1.4.2 Turn Insulation

The purpose of the turn insulation in both random- and form-wound stators is to prevent
shorts between the turns in a coil. If a turn short occurs, the shorted turn will appear as the
secondary winding of an autotransformer. If, for example, the winding has 100 turns between
the phase terminal and neutral (the “primary winding”), and if a dead short appears across
one turn (the “secondary”), then 100 times normal current will flow in the shorted turn. This
follows from the transformer law:
npIp = nsIs (1.1)
where n refers to the number of turns in the primary or secondary, and I is the current in the
primary or secondary. Consequently, a huge circulating current will flow in the faulted turn,
rapidly overheating it. Usually, this high current will be followed quickly by a ground fault
due to melted copper burning through any groundwall insulation. Clearly, effective turn insulation
is needed for long stator winding life.

|1.3.3|Form-Wound Stators—Roebel Bar Type, |1.4|STATOR WINDING INSULATION SYSTEM FEATURES, |1.4.1|Strand Insulation

1.3.3 Form-Wound Stators—Roebel Bar Type

In large generators, the more the power output, the larger and mechanically stiffer each coil usually is. In stators larger than about 50 MW, the form-wound coil is large enough that there are difficulties in inserting both legs of the coil in the narrow slots in the stator core without risking mechanical damage to the coil during the insertion process. Thus, most large generators today are not made from multiturn coils, but rather from “half-turn” coils, often referred to as Roebel bars. With a Roebel bar construction, only one half of a “coil” is inserted into
the slot at a time, which is considerably easier than inserting two sides of a coil in two slots simultaneously. With the Roebel bar approach, electrical connections to make the “coils” are needed at both ends of the bar (Figure 1.7).

 1.4 STATOR WINDING INSULATION SYSTEM FEATURES

The stator winding insulation system contains several different components and features, which together ensure that electrical shorts do not occur, that the heat from the conductor I2R losses are transmitted to a heat sink, and that the conductors do not vibrate in spite of the magnetic forces. The basic stator insulation system components are the:

• Strand (or subconductor) insulation
•  Turn insulation 
• Groundwall (or ground or earth) insulation

Figures 1.8 and 1.9 show cross sections of random-wound and form-wound coils in a stator slot, and identify the above components. Note that the form-wound stator has two coils per slot; this is typical. Figure 1.10 is a photograph of the cross section of a multiturn coil. In addition to the main insulation components, the insulation system sometimes has high-voltage stress-relief coatings and end-winding support components.
       The following sections describe the purpose of each of these components. The mechanical, thermal, electrical, and environmental stresses that the components are subjected to are also described.

1.4.1 Strand Insulation
In random-wound stators, the strand insulation can function as the turn insulation, although extra sleeving is sometimes applied to boost the turn insulation strength in key areas. Many form-wound machines employ separate strand and turn insulation.

Figure 1.7. Photo of a turbogenerator stator winding using Roebel bars.

The following mainly addresses the strand insulation in form-wound coils and bars. Strand insulation in randomwound machines will be discussed as turn insulation. Section 1.4.8 discusses strand insulation in its role as transposition insulation.
       There are both electrical and mechanical reasons for stranding a conductor in a formwound coil or bar. From a mechanical point of view, a conductor that is big enough to carry

Figure 1.8. Cross section of a random stator winding slot.

the current needed in the coil or bar for a large machine will have a relatively large cross-sectional area. That is, a large conductor cross section is needed to achieve the desired ampacity. Such a large conductor is difficult to bend and form into the required coil/bar shape. A conductor formed from smaller strands (also called subconductors) is easier to bend into the required shape than one large conductor.
       From an electrical point of view, there are reasons to make strands and insulate them from one another. It is well known from electromagnetic theory that if a copper conductor has a large enough cross-sectional area, the current will tend to flow on the periphery of the conductor. This is known as the skin effect. The skin effect gives rise to a skin depth through which most of the current flows. The skin depth of copper is 8.5 mm at 60 Hz. If the conductor has a cross section such that the thickness is greater than 8.5 mm, there is a tendency for the current not to flow through the center of the conductor, which implies that the current is not making use of all the available crossection. This is reflected as an effective AC resistance that is higher than the DC resistance. The higher AC resistance gives rise to a larger I2R loss than if the same cross section had been made from strands that are insulated from one another to prevent the skin effect from occurring. That is, by making the required cross section from strands that are insulated from one another, all the copper cross section is used for current flow, the skin effect is negated, and the losses are reduced.

       In addition, Eddy current losses occur in solid conductors of too large a cross section. In the slots, the main magnetic field is primarily radial, that is, perpendicular to the axial direction. There is also a small circumferential (slot leakage) flux that can induce eddy currents to flow. In the end-winding, an axial magnetic field is caused by the abrupt end of the rotor and stator core. This axial magnetic field can be substantial in synchronous machines that are under- excited. By Ampere’s Law, or the ‘right hand rule’, this axial magnetic field will tend to cause a current to circulate within the cross section of the conductor (Figure 1.11). The larger
the cross sectional area, the greater the magnetic flux that can be encircled by a path on the periphery of the conductor, and the larger the induced current. The result is a greater I2R loss from this circulating current. By reducing the size of the conductors, there is a reduction in stray magnetic field losses, improving efficiency.
       The electrical reasons for stranding require the strands to be insulated from one another. The voltage across the strands is less than a few tens of volts; therefore, the strand insulation can be very thin. The strand insulation is subject to damage during the coil manufacturing process, so it must have good mechanical properties. Since the strand insulation is immediately adjacent to the copper conductors that are carrying the main stator current, which produces the I2R loss, the strand insulation is exposed to the highest temperatures in the stator.

              (a)

(b)

Figure 1.9. Cross sections of slots containing (a) form-wound multiturn coils; (b) directly cooled
Roebel bars.

Figure 1.10. Cross-section of a multiturn coil, with three turns and three strands per turn.

Therefore, the strand insulation must have good thermal properties. Section 3.8 describes in detail the strand insulation materials that are in use. Although manufacturers ensure that strand shorts are not present in a new coil, they may occur during service due to thermal or mechanical aging (see Chapter 8). A few strand shorts in form-wound coils/bars will not cause winding failure, but will increase the stator winding losses and cause local temperature increases due to circulating currents

Figure 1.11. Side view of a generator showing the radial magnetic flux in the air gap and the bulging
flux at the core end, which results in an axial flux.

Section : [1.3], [1.3.1] , and [1.3.2]

1.3 TYPES OF STATOR WINDING CONSTRUCTION

Three basic types of stator winding structures are employed over the range from 1 kW to more than 1000 MW:

1. Random-wound stators
2. Form-wound stators using multiturn coil
3. Form-wound stators using Roebel bars

In general, random-wound stators are typically used for machines less than several hundred kW. Form-wound coil windings are used in most large motors and many generators rated up to 50 to 100 MVA. Roebel bar windings are used for large generators. Although each type of construction is described below, some machine manufacturers have made hybrids that do not fit easily into any of the above categories; these are not discussed in this blog.

1.3.1 Random-Wound Stators

Random-wound stators consist of round, insulated copper conductors (magnet wire or winding wire) that are wound continuously (by hand or by a winding machine) through slots in then stator core to form a coil (Figure 1.5). Figure 1.5 shows that most of the turns in the coils can be easily seen. Each turn (loop) of magnet wire could, in principle, be placed randomly against any other turn of magnet wire in the coil, independent of the voltage level of the turn, thus the term “random.” Since a turn that is connected to the phase terminal can be adjacent to a turn that is operating at low voltage (i.e., at the neutral point), random-wound stators usually operate at voltages less than 1000 V. This effectively limits random-wound stators to machines less than several hundred kW or HP.

1.3.2 Form-Wound Stators—Coil Type

Form-wound stators are usually intended for machines operating at 1000 V and above. Such windings are made from insulated coils that have been preformed prior to insertion in the slots in the stator core (Figure 1.6). The preformed coil consists of a continuous loop of magnet wire shaped into a coil (sometimes referred to as a diamond shape), with additional insulation applied over the coil loops.

Figure 1.5. Photograph of the end-winding and slots of a random-wound stator. (Courtesy TECOWestinghouse.)

(a)

(b)

 

 

Figure 1.6. (a) Photograph of a form-wound motor stator winding. (Courtesy TECO-Westinghouse.) (b)
A single form-wound coil being inserted into two slots.

Usually, each coil can have from two to 12 turns, and several coils are connected in series to create the proper number of poles and turns between the phase terminal and ground (or neutral); see Figure 1.4. Careful design and manufacture are used to ensure that each turn in a coil is adjacent to another turn with the smallest possible voltage difference. By minimizing the voltage between adjacent turns, thinner insulation can be used to separate the turns. For example, in a 4160 volt stator winding (2400 V line-to-ground), the winding may have 10 coils connected in series, with each coil consisting of 10 turns, yielding 100 turns between the phase terminal and neutral. The maximum voltage between adjacent turns is 24 V. In contrast, if the stator were of a random-wound type, there might be up to 2400 V between adjacent turns, since a phase-end turn may be adjacent to a neutral-end turn. This placement would require an unacceptably large magnet wire insulation thickness.

1.2.2 Insulated Rotor Windings and 1.2.3 Squirrel Cage Induction Motor Rotor Windings

1.2.2 Insulated Rotor Windings
In many ways, the rotor winding has the same components as the stator, but with important changes. In all cases, copper, copper alloy, or aluminum conductors are present to act as a conduit for current flow. However, the steady-state current flowing through the rotor winding is usually DC (in synchronous machines), or very low frequency AC (a few Hz) in induction machines. This lower frequency makes the need for a laminated stator core less critical.
       The conductors in rotor windings are often embedded in the laminated steel core or surround laminated magnetic steel. However, round rotors in large turbogenerator and highspeed salient pole machines are usually made from forged magnetic steel, since laminated magnetic steel rotors cannot tolerate the high centrifugal forces.
       Synchronous machine rotor windings, as well as wound rotor induction motors, contain electrical insulation to prevent short circuits between adjacent conductors or to the rotor body. As will be discussed in Chapters 3 and 5, the insulating materials used in rotor windings are largely composites of organic and inorganic materials, and thus have poor thermal and mechanical properties compared to copper, aluminum, or steel. The insulation then often determines the expected life of a rotor winding.

1.2.3 Squirrel Cage Induction Motor Rotor Windings

SCI rotor windings are unique in that they usually have no explicit electrical insulation on the rotor conductors. Instead, the copper, copper alloy, or aluminum conductors are directly installed in slots in the laminated steel rotor core. (Smaller SCI rotors may have the aluminum conductors cast in place.) In normal operation, there are only a few volts induced on the rotor conductors, and the conductivity of the conductors is much higher than that of the steel core. Because the current normally only flows in the conductors, electrical insulation is not needed to force the current to flow in the right paths. Reference 1.9 describes the practical aspects of
rotor design and operation in considerable detail.
       The only time that significant voltage can appear on the rotor conductors is during motor starting. This is also the time that extremely heavy currents will flow in the rotor windings. Under some conditions during starting, the conductors make and break contact with the rotor core, leading to sparking. This is normally easily tolerated. However, some SCI motors operate in a flammable environment, and this rotor sparking may ignite an explosion. Therefore, some motor manufacturers do insulate the conductors from the rotor core to prevent the sparking [1.10]. Since such applications are rare, for the purposes of this book, we assume
that the rotor is not insulated.
       Although SCI rotor windings are generally not insulated, for completeness, Section 9.4 does discuss such rotors, and Chapters 12 and 13 present some common tests for SCI rotor winding integrity.

1.2 PURPOSE OF WINDINGS & 1.2.1 Stator Winding

1.2 PURPOSE OF WINDINGS

The stator winding and rotor winding consist of several components, each with their own function. Furthermore, different types of machines have different components. Stator and rotor windings are discussed separately below.

1.2.1 Stator Winding

The three main components in a stator are the copper conductors (although aluminum is sometimes used), the stator core, and the insulation. The copper is a conduit for the stator winding current. In a generator, the stator output current is induced to flow in the copper conductors as a reaction to the rotating magnetic field from the rotor. In a motor, a current is introduced into the stator, creating a rotating magnetic field that forces the rotor to move. The copper conductors must have a cross section large enough to carry all the current required without overheating.
       Figure 1.4 is the circuit diagram of a typical three-phase motor or generator stator winding. The diagram shows that each phase has one or more parallel paths for current flow. Multiple parallels are often necessary since a copper cross section large enough to carry the entire phase current may result in an uneconomic stator slot size. Each parallel consists of a number of coils connected in series. For most motors and small generators, each coil consists of a number of turns of copper conductors formed into a loop. The rationale for selecting the number of parallels, the number of coils in series, and the number of turns per coil in any particular machine is beyond the scope of this book. The reader is referred to any book on motors and generators, for example references 1.1 to 1.3.

Figure 1.4. Schematic diagram for a three-phase, Y-connected stator or winding, with two parallel circuits
per phase.

       The stator core in a generator concentrates the magnetic field from the rotor on the copper conductors in the coils. The stator core consists of thin sheets of magnetic steel (referred to as laminations). The magnetic steel acts as a low-reluctance (low magnetic impedance) path for the magnetic fields from the rotor to the stator, or vice versa for a motor. The steel core also prevents most of the stator winding magnetic field from escaping the ends of the stator core, which would cause currents to flow in adjacent conductive material. Chapter 6 contains more information on cores.
       The final major component of a stator winding is the electrical insulation. Unlike copper conductors and magnetic steel, which are active components in making a motor or generator function, the insulation is passive. That is, it does not help to produce a magnetic field or guide its path. Generator and motor designers would like nothing better than to eliminate the electrical insulation, since the insulation increases machine size and cost, and reduces efficiency, without helping to create any torque or current [1.8]. Insulation is “overhead,” with a primary purpose of preventing short circuits between the conductors or to ground. However, without the insulation, copper conductors would come in contact with one another or with the grounded stator core, causing the current to flow in undesired paths and preventing the proper operation of the machine. In addition, indirectly cooled machines require the insulation to be a thermal conductor, so that the copper conductors do not overheat. The insulation system must also hold the copper conductors tightly in place to prevent movement.
       As will be discussed at length in Chapters 3 and 4, the stator winding insulation system contains organic materials as a primary constituent. In general, organic materials soften at a much lower temperature and have a much lower mechanical strength than copper or steel. Thus, the life of a stator winding is limited most often by the electrical insulation rather than by the conductors or the steel core. Furthermore, stator winding maintenance and testing almost always refers to testing and maintenance of the electrical insulation. Section 1.3 will describe the different components of the stator winding insulation system and their purposes.

1.1.3 Classification by Cooling

Another important means of classifying machines is by the type of cooling medium they use: water, air, and/or hydrogen gas. One of the main heat sources in electrical machines is the DC or AC current flowing through the stator and rotor windings. These are usually called I2R losses, since the heat generated is proportional to the current squared times the resistance of the conductors (almost always copper in stator windings, but sometimes aluminum in SCI rotors). There are other sources of heat: magnetic core losses, windage losses, and eddy current losses. All these losses cause the temperature of the windings to rise. Unless this heat is removed, the winding insulation deteriorates and the machine fails due to a short circuit.

Indirect Air Cooling. Motors and modern generators rated less than about 100 MVA are almost always cooled by air flowing over the rotor and stator. This is called indirect cooling since the winding conductors are not directly in contact with the cooling air due to the presence of electrical insulation on the windings. The air itself may be continuously drawn in from the environment, that is, not recirculated. Such machines are termed open-ventilated, although there may be some effort to prevent particulates (sand, coal dust, pollution, etc.) and/or moisture from entering the machine using filtering and indirect paths for drawing in the air. These open-ventilated machines are referred to as weather-protected or WP.
       A second means of obtaining cool air is to totally enclose the machine and recirculate air via a heat exchanger. This is often needed for motors that are exposed to the elements. The recirculated air is most often cooled by an air-to-water heat exchanger in large machines, or cooled by the outside air via radiating metal fins in small motors or a tube-type cooler in large ones. Either a separate blower motor or a fan mounted on the motor shaft circulates the air. IEC and NEMA standards describe the various types of cooling methods in detail [1.4, 1.5].
       Although old, small generators may be open-ventilated, the vast majority of hydrogenerators and turbogenerators (rated less than about 50 MVA) have recirculated air flowing through the machine. Virtually all hydrogenerators use recirculated air, with the air often cooled by air-to-water heat exchangers. For turbogenerators rated up to a few hundred megawatts, recirculated air is now the most common form of cooling.

Indirect Hydrogen Cooling. Almost all large turbogenerators use recirculated hydrogen as the cooling gas. This is because the smaller and lighter hydrogen molecule results in a lower windage loss and better heat transfer than air. It is then cost effective to use hydrogen in spite of the extra expense involved, due to the few percent gain in efficiency. The dividing line for when to use hydrogen cooling is constantly changing. In the 1990s, there was a definite trend to reserve hydrogen cooling for machines rated more than 300 MVA, whereas in the past, hydrogen cooling was sometimes used on steam and gas turbine generators as small
as 50 MVA [1.6, 1.7].

Directly Cooled Windings. Generators are referred to as being indirectly or conventionally cooled if the windings are cooled by flowing air or hydrogen over the surface of the windings and through the core, where the heat created within the conductors must first pass through the insulation. Large generator stator and rotor windings are frequently “directly” cooled. In direct-cooled windings, water or hydrogen is passed internally through the conductors or through ducts immediately adjacent to the conductors. Direct water-cooled stator
windings pass very pure water through hollow copper conductors strands, or through stainless steel tubes immediately adjacent to the copper conductors. Since the cooling medium is directly in contact with the conductors, this very efficiently removes the heat developed by I2R losses. With indirectly cooled machines, the heat from the I2R losses must first be transmitted through the electrical insulation covering the conductors, which forms a significant thermal barrier. Although not quite as effective in removing heat, in direct hydrogen-cooled windings the hydrogen is allowed to flow within hollow copper tubes or stainless steel tubes,
just as in the water-cooled design. In both cases, special provisions must be taken to ensure that the direct water or hydrogen cooling does not introduce electrical insulation problems.
See Sections 1.4 and 8.13.
       Direct water cooling of hydrogenerator stator windings is applied to machines larger than about 500 MW. There are no direct hydrogen-cooled hydrogenerators. In the 1950s, turbogenerators as small as 100–150 MVA had direct hydrogen or direct water stator cooling. Modern turbogenerators normally only use direct cooling if they are larger than about 200 MVA.
       Direct cooling of rotor windings in turbogenerators is common whenever hydrogen is present, or in air-cooled turbogenerators rated more than about 50 MVA. With the exception of machines made by ASEA, only the very largest turbo and hydrogenerators use direct water cooling of the rotor.

1.1.2 Synchronous Generators

Although induction generators do exist, particularly in wind turbine generators, they are relatively rare compared to synchronous generators. Virtually all generators used by electrical
utilities are of the synchronous type. In synchronous generators, DC current flows through the rotor (field) winding, which creates a magnetic field from the rotor. At the same time, the rotor is spun by a steam turbine (using fossil or nuclear fuel), gas turbine, diesel engine or, a hydroelectric turbine. The spinning DC field from the rotor induces current to flow in the stator (armature) winding. As for motors, the following types of synchronous generators are determined by the design of the rotor, which is primarily a function of the speed of the driving
turbine.

Round Rotor Generators (Figure 1.2). Also known as cylindrical rotor machines, round
rotors are most common in high-speed machines, that is, machines in which the rotor revolves
at about 1000 rpm or more. Where the electrical system operates at 60 Hz, the rotor
speed is usually either 1800 rpm or 3600 rpm. The relatively smooth surface of the rotor reduces “windage” losses, that is, the energy lost to moving the air (or other gas) around in the air gap between the rotor and the stator—the fan effect. This loss can be substantial at high speeds in the presence of protuberances from the rotor surface. The smooth cylindrical shape also lends itself to a more robust structure under the high centrifugal forces that occur in high-speed machines. Round rotor generators, sometimes called “turbogenerators,” are usually driven by steam turbines or gas turbines (jet engines). Turbogenerators using round rotors have been made in excess of 1500 MW. (1000 MW is a typical load for a city of 500,000
people in an industrialized country).

Figure 1.2. Phototgraph of a small round rotor. The retaining rings are at each end of the rotor body

Such a machine may be 10 m in length and about 5 m in
diameter, with a rotor on the order of 1.5 m in diameter. Such large generators almost always have a horizontally mounted rotor and are hydrogen-cooled (see Section 1.1.3).

Salient Pole Generators (Figure 1.3). Salient pole rotors usually have individual magnetic field poles that are mounted on a rim, with the rim in turn fastened to the rotor shaft by a “spider”—a set of spokes. Since the magnetic field poles protrude from the rim with spaces between the poles, the salient pole rotor creates considerable air turbulence in the air gap between the rotor and the stator as the rotor rotates, resulting in a relatively high windage loss.
However, since the rotational speed is usually significantly less than 1000 rpm, the loss is
considered moderate. Salient pole machines typically are used with hydraulic turbines, which have a relatively low rpm (the higher the penstock, i.e., the larger the fall of the water, the faster the speed). To generate 50 or 60 Hz current in the stator, a large number of field poles are needed (recall that the generated AC frequency is the number of pole pairs times the rotor speed in revolutions per second). Fifty pole pairs are not uncommon on a hydrogenerator,
compared to one or two pole pairs on a turbogenerator. Such a large number of pole pairs requires
a large rotor diameter in order to mount all the poles. Hydrogenerators have been
made up to about 800 MW. The rotor in a large hydrogenerator is almost always vertically
mounted, and may be more than 10 m in diameter.

Pump/Storage Generator. This is a special type of salient pole machine. It is used to
pump water into an upper reservoir during times of low electricity demand. Then, at times of high demand for electricity, the water is allowed to flow from the upper reservoir to the lower reservoir, where the machine operates in reverse as a generator.

chine from the pump to generate mode is commonly accomplished by changing the connections on the machine’s stator winding to reverse rotor direction. In a few cases, the pitch of the hydraulic turbine blades is changed. In the pump motor mode, the rotor can come up to speed by using a SCI-type winding on the rotor (referred to as an amortisseur or damper winding), resulting in a large inrush current, or by using a “pony” motor. If the former is used, the machine is often energized by an inverter-fed drive (IFD) that gradually increases the rotor speed by slowly increasing the AC frequency to the stator. Since the speed is typically less than a few hundred rpm, the rotor is of the salient pole type. Pump storage units have been made up to 500 MW.