore generators is the cost of replacing them when they fail. Cranes can be used onshore, but heavy lift vessels – usually rented on daily rates of over £100,000 per day – are required offshore [5]. As a means of overcoming this, manufacturers are aiming to make their offshore generators modular, to allow them to be repaired and replaced without, or with reduced, use of heavy lift vessels.

Generally, offshore wind turbines have larger rotor diameters than those onshore. Partly this is to mitigate the expense of offshore foundations and cabling by increasing the power rating of each turbine and increasing each turbine's annual energy production. Reducing the rotor power density also leads to larger swept areas – this helps improve the turbine's capacity factors and utilisation of the transmission cabling. One advantage of offshore wind turbines is that there tend to be fewer objections from the local community than for onshore wind turbines. These objections can be sensitive to hub height, blade length and blade noise. This relaxation tends to lead to longer blades offshore as can be seen in Fig. 8.4. This larger blade size allows for greater mechanical torque to be produced offshore, resulting in generators offshore having a higher rated power. This is evident in the portfolio of major manufacturers who offer 7- and 8-MW turbines with rotor diameters up to 154 m (Siemens D7) and 164 m (Vestas V164) offshore but the largest onshore turbine is rated at 7.5 MW and has a rotor diameter of 126 m (Enercon E−126).

Rotating electrical machines can perform the roles of either generator or motor depending on the way they are controlled, the electrical connection and the nature of the mechanical interface. Fig. 8.5 shows this. There are two major categories of rotating electrical machines, those that use/produce alternating currents (AC) and those that use/produce direct current (DC). In conventional power systems – which are based on AC motors and generators – the majority of generators are synchronous machines and most motors are asynchronous machines. In wind turbines both synchronous and asynchronous machines (also known as induction machines) are used as generators.

The machine described in Section 8.1.1 is a primitive synchronous machine. In the case of the synchronous generator the rotor rotates at the same speed as the electromagnetic flux waveform in the air gap of the machine. For the wind turbine rotor to have a variable speed, the electrical frequency of the machine must also be variable in nature and this means that the generator stator must be connected to a power converter to decouple it from the grid frequency. The synchronous machine is described in more detail in Section 8.2.3.

Although similar in construction to the synchronous machine, induction machines have a different principle of operation. In these machines, an alternating current is present on the rotor windings, giving rise to a rotor magnetic field which rotates at a speed slower or faster than the rotor. This field is either induced from the electromagnetic field on the stator (ie, squirrel cage induction generators) or else set up by a current from a static power converter connected to the rotor windings through brush gear and slip rings (ie, doubly fed induction generators). The principle of operation of these induction machines is explained in the Section 8.2.4.

To find the induced voltage, e, (units V), in the conductor in Fig. 8.1, one can use Eq. [8.1],

In rotating electrical machines, one often thinks of coils (a combination of pairs of antiparallel conductors distributed around the stator of the machine so that they experience flux density of the same magnitude but opposing sign). For a coil, Faraday's law states that the induced voltage in a coil is proportional to the negative rate of change of magnetic flux. This is given in Eq. [8.2],

To find the force, f, (units N), applied to the current-carrying conductor in Fig. 8.6, one can use Eq. [8.3],

The manufacturing process of a generator rotor depends on the type of generator it is to be used in. Two of the most common types of rotors in wind energy are the permanent magnet rotor in a permanent magnet synchronous generator and the wound rotor in a DFIG.

As the name suggests the permanent magnet rotor is manufactured using permanent magnets, such as neodymium (NdFeB). These have the advantages of having high maximum energy product, meaning that they can produce a strong flux density using a relatively small mass of material. They have a high coercivity and remnant flux density but a relatively modest relative permeability see, e.g., [12]. This means that they are able to resist demagnetisation but that the magnets themselves are not very good at ‘conducting’ magnetic flux and care must be taken to design the magnets to get best use of the expensive magnet material. Fig. 8.7(c) shows these magnets mounted onto the surface of the rotor. Even though they appear to be salient pole machines, this low permeability means that the air gap reluctance is fairly constant as one travels circumferentially around the air gap radius. As such they can be treated as non-salient machines. Some machines, using less energy-dense permanent magnetic materials – such as Ferrite magnets – need to use flux concentration techniques, burying magnets between steel poles. In this case these machines are salient in nature.

Sometimes the magnets on the rotor are skewed relative to the teeth and slots on the stator. This helps reduce the effect of cogging torque. This effect, whereby the rotor magnets favour being in alignment with the teeth on the stator can add an unwanted ripple onto the machine torque and mean that there is a minimum starting torque to get the generator to rotate. Skewing means that the circumferential position of the magnet pole varies as one moves axially from one end of the machine to the other. Other techniques for reducing cogging torque also exist, including shaping the magnets in the radial, axial and circumferential dimensions.

The wound rotor in the DFIG is manufactured using copper wire wound around steel teeth in a cylindrical core. In a squirrel cage induction machine, electrically conducting bars are embedded into the same types of slots in the rotor steel. The core of the rotor is made from steel laminations. In a DFIG the shaft of the rotor is then connected to slip rings and brushes.

From a magnetic perspective, the air gap is important as it leads to reluctance which reduces the flux per pole. The magnetic or effective air gap is the notional dimension of air that gives rise to the same magnetic reluctance as the real machine. The effect of slotting (which tends to increase reluctance) can be factored in. The biggest difference between ‘real’ and ‘effective’ air gap values is found in surface-mounted magnet machines. Here the height of the magnet – a material which has almost the same permeability as air – is also included in the air gap. So, a 15-mm-high magnet and a physical air gap dimension of 5 mm leads to an effective magnetic air gap of 20 mm (if we neglect the effect of slotting). Reducing the physical air gap from 5 to 4 mm (−20%) only reduces the magnetic air gap from 20 to 19 mm, and hence reluctance reduces by only 5%.

The steel laminations are thin sections coated with a thin layer of insulating material. Laminations are used instead of solid steel because they reduce eddy current losses. Special electrical steel with silicon content is used which is designed to reduce hysteresis losses.

The generator frame and housing provide the structural support and protection for the internal workings of the generator. Heat can build up in a generator during the generation process leading to losses in efficiency, so cooling systems are required. Two of the basic types of cooling systems in wind energy generators are forced air and liquid cooling [13].

Low-speed machines in direct-drive wind turbines are much larger and so tend to be significantly integrated into the wind turbine structure. As the torque rating increases, so too does the air gap diameter and axial length. Due to the magnetic forces of attraction (between the ferromagnetic parts of the rotor and the stator), there tend to be large forces trying to close the air gap. This, along with the torque loading, can mean that these generators need considerable structures in order to give sufficient stiffness [13]. For large direct-drive machines, this mass can dominate the mass of the material used to produce electrical power [15].

Copper losses result from current flowing through the resistance of any windings (field or armature). This can be reduced by reducing resistance, either by increasing the cross-sectional area of conductors (generally increases the size of the machine), lowering the winding temperature (reduces the resistivity) or using materials which have lower resistivity. Copper is usually used as the conductor material and increasing current densities can be achieved – without significant additional losses – with improving cooling systems.

Iron loss is an important loss component in the components that conduct flux, which changes with time. They are sometimes described as ‘core losses’. The two loss mechanisms are hysteresis and eddy current losses. Both of these increase with increasing flux density in the teeth and back iron. Hysteresis losses are proportional to the electrical frequency and the eddy current losses are proportional to the square of the electrical frequency. The electrical steel used in the stator and rotor of induction machines is affected, whereas the rotors of synchronous machines are less affected as the main field is stationary relative to the rotor.

No slip ring or brushes are required for this generator type because the squirrel cage is excited through electromagnetic induction. This allows for increased reliability because slip rings and brushes make up ∼50% of the overall generator failure rate [2]. Unlike PMGs, the SCIG does not require any rare earth materials in its manufacture. This is advantageous for manufacturers as it allows for better cost planning. It is notable that Vestas also opted for the SCIG in the second version of their V112 3.0 MW [17] rather than the PMG as used in the first version of their V112 3.0 MW [18].

Slip rings and brushes are required for this generator type for power take off. The inclusion of these slip rings and brushes decreases reliability in comparison to the PMG [2], however a DFIG is cheaper than a PMG by up to ∼40% [19]. Like the SCIG and unlike PMGs, the DFIG does not require any rare earth materials in its manufacture. The use of the partially rated power converters with the DFIG allows for compliance to many grid codes, however the DFIG is not completely decoupled from the grid as in the PMG or SCIG, which both use fully rated converters.

As the input to the generator is from a single-stage gearbox the generator has a lower speed range than the previous turbine types, this means higher torque and a larger generator with more windings and insulation. As a result of these extra failure modes it is expected that lower-speed generators will have more winding failures than the high-speed generators in the previous two turbines but less than in an equivalent direct-drive generator [20].

The single rotor main bearing, single-stage planetary gearbox and the generator are packaged together in a compact single cast structure leading to a relatively lightweight solution.

As generator sizes increase, the cost of materials used in generator construction will naturally increase, however the big challenge in this area will come from the larger generators requiring more rare earth material for their permanent magnets. Rare earth materials used in permanent magnets – such as Dysprosium – are in short supply and are readily mined in only a few countries. This allows the supply and hence price to be manipulated. Price volatility – which generator manufacturers try to avoid – represents a risk which turbine designers would prefer to avoid.

Reliability of generators becomes even more important as turbines move further offshore and access becomes more challenging. Figs 8.24 and 8.25 show how the same technologies at different distances have lower availability and increased O&M costs as the distance to shore increases. This implies that future generators need to have lower failure rates and that the time and cost of repair need to be reduced.

One way to reduce the cost of energy for offshore wind farms is to increase the power rating and rotor diameter size. Assuming the same tip speed limit (eg, 90 m/s) and rated power per swept area (eg, 350 W/m²) figures, one can find that the rated torque grows from 3.75 MNm for a 5-MW turbine to 30.0 MNm for a 20-MW turbine. Especially for direct drive machines, this eightfold increase in torque for a fourfold increase in power means that – unless different technologies are used – the cost and size increase in the generator will rapidly outpace improvements in energy capture.

Brushless DFIGs are of interest to the wind energy industry because they would remove the brush and slip ring failure mode while keeping the lower failure rate of the partially rated converters [23]. The brushless DFIG operates through the use of a second stator winding – often termed the control winding – which fulfils the role of the brush gear and slip rings. The two stator windings are housed within the same generator frame but no direct coupling occurs between them.

Air-cored permanent magnet machines use air-cored windings in the stator of the machine. This allows for the elimination of ferromagnetic material in the stator which leads to the elimination of the force between rotor and stator described in Section 8.3.4. This in turn leads to advantages such as weight saving of up to 30–50%, making stator repairs easier and eliminating cogging torque [24].

As a means of making generators more maintainable and reliable, researchers and designers are designing for maintainability and fault tolerance. Generators and wind turbines that include modularity and redundancy are now being seen in academia and industry [25]. Like all reliability-related issues the business case for modularity and redundancy greatly improves as turbines move further offshore. Modularity allows for a failed module of the generator to be changed without requiring the replacement of the full generator. As well as the obvious cost advantage of replacing a module instead of a full unit, there are also huge advantages in the cost savings from vessel hire due to a module not requiring heavy lift vessels that may have long waiting times and have a day rate of over £100,000 [5]. Redundancy allows for generators to continue producing energy at a full or partial percent of the rated power even after a failure occurs. Although capital costs are usually higher due to the duplication of parts of the generator or the full generator, recent studies have shown that O&M savings may justify the higher capital costs [26].

High-temperature superconducting machines encountered in the literature are both synchronous and asynchronous. Superconducting machines replace some of the materials used in conventional generators with superconducting materials. Superconductors offer greater efficiency across all loads due to having no DC resistance. They also allow for far stronger magnetic fields to be produced in comparison to conventional generators. This leads to higher torque density and reduced size, attractive qualities especially for direct-drive turbines. One significant challenge is that the machines have to be cryogenically cooled to around 50 K. The reliability and maintenance intervals of these cryostats and generators is one of the biggest hurdles: it can take more than a day to cool a generator to its operating temperature [27].

Synchronous reluctance machines are another form of electrical machine that are known for their low cost as they avoid the use of permanent magnets. The synchronous reluctance machine makes use of the effect of saliency, the component of torque which is different between cylindrical and salient pole synchronous machines. The lack of brush gear slip rings is attractive and a few authors have started to look at these machines in some detail [28].

The hydraulic torque/speed converter in a wind turbine uses a hydraulic pump, accumulator and motor, which drives a generator at a constant speed. It is the modulation of these hydraulic units that gives the turbine its variable speed capability, rather than the control of the generator. The fact that the generator can be driven at a constant speed allows for the use of a constant-speed synchronous generator. In one of the wind turbine types that uses hydraulic torque/speed conversion, brushless synchronous generators are used (in this configuration, an auxiliary electrical machine is mounted on the shaft and is used to generate and regulate the DC) [31].

Electromagnetic torque conversion and power generation have also been developed in a single unit for a wind turbine. This configuration is being called pseudo direct drive [32]. The low-speed input is converted to a higher speed through three hollow cylinder rings placed inside each other. The outer and inner rings have permanent magnets arranged in alternating north–south patterns. The middle ring consists of steel segments that alter the magnetic field between the inner and outer ring. It is this ring which is connected to the low-speed shaft. Stator windings are then placed around the outside magnetic ring as seen in Fig. 8.27. This allows for a current to be induced in the windings as the magnetic outer ring rotates inside the stator.