29.1.3.1 Energy Saving

Electric VSD provides savings in two ways: (a) directly by consuming less energy and (b) indirectly by improving the product quality. The latter is often more difficult to quantify.

Direct energy saving is possible only with centrifugal loads such as centrifugal pumps and fans. Such loads are often run at fixed speeds. Traditionally, an automatic valve, or some other mechanical means, is used to vary fluid flow rates in pumps. However, if a VSD is used, then the motor speeds can be controlled electronically to obtain a desired flow rate and can result in significant energy savings.

On the basis of the laws of affinity for centrifugal loads,

• the volume of flow is directly proportional to speed,

• pressure is proportional to the square of the speed,

• input power is proportional to the cube of the speed.

The affinity law states that the power consumption is proportional to the cube of the motor speed. This implies that if the speed is halved, then the power consumption is reduced to one-eighth. So, energy savings occur as the requirement for volume decreases. If, for example, a cooling system calls for operation at 50% airflow volume, it requires only 12.5% of the power needed to run the system at 100% volume. Because power requirements decrease faster than the reduction in volume, there is a potential for significant energy reduction at lower volume.

Generally, centrifugal pumps and fans are sized to handle peak volume requirements that typically occur for short periods. As a result, centrifugal pumps and fans mostly operate at reduced volumes.

Opening or closing of a damper allows the airflow of fans to be controlled. Restricting the airflow causes the motor to work hard even with a low throughput.

With a VSD, the speed of the fan can be reduced, thus giving the opportunity to reduce energy consumption. Adjusting the speed of the motor regulates the airflow. The control can be achieved by monitoring humidity, temperature, flow, etc. The lower the required throughput, the greater the energy saved.

It has been estimated that the payback period of a 50 kW fan or pump VSD equipment operating 2000 h/year is 1.9 years for operation at 75% speed and 1.23 years for 50% speed. It has been assumed that the cost of the VSD is £5.5 k and the cost of power is £0.05/kW.

29.1.3.2 Improved Process Control

Using VSDs to improve process control results in more efficient operating systems. The throughput rates of most industrial processes are functions of many variables. For example, throughput in continuous metal annealing depends on, among other factors, the material characteristics, the cross-sectional area of the material being processed, and the temperature of one or more heat zones. If constant speed motors are used to run conveyors on the line, it must either run without material during the time required to change the temperature in a heat zone or produce scrap during this period. Both choices waste energy or material.

With VSDs, however, the time needed to change speed is significantly less than the time it takes to change heat-zone temperature. By adjusting the material flow continuously to match the heat-zone conditions, a production line can operate continuously. The results are less energy use and less scrap metal.

29.1.3.3 Reduced Mechanical Stress (Soft Starts)

Starting a motor on line power increases stress on the mechanical system, for example, belts and chains. Direct on-line start-up of an induction motor is always associated with high inrush current with poor power factor.

VSD can improve the operating conditions for a system by giving a smooth, controlled start and by saving some energy during starting and running. Smoother start-up operation will prolong life and reduce maintenance, but it is difficult to do more than make an estimate of the cost advantages of these. The benefits of soft start, inherent in VSD, are that it eliminates the uncontrolled inrush of current that occurs when stationary motor is connected to full line voltage, and also the inevitable suddenly applied high start-up torque. Benefits are that the power wasted by current inrush is eliminated and that the life of the motor and the driven machine are prolonged by the gentle, progressive application of torque.

29.1.3.4 Improved Electrical System Power Factors

When a diode supply bridge is used for rectification, electric VSDs operate at near-unity power factor over the whole speed range (the supply delivers mostly real power). When a fully controlled thyristor supply bridge is used (as in DC, cyclo, and current source drives), the power factor starts at around 0.9 at full speed and proportionately worsens as speed declines due to front-end thyristors (typically 0.45 at 50% speed and 0.2 at 25% speed).

Modern pulse-width modulated (PWM) drives convert the three phases AC line voltage to a fixed-level DC voltage. They do this regardless of inverter output speed and power. The PWM inverters, therefore, provide a constant power factor regardless of the power factor of the load machine and the controller installation configuration, for example, by adding a reactor or output filter between the VSD and the motor.

29.1.4.1 Acoustic Noise

In some installations, placing a VSD on a motor increases the motor's acoustic noise level. The noise occurs when the drive's nonsinusoidal (current and voltage) waveforms produce vibration in the motor's laminations. The nonsinusoidal current and voltage waveforms produced by the VSD are the result of the transistor switching frequency and modulation in the DC-to-AC inverter. The switching frequency, fixed or variable, determines the audible motor noise. In general, the higher the carrier frequency, the closer the output waveform is to a pure sine wave. One method of reducing audible motor noise is full-spectrum switching (random switching frequency). The VSD manufacturers accomplish full-spectrum switching by an algorithm within the VSD controller. The motor performance is optimized by evaluating motor characteristics, including motor current, voltage, and the desired output frequency. The resulting frequency band, though audible to humans, produces a family of tones across a wide frequency band. So, the perceived motor noise is considerably less than it would be with a single switching frequency.

Motor noise may not present a problem. Relevant factors include motor locations and the amount of noise produced by other equipment. Traditionally, motor noise level is reduced by adding an LC filter between the VSD and the motor, that is, reducing the high-frequency component of the motor voltage waveform. Modern PWM inverter drives run at very high switching frequency and with random switching frequency thus reducing the noise level too. Various methods have been proposed to reduce the magnetically generated noise, which is radiated from inverter-fed induction motors.

29.1.4.2 Motor Heating

Most motor manufacturers design their products according to NEMA standards to operate on utility-supplied power. Designers base their motors' heating characteristics and cooling methods on power supplied at fixed voltage and frequency.

For many drive applications, particularly those requiring relatively low power, inverters with a high switching speed can produce variable voltage and variable frequency with little significant harmonic content. With these, either standard or high-efficiency induction motors can be used with little or no motor derating. However, the inverters used in larger drives have limits on switching rate that cause their output voltages to contain substantial harmonics of orders 5, 7, 11, 13, and so on. These, in turn, cause harmonic currents and additional heating (copper and iron losses) in the stator and rotor windings. These harmonic currents are limited mainly by the leakage inductance. For simple six-step inverters, the additional power losses, particularly those in the rotor, may require derating of the motor by 10%–15%.

Existing constant speed drives often have an oversized induction motor. These can usually be converted to variable speed operation using the original induction motor. Most of the subsequent operation will be at lower load and lower loss than that for which the motor was designed.

Modern PWM-VSI drives produce a voltage wave with negligible lower-order harmonics. The wave consists of pulses formed by switching at relatively high frequency between the positive and negative sides of the DC link voltage supply. With larger motors that operate from AC supplies up to 6600 V, the rapid rate of change of the voltage applied to the winding may cause deterioration and failure in the insulation on the entry turns of standard motors.

On self-ventilated (fan-cooled) motors, reducing the motor shaft speed decreases the available cooling airflow. Operating a motor at full torque and reduced speed results in inadequate airflow. This consequently results in increased motor insulation temperature. This potentially can be damaging and can reduce the life of the motor's insulation or cause the motor to fail. One potential solution is to add a constant speed, separately driven cooling fan to the motor. This approach ensures adequate stator cooling over the whole speed range. However, the rotor will run hotter than designed as internal airflow remains a function of speed. As there are no windings in the rotor, insulation failure is not an issue, but bearings may run hotter and require more frequent lubrication.

Fan-cooled motors with centrifugal loads present less of a problem. Pumps and fans, for example, do not require full torque at reduced speeds. So, in these cases, there is less thermal stress on motors at reduced speeds. Centrifugal load does not cause the motor to exceed thermal limits defined by the insulation system.

29.1.4.3 Supply Harmonics

Current and voltage harmonics in the AC supply are created by VSD (as a nonlinear load) connected on the power distribution system. Such harmonics pollute the electric plant, which could cause problems if harmonic level increases beyond a certain level. The effect of harmonics can be overheating of transformers, cables, motors, generators, and capacitors connected to the same power supply with the devices generating the harmonics.

The IEEE 519 recommends practices and requirements for harmonic control in electric power systems. The philosophy of such regulations is to limit the harmonics injection from customers so that they will not cause unacceptable voltage distortion levels for normal system characteristics and to limit the overall total harmonic distortion of the system voltage supplied by the utility.

In order to reduce supply harmonics that are generated by VSDs equipped with a 6-pulse diode bridge rectifier, VSD equipment manufacturers adopt various techniques. Table 29.2 summarizes the most common methods and their advantages and disadvantages [2].

Ref. [2] quantifies the cost of these options as a percentage of the cost of a basic system with 6-pulse diode bridge. For low-power VSDs, the cost of a drive with a line reactor is estimated to be 120% of that without. A VSD with a 12-pulse diode bridge with a polygon transformer is 200% while for a double-wound transformer is 210%. The most expensive solution is that with active front end, estimated at 250%.

For 6-pulse converter, $\mathrm{n}6\mathrm{p}\pm 1$ (5, 7, 11, 13, 17, 19, etc) order harmonics are generated. To minimize the effects on the supply network, recommendations are laid down by IEEE 519 as to the acceptable harmonic limits. For higher drive powers, therefore, either harmonic filtering or use of a higher converter pulse number is necessary. It is generally true that the use of a higher pulse number is the cheaper alternative. Ref. [2] also quantifies the harmonic levels generated by each of the above method; refer to Table 29.3 for a direct comparison.

29.2.1.1 The Mining Industry

The majority of early-generation large mine winders are DC drives. Modern plants and retrofits generally employ cycloconverters with AC motors. However, small mine winders (below 1 MW) tend to remain DC.

The main requirements are the following:

• High reliability and availability

• Fully regenerative

• Small number requiring single-quadrant operation

• High range of speeds

• High starting torque required

• High torque required continuously during slow-speed running

• Low-torque ripple required

• Low supply harmonics

• Low audible noise emissions

• Flameproof packaging

29.2.1.2 The Marine Industry

The requirements of this industry are the following:

• Initial purchase price

• Reliability

• Ease of maintenance, that is, minimum component count and simple design

• Size and weight of equipment

• Transformerless, water-cooled VSD equipment is always preferred

Other desirable features include the following:

• A requirement for the integration of power management functions

• High volumetric power density (the smallest possible)

• Remote diagnostics, to allow faultfinding by experts onshore in critical situations

Drive powers are commonly in the range of 0.75–5.8 MW for thrusters and 6–24 MW for propulsion. The evolution in the commercial market is toward powers from 1 to 10 MW for propulsion. Higher powers are required for naval applications. The package drive efficiency must be equal to or better than 96%. Noise and harmonics problems are to be considered when using PWM inverters. The supply side harmonics produced must be capable of being filtered. Above 1 MW, power converters are usually equipped with a 12-pulse supply bridge, given today's technology.

Two-quadrant operation is required in general; hence, a diode supply bridge is adequate. Occasional requirement for crash stops forces the use of dynamic brake chopper. DC bus can be advantageous for supply to wharf loading equipment, but the drive power ranges are such that commercially available products already adequately serve this application.

The use of standard AC machines is desirable; however, if motors matched to the inverter prove to be cheaper, their use could be preferred. Low-noise emission (acoustic and electromagnetic) is very important. There is no requirement for high torque at low speed. Programming and expanded input and output capabilities are required to avoid the need for additional programmable logic control (PLC).

29.2.1.3 The Process Industries

The main requirements of this market are the following:

• Initial purchase price (long-term cost of ownership does not generally influence purchasing decision)

• Efficiency in continuous processes

• Reliability

• Ease of maintenance

• Bypass facility

The industry preference is for air-cooled drives. It is perceived that air-cooled drives are less costly than their water-cooled equivalents. Customers often have the belief that water and electricity does not mix well and are wary of problems with leaks. The exception is the offshore industry where equipment size is paramount, and therefore, water cooling is standard. In general, there is no perceived requirement for space saving in majority of process plants. The desirable features often requested by customers are ease of maintenance and good diagnostic facilities.

The market requirement is for cost-effective, stand-alone drives at various power levels from a fraction of a kW up to 30 MW. The use of standard AC machines is desirable. However, if nonstandard but simpler and cheaper machines can be offered, an advantage could be gained:

• Two-quadrant operation for fans, pumps, and compressors.

• Four-quadrant operation for some test benches.

• Control must allow additional functions such as temperature protection, motor bearing temperature, and flow and pressure control.

• There is no requirement, in general, for field weakening.

• The harmonics produced by the drive, imposed on the power system, should not require a harmonic filter. Harmonics must be minimized.

In the low-voltage (LV) arena, the PWM-VSI is dominating the market. In the medium-voltage (MV) arena, there are a number of viable drive solutions—load-commutated inverters (LCIs) and cycloconverters. However, there is a developing market for MV PWM-VSI drives.

29.2.1.4 The Metal Industries

The requirements of this industry are the following:

• Reliability—high availability.

• Efficiency of the equipment—long-term costs of ownership.

• Low maintenance costs—this has been a key factor in the move from DC to AC.

• Power supply system distortion—more onerous regulations from the supply authorities.

• Initial purchase cost—very competitive market and large drive costs have a big impact on total project costs.

• Confidence in the supplier and their solution.

The following is a list of desirable features:

• Programmable system drives with powerful programming tools.

• Preference for air-cooled stacks but water-cooled is acceptable if a water-to-air heat exchanger is used.

• Powerful maintenance and diagnostic tools.

• Low EMC noise signature.

• Ability to interface to existing automation system via network, fieldbus, or serial link.

• Physical size of equipment is often not an important consideration.

• Fire protection systems integral to drive equipment.

The main market concerns are (a) EMC regulations, (b) effects on motor insulation of higher voltage levels, and (c) cooling with “dirty” mill water is not acceptable. The maintenance of deionized water circuits is a big issue.

29.3.3.1 DC Static Converter

This drive employs the simplest static converter. It is easily configured to be a regenerative drive with a wide speed range. Table 29.8 summarizes its key features.

Table 29.8

Converter topologies

Converter	Schematic	Features
(a) Controlled rectifier		• DC motor • Fully controlled SCR converter • Controlled DC voltage • Simple converter topology • Power factor is a function of speed
(b) Cyclo		• Induction motor and synchronous motor • Direct AC/AC power conversion • 3×6-pulse SCR-based fully controlled converters—APT for fully regenerative • Natural commutation • Low supply harmonics, 18-pulse • Power factor is a function of speed
(c) Matrix		• Squirrel-cage induction motor • Synchronous motor • Direct AC/AC power conversion • Forced commutated, reverse conducting switches • Four-quadrant operation inherent PWM in/PWM out • Controlled power factor
(d) LCI		• Synchronous motor • Simple converter arrangement • Power factor is a function of speed • Load-commutated SCRs • Synchronous motor requires excitation • Suffer from torque pulsation at low speeds
(e) FCI		• Squirrel-cage induction motor • Similar to LCI • Requires output capacitors for commutation • Requires a diverter commutation circuit for commutation at low speeds • Torque pulsation and resonance
(f) VSI		• Synchronous and squirrel-cage induction motors • 6-pulse diode front end • Good power factor across speed range • DC link voltage source • PWM output voltage
(g) Kramer		• Wound-rotor induction motor with slip rings • Small energy recovery converter • Any type of converter may be used between slip ring and AC input

High torque is available throughout the speed range with excellent dynamic performance. Unfortunately, the motor requires regular maintenance, and the top speed often is a limiting factor. Commutator voltage is limited to around 1000 V, and this limits the maximum power available. The continuous stall torque rating is very limited due to the motor's commutator.

29.3.3.3 Current Source Inverter

The output of this inverter is rectangular blocks of current from the motor bridge supplied from a supply converter whose output is kept at constant current by a DC link reactor and current servo. This type of inverter is typically based on fast thyristors.

29.3.3.4 Slip Power Recovery (Kramer)

In this type of converter, which is described in Table 29.8, the rotor current of a slip-ring wound-rotor induction motor is rectified and the power then reconverted to AC at fixed frequency and fed back into the supply network. For traditional designs, the low-frequency slip-ring currents are rectified with a diode bridge, and the DC power is then inverted into AC power at mains frequency.

The traditional designs had poor AC mains dip immunity, high torque pulsation, and high levels of low-frequency AC supply harmonics. The latest generation of this type of drive is called the rotor drive and uses PWM-VSI inverters for the rotor and AC supply bridges.

This keeps sine-wave currents in the AC rotor circuits, and the drive has many advantages over traditional circuits including the following:

• No torque pulsation

• Low AC harmonics

• Very high immunity to AC supply dips

• Very cost-effective if a limited speed range is required but still requires a separate starter

• Inherent ability to run at rated speed without electronic circuits

• Converter cost reduced by 2:1 if it uses the ± speed ability to give a speed range

29.3.3.5 PWM-VSI Converter

The availability of power electronic switches with turn-off capability; for example, FETs, BJTs, IGBTs, and GTOs have currently favored drives with voltage-fed PWM converters on induction.

The PWM-VSI drives offer the highest possible performance of all VSDs; refer to Table 29.9. Recent improvements in switching technology and the use of microcontrollers have greatly advanced this type of drive. The inverters are now able to operate with an infinite speed range. The supply power factor is always near unity. Additional hardware is easily added if there is a requirement to regenerate power back into the mains supply. Motor ripple current is related to the switching frequency, and in large drives, the motor may be derated by <3%.

29.3.3.6 Comparison

Table 29.9 summarizes the main features of all types of converter drives discussed above and assesses their merits and drawbacks. It also illustrates typical applications.

29.4.2 Motor Drive Duty

29.4.2.1 Duty Cycle

The size of the driven motors is generally chosen for continuous operation at rated output, yet a considerable proportion of motor drives is used for duties other than continuous. As the output attainable under such deviating conditions may differ from the continuous rating, fairly accurate specification of the duty is an important prerequisite for proper planning. There is hardly a limit to the number of possible duty types.

In high-performance applications, such as traction and robotics, the load and speed demands vary with time. During acceleration of traction equipment, a higher start-up torque (typically twice the nominal torque) is required; this is usually followed by cruising and deceleration intervals. As the torque varies with time so does the motor current (and motor flux linkage level). The electric, magnetic, and thermal loading of the motor and the electric and thermal loading of the power electronics converter are definite constraints in a drive specification.

Table Table 29.11 categorizes operating duties into eight major types (Ref. [4]).

Table 29.11

Definition of load cyclic duties—VDE0530, in accordance with IEC 34-1

Duty type	Representation	Description
S1: Continuous running duty		Operation at constant load of sufficient duration for the thermal equilibrium to be reached. Specify by indicating “S1” and required output
S2: Short-time duty		Operation at constant load during a given time, less than required to reach thermal equilibrium, followed by a rest and deenergized period of sufficient duration to reestablish machine temperatures within 2°C of the coolant
S3: Intermittent periodic duty with a high start-up torque		A sequence of identical duty cycles, each including a period of operation at constant load and a rest and deenergized period. In this duty type, the cycle is such that the starting current does not significantly affect the temperature rise
S4: Intermittent periodic with a high start-up torque		A sequence of identical duty cycles, each cycle including a significant period of starting, a period of operation at constant load, and a rest and deenergized period
S5: Intermittent periodic duty with high start-up torque and electric braking		A sequence of identical cycles, each cycle consisting of a period of starting, a period of operation at constant load, a period of rapid electric braking, and a rest and deenergized period
S6: Continuous-operation periodic duty		A sequence of identical duty cycles, each cycle consisting of a period of operation at constant load and a period of operation at no load. There is no rest and deenergized period
S7: Continuous-operation periodic duty with high start-up torque and electric braking		A sequence of identical duty cycles, each cycle consisting of a period of starting, a period of operation at constant load, and a period of electric braking. There is no rest and deenergized period
S8: Continuous-operation periodic duty with related load/speed changes		A sequence of identical duty cycles, each cycle consisting of a period of operation at constant load corresponding to a predetermined speed of rotation, followed by one or more periods of operation at other constant loads corresponding to different speeds of rotation. There is no rest and deenergized period

29.4.2.2 Mean Output

Variation of the required motor output during the periods of loaded operation is among the most frequent deviations from the duty types defined in Table Table 29.11. In such cases, the load (defined as current or torque) is represented by the mean load. This represents the root-mean-square (RMS) value, calculated from the load versus time characteristics. The maximum torque must not exceed 80% of the breakdown torque of an induction motor.

If the ratio of the peak torque to the minimum power requirements is greater than 2:1, the error associated with using the RMS output becomes excessive, and the mean current has to be used instead. No such mean value approximation is possible with duty type S2, which therefore necessitates special inquiry.

Careful assessment of duty types S2–S8 reveals that there exist two distinct groups:

1. Duties S2, S3, and S6 permit uprating of motors relative to the output permissible in continuous running duty (S1).

2. Duties S4, S5, S7, and S8 requiring derating relative to the output permissible in continuous running duty (S1).

29.4.2.3 Thermal Cycling

The drive duty cycle also affects the reliability and the life expectancy of power devices. Repetitive load cyclic duty results in additional thermal stresses on power devices. Frequent acceleration and deceleration of drives results in repetitive junction temperature rise and falls at the cyclic duty. The life expectancy of devices is often determined by the maximum allowed number of cycles for a given power device junction temperature rise.

Although this is true for all types of power devices, it is more critical for IGBTs where wire bonds and solder layers are used.

In modern IGBT-based converter design, the maximum junction temperature rise of the IGBTs is limited to a level, which ensures a conservative number of thermal cycles over the lifetime of the drive. Typical junction temperature rise is 30°C for a repetitive cyclic duty (e.g., steel mill) and 40°C for nonrepetitive cyclic duty (e.g., fan pumps).

29.4.2.4 Multiquadrant Operation

Fully regenerative electric VSDs offer a rapid regenerative dynamic braking in both forward and reverse directions. Operation in motoring implies that torque and speed are in the same direction (QI, III). In regenerative braking, the torque is opposite to the speed direction (QII and IV), and the electric power flow in the motor is reversed. (See Fig. 29.2.)

Positive power flow of electric energy means that electric power is drawn from the power supply via the power electronics converter by the motor, while negative power flow refers to electric power delivered by the motor in the generator mode to the power electronics converter. This could be regenerated back to the supply or dissipated as a heat in the dynamic brake dissipative mechanism.

For regenerative drive, the power electronics converter has to be designed to be able to handle bidirectional power flow. In low and medium power converter (say <500 kW) with slow dynamic braking demands, the generated power during the braking period is interchanged with the strong filter capacitor of power electronics converter, or DC (dynamic) braking is used.

29.4.2.5 Dynamic Braking Energy

There exist two types of energy stored in VSD, which need to be dealt with during dynamic braking:

• Inertia or kinetic energy loads: Typically moving (rotating or linear) machines. These would decelerate naturally to rest. Braking can speed up the process cycle for the sake of productivity.

• Mass or potential energy loads: Typically hoists or lifts—which would run on or even accelerate. Braking must apply full power to maintain constant speed while the load is lowered.

The drive losses, mechanical resistance, and the transmission efficiency, work in favor of deceleration, reducing the braking power demand. The energy regenerated by potential energy loads depends on maximum power and both the overrun time and the decelerating time.

The braking time and the duty cycle time are decided by the requirements of the process system, but note particularly the effect of varying the braking duty cycle time and the deceleration time.

For DC injection braking, the kinetic energy of the motor-load system is converted to heat in the motor rotor. For fast and frequent generator braking, the power electronics converter has to handle the generated power either by a controlled dynamic brake chopper (with braking resistor) or through bidirectional power flow. The power losses in the converter can assist in dynamic braking.

For a fast speed response, modern VSDs may develop a maximum transient torque up to the base speed and maximum transient power up to a maximum speed, provided that both the motor and the power electronics converter can handle these powers. For a 200 kW dynamometer drive application, a rapid change of torque from full positive torque to full negative torque is required in <10 ms.

29.5.2 Induction Motor Drive

29.5.2.1 Squirrel Cage Induction Motor

Squirrel-cage induction motors are simpler in structure than DC motors and are most commonly used in the VSD industry. They are robust and reliable. They require little maintenance and are available at very competitive prices. They can be designed with totally enclosed motors to operate in dirty and explosive environments. Their initial cost is substantially less than that of commutator motors, and their efficiency is comparable. All these features make them attractive for use in industrial drives.

The three-stator windings develop a rotating magnetic flux rotating at synchronous speed. This speed depends on the motor pole number and supply frequency: the rotating flux intersects the rotor windings and induces an EMF in the rotor winding, which in turn results in circulating current. The rotor currents produce a second magnetic flux, which interacts with the stator flux to produce torque to accelerate the machine. As the rotor accelerates, the induced rotor voltage falls in magnitude and frequency until an equilibrium speed is reached. At this point, the induced rotor current is sufficient to produce the torque demanded by the load. The rotor speed is slightly lower than the synchronous speed by the slip frequency, typically 3%.

In order to ensure constant excitation of the machine, and to maximize torque production up to the base speed, the ratio of stator voltage to frequency needs to be kept approximately constant.

Induction motor drive has three distinct operating regions:

(a) Constant torque: The inverter voltage is controlled up to a maximum value limited by the supply voltage. As the motor speed and the voltage are increased in proportion, constant V/F, the rated flux linkage is maintained up to the base speed. Values of torque up to the maximum value can be produced at speeds up to about this base value. The maximum available torque is proportional to the square of the flux linkage. Typically, the induction motor is designed to provide a continuous torque rating of about 40%–50% of its maximum torque.

(b) Constant power: For higher speed, the frequency of the inverter can be increased, but the supply voltage has to be kept constant at the maximum value available in the supply. This causes the stator flux linkage to decrease in inverse proportion to the frequency. Constant power can be achieved up to the speed at which the peak torque available from the motor is just sufficient to reach the constant power curve. A constant power speed range of 2–2.5 can usually be achieved. Within this range, the motor frequency is increased until at maximum speed.

(c) Machine limit (pullout torque): Once the machine limit has been reached, the torque falls off in proportion to the square of motor frequency. Operation at the higher end of this speed range may not be feasible as the motor power factor worsens. This in turn results in a higher stator current than the rated value. The motor heating may be excessive unless the duty factor is low.

Induction motors are used in applications requiring fast and precise control of torque, speed, and shaft position.

The control method widely used in this type of application is known as vector control, a transient response at least equivalent to that of a commutator motor can be achieved.

The voltage, current, and flux linkage variables in this circuit are space vectors from which the instantaneous values of the phase quantities can be obtained by projecting the space vector on three radial axes displaced 120 degrees from each other. The real and imaginary components of the space vectors are separated, resulting in separate direct and quadrature axis equivalent circuits but with equal parameters in the two axes.

Changes in the rotor flux linkage can be made to occur only relatively slowly because of the large value of the magnetizing inductance of the induction motor. Vector control is based on keeping the magnitude of the instantaneous magnetizing current space vector constant so that the rotor flux linkage remains constant. The motor is supplied from an inverter that provides an instantaneously controlled set of phase currents that combine to form the space vector, which is controlled to have constant magnitude to maintain constant rotor flux linkage. The second component is a space vector, which is in space quadrature with the instantaneous magnetizing current space vector. This component is instantaneously controlled to be proportional to the demand torque.

To the extent that the inverter can supply instantaneous stator currents meeting these two requirements, the motor is capable of responding without time delay to a demand for torque. This feature, combined with the relatively low inertia of the induction motor rotor, makes this drive attractive for high-performance control systems.

Vector control requires a means of measuring or estimating the instantaneous magnitude and angle of the space vector of the rotor flux linkage. Direct measurement is generally not feasible. Rapid advances are being made in devising control configurations that use measured electric terminal values for estimation.

29.5.2.2 Slip-Ring (Wound-rotor) Induction Motor Drive

Wound-rotor induction motors with three rotor slip rings have been used in adjustable speed drives for many years. In an induction motor, torque is equal to the power crossing the air gap divided by the synchronous mechanical speed. In early slip-ring induction motor drives, power was transferred through the motor to be dissipated in external resistances, connected to the slip-ring terminals of the rotor. This resulted in an inefficient drive over most of the speed ranges. More modern slip-ring drives use an inverter to recover the power from the rotor circuit, feeding it back to the supply system.

The speed of slip-ring induction motor can be controlled by the following:

• Stator frequency control as with a cage rotor machine.

• Rotor frequency control.

• Rotor resistance control.

• Slip energy recovery (Kramer system). For capital cost reasons, the last two are commonly used.

Addition of rotor resistance especially for starting large induction motors is well-known. The basic effect produced by adding rotor resistance is to alter the speed at which maximum motor torque is developed. Unfortunately, power dissipation as heat in the rotor resistance bank takes place; earlier means to overcome this shortcoming were to convert the rotor power to DC and feed a DC motor on the same shaft. The rotor slip energy, when running at reduced speed, is therefore reconverted to mechanical power. This is the “Kramer” system. The disadvantages of this approach were the extra maintenance and capital costs.

The static Kramer system overcomes these shortcomings by replacing the DC machine with a line-commutated inverter that returns the slip energy directly to the AC line, either directly (on lower power systems) or via a transformer. A key advantage of the Kramer drive system is that the slip-energy recovery equipment (DC machines or static inverter) needs only be rated for a fraction of the maximum motor rating. This is true when a small speed range is required and provided that a separate means is provided of starting the motor. This is because the motor rotor current is proportional to torque and the rotor voltage inversely proportional to speed.

Naturally, if the slip-energy recovery network can be rated to withstand full rotor voltage (developed at standstill), a controlled speed range of zero to maximum could be achieved. However, this is generally only feasible on smaller motors (below 2000 kW) where the rotor voltage is sufficiently low for an economic inverter package. Secondly, if a full speed range is needed, the slip-energy recovery network has to be rated at full motor power, so static Kramer drives become uneconomic for wide speed ranges. The overall system power factor would be very low for a wide speed range system.

For the above reasons, Kramer drives are very suitable for high-power drives (>200 kW) where a small speed range is required. Pump and fan drives present therefore good economic applications. Kramer drives have also been used for low-speed-range endurance dynos using the recovery system to control torque of induction generator. As with all line-commutated converters and inverters, current harmonics are produced, and these can be reduced to acceptable values. However, as the slip-energy recovery network is only power-rated in direct proportion to the speed reduction required (assuming constant load torque), the magnitudes of the harmonic currents generated are proportionally less than with drives where the solid-state converters have to handle the whole drive power. Harmonics of the rotor rectifiers are transmitted through the rotor and appear as noninteger harmonics in the main supply.

The main disadvantages of the slip-ring induction motor drive are (a) the increased cost of the motor in comparison with a squirrel cage, (b) the need for slip-ring maintenance, (c) difficulty in operating in hazardous environments, (d) the need for switchable start-up resistors, and (e) the poor power factor compared with other types of drive.

29.5.4.1 Brushless DC Motor Drive

This type of machine has a similar construction to a standard synchronous machine, but the rotor magnetic field is produced by permanent magnet material. A position sensor is used to ensure synchronism between the rotor position and the stator magnetomotive force (MMF) via drive signals to the inverter. The use of new magnet materials characterized by high coercive force levels has reduced magnet sizes and largely overcome the demagnetization problem. The absence of the field copper losses improves the machine efficiency.

As the permanent magnet is the source for excitation, the brushless DC motor (BDCM) can be viewed as a constant flux motor. A limited amount of flux weakening can be achieved by increasing the load angle of the stator current. Achieving a useful constant power range is not usually practical with this type of motor. A large demagnetizing component of stator current would be required to produce a significant reduction in magnet flux, and this would increase the stator loss substantially.

The required base torque determines the motor size, and the losses are essentially independent of the number of stator turns. At speeds up to the base speed of the constant power range, the efficiency of the motor is essentially the same as for one designed for rated voltage at base speed. For operation above base speed, the stator current from the inverter is reduced in inverse proportion to the speed. This mode of operation in the high-speed range reduces the dominant stator winding losses relative to a machine in which the flux is reduced and the current kept constant. The losses in the inverter are, however, increased due to its higher current rating. For an electric road vehicle that must carry its energy store, the net energy saving may be sufficiently valuable to overcome the additional cost of the larger inverter. A further advantage of this approach is that, if the DC supply to the inverter is lost, the open-circuit voltage applied to the inverter switches will be within their normal ratings.

The BDCM has higher volumetric power density compared with other types of motors (induction or synchronous). They are particularly suited for the high values of acceleration required in drives (e.g., machine tools). They are often operated with high acceleration for a short time followed by a longer period of low torque. At such low values of load factor, the cooling capability is frequently not a limitation. The major interest is in obtaining the maximum acceleration from the motor. The short-term stator current of a BDCM is limited to the value required for magnet protection. These values of acceleration are significantly higher than that can be achieved with either induction or DC motors of similar maximum torque rating.

29.5.4.2 Switched Reluctance Motor Drive

This motor can be regarded as a special case of a salient synchronous machine in which the field MMF is zero and the torque is produced by reluctance or saliency action only. The rotor has no winding. The switched reluctance motor (SRM) drive needs an inverter whose frequency is locked to the shaft speed, but since the torque is linearly proportional to the square of the stator current, the use of unidirectional current involves little sacrifice in performance.

Generally, the use of position sensors in the SRM and BDCM is something of a disadvantage in both cases. The SRM does not require permanent magnets, which can cost and may involve demagnetization risks and limit top speeds due to centrifugal forces. The SRM hence has a simpler construction and is more robust. However, the need to magnetize the motor from the AC side adds to inverter costs and may increase peak current levels significantly, hence raising stator copper losses.

Switched-reluctance synchronous motors have a cylindrical stator with three AC windings and a solid rotor (without any winding) with a moderate orthogonal axis magnetic saliency up to 4 (6) to 1. High magnetic saliency is obtained with multiple flux barriers. The conventional SRMs are to some extent (up to 100 kW) used in low-dynamics VSDs with open-loop speed control, as the speed does not decrease with load. Consequently, the control is simpler than with induction motors.

The main drawback of the conventional SRD is the low motor power factor and the relatively poor torque density, which leads to a higher kVA rating of the power converter (~20%). The main advantage of the synchronous reluctance motor over the induction motor of similar rating is the higher efficiency. Compared with the squirrel-cage induction motor, the rotor loss is small or negligible in synchronous reluctance machines. If the saliency ratio is sufficient to produce a power factor equal to that of the induction motor, the stator winding loss will be the same. Also, the stator iron losses will be similar for the two motors.

The reluctance motor is capable of operation in the constant power mode of operation. As for all AC drives, when the supply voltage limit is reached above the base speed, the flux linkage is reduced in inverse proportion to the shaft speed, and the torque is inversely proportional to speed squared.

29.5.4.3 Linear Motors

There are applications in which linear motion, as opposed to rotational, is required. A linear machine has the same operating principles as those applied to all other rotating machines. The PWM-VSI converters and motor control principles discussed in this chapter are also applicable to this type of motor.

There are two types of linear motors:

• LIM—linear induction motor

• LSM—linear synchronous motor with permanent magnetic excitation

The LSM type has the following advantages over the LIM:

• Better power factor

• More responsive control

• Higher efficiency

The disadvantages of LSM are the following:

• Very accurate position feedback is required.

• The use of PM—expensive and heavy.

Transport, material handling, and extrusion processes are a few examples in which linear motors have successfully been employed.

29.5.4.4 Stepper Motors

Stepper motors are either built in a similar manner to BDCM, with permanent magnets embedded in or bonded to the rotor or a rotor with no magnets. The latter type is made of a ferrite magnetic material, and its circumference is cut to form a number of slots, forming teeth lengthwise to the rotor axis.

Torque production can be based on (a) magnetic reluctance (as in SRM), (b) magnetic attraction (as in BDCM), or (c) both magnetic reluctance and attraction.

Stepper drives do not offer dynamic speed control, and the main action is to accelerate at full torque to full speed, maintain the speed, and decelerate at full torque. In comparison with the reluctance-type stepper motor, the permanent magnet type offers greater torque for a given speed, particularly at start and low speeds.

Most drives incorporate controllers with connections for a communications link for supervisory control by PLC, hard-wiring connectors for analog/digital inputs and outputs, and some are equipped with software for communications with a computer or handheld keypad. Table 29.12 lists typical options.

Unlike above motor drives, the stepper motor can achieve precise position control without the need for any external feedback.

29.5.4.5 Actuators

Actuators are widely used in industry, primarily for positioning tasks. Their designs are based on all sorts of force-producing principles. Ref. [7] describes several types of direct drive electric actuators, including (a) the DC actuator (moving coil type), (b) induction actuators, (c) synchronous actuators (moving magnet DC type), (d) reluctance actuators, and (e) inductor actuators (polarized reluctance type).

Electric actuators are used increasingly in control systems and automated electromechanical equipment. Typical specification factors include range of motion, type of motion (linear or rotary and stepwise or continuous), resolution needs, speed of response, environmental conditions, supply conditions, allowable electromagnetic noise emission level, need for integrated position and velocity sensors, maintenance needs, eligibility, cost, peak, and continuous torque.

The main demands of industry for high-performance systems are the following:

(a) A convenient supply and low power consumption

(b) Reliability and robustness

(c) Low initial cost and maintenance

(d) Fast response

(e) Linear “torque-excitation” characteristics

29.5.4.6 Integrated Motors

The integral motor consists of a standard AC motor with an integrated frequency inverter and EMC filter. It is robust and specified for reliable operation and often designed to handle rough working conditions, including ambient temperature −25°C to −40°C and dusty, corrosive, and humid environments (enclosure IP55). This type of drive uses a standard induction motor with the AC/AC converter integrated in the motor frame often as a separate converter box mounted directly above the motor frame in place of the terminal box. The power popularity of this type of drive is limited to 0.5–7.5 kW.

This type of motor offers the following advantages:

• Save space by eliminating the need for a separate controller

• Reduce installed costs because cabling between motor and converter is eliminated

• Eliminate motor problems caused by high voltage transient due to output cable capacitance

• Minimize EMC due to high dV/dt

The integrated drive includes most features including start, stop, forward, reverse, speed and torque controls, and controlled acceleration and deceleration.

29.6.6 Impact of PWM Waveform

29.6.6.1 PWM Voltage Waveform

Fast switching of IGBTs (typically <1 μs) results in high dV/dt, typically 3–5 kV/μs, and possible voltage overshoot at turn-off that can last for a few microseconds. The fast rate of rise/fall of voltage combined with high peak voltage at the turn-off results in a premature failure of motors and EMC. Refs. [9–11] deal with the effect of PWM waveforms of VSD.

The following is a brief summary of the effect of the unfiltered waveforms.

29.6.6.2 Effect on Motor

• Premature insulation failure due to partial discharge as a result of peak voltage, high dV/dt, and high frequency.

• Motor shaft voltage that forces current into the shaft bearing, leading to early bearing failure.

• Motor stray capacitance (between windings and earthed frame) leads to earth-current flow caused by high dV/dt.

• High dV/dt creates nonuniform distribution of voltage across the winding, with high voltage drop across the first few turns and consequential failures.

• In a large motor, voltage differential on the frame is likely to develop in spite of protective earthing of the motor. More than one earthing point is needed.

29.6.6.3 Effect on Cables

• Voltage-doubling effect at the rising/falling edges of voltage waveform due to wave propagation in long cables [9].

• Earth-current flows in cable stray capacitance due to dV/dt.

• Restriction on cable type used and earthing methods employed.

• Cable type (armored, screen, and multicore).

• Likelihood of cross talk with other surrounding cables running in parallel.

• For PWM drives, the cost of cabling is likely to be significant due to special requirements of cables and termination methods employed.

29.6.6.4 Effect on EMC/Insulation/Earthing

• Inductive and capacitive couplings between live components and earth result in common-mode and differential-mode noise. This could lead to malfunctioning of nearby sensitive equipment.

• The voltage to earth applied on drive components pulsates at the switching frequency, adding voltage stresses (worst at low speeds and low modulation index). This poses additional insulation requirements on main power components (motor, cable, output filter, and transformer).

• A fourth wire may be required between the motor frame and the converter virtual earth so that a low impedance path is provided for the motor earth current.

• Strict rules must be observed when cabling and earthing.

29.6.6.5 Motor Insulation

High peak voltages can be experienced at the motor terminals especially when long cable is employed (10–100 m depending on the size of motor). This is usually caused by voltage-doubling phenomena of a transmission line with unequal line and load impedance. Motor line voltage can reach twice the DC link voltage with long cables.

Fast voltage rise times of 5000 V/μs can be measured at the motor terminals. Under this condition, the motor insulation becomes stressed and can lead to a premature breakdown of a standard motor insulation. When motor fails due to insulation stress caused by high peak voltage and fast voltage rise times, failure occurs in the first turn as phase-to-phase short or phase-to-stator short. The highest voltage is normally seen by the first turn of the winding.

Standard motor capabilities, established by the National Electrical Manufacturers Association (NEMA) and expressed in the MG-I standard (part 30), indicate that standard NEMA type B motors can withstand 1000 V peak at a minimum rise time of 2 μs (500 V/μs). Ref. [10] describes the effect of PWM inverter waveform on motor insulation in more detail.

29.6.6.6 Bearing Current

Bearing current and shaft voltages under 50/60 Hz sine-wave operation has been recognized since 1924. The bearing impedance characteristics largely determine the resulting bearing current that will flow for a given shaft voltage [11].

The rotating machines have three basic sources of shaft voltage. These are the following:

• Electromagnetic induction from the stator winding to the rotor shaft (due to small asymmetries of the magnetic field in the air gap that is inherent in a practical machine design). The design limit is <1 V RMS.

• Electrostatic coupled from internal sources: such as voltage in motors where rotor charge accumulation may occur (belt-driven coupling and ionized air passing over rotor fan blades).

• Electrostatic coupled from external sources such as PWM inverter. The presence of high dV/dt across the stator neutral to frame ground causes a portion of the voltage to ground due to capacitor divider action. The presence of PWM-related voltage components is undesirable and leads to a premature bearing failure.

• The fundamental cause of the shaft voltage is magnetic asymmetry between the stator and the rotor or possibly a phase shift of the motor voltage waveform. System ground may also contribute to this condition through unbalance system voltage.

• NEMA-500 recommends the consideration of insulated bearing for motor frame of certain sizes.

29.6.6.7 Electromagnetic Compatibility

The main sources of electromagnetic emission of PWM-VSI drives are described in [12] as follows:

AC/DC converter: Supply harmonics caused by supply bridge rectifier (from 100 Hz to 2.5 kHz): as already explained, the input bridge circuit with an SCR or diode bridge is a source of supply harmonics in the input current.

DC/AC inverter: Harmonics caused by the switching of the inverter bridge (from 3 kHz to 20 MHz): the inverter bridge uses fast-switching devices to create PWM voltage output. The inverter is a source of a wide band of frequencies, typically extending from the basic switching frequency (usually several kilohertz) to the radio high-frequency bands at 20 MHz. The radio frequency current spreads out into both the supply and motor connections. An EMC filter is often used to limit the spread of high-frequency harmonics into the supply.

Control electronics: The control circuit employs a microprocessor with clock frequency of several megahertz, typically 20 MHz. The clock wave produces frequencies, which are multiple of 20 MHz up to 300 MHz (Table 29.16).

29.6.7 Techniques Used to Reduce the Effect of PMP Voltage Waveform

29.6.7.1 Output Line Reactor

A reactor increases the rise time, but the benefit of its connection may be negated as follows:

• Beneficial connection if cable length is short enough for reflections to be superimposed within rise time, that is, if rise time is increased beyond critical value of cable length.

• Harmful connection if cable length is too long; the reactor may have negligible effect on peak voltage (theoretically, its presence is insignificant in this case) or ringing period, but it will increase the duration of each overshoot, thus increasing the probability of partial discharge.

Adding a series line reactor between the motor and inverter is not as simple as illustrated above because the reactor adds or adjusts other resonant modes where the reactor rings with lumped capacitance. These resonant modes are pure transmission line modes and can double voltage. Some line inductance helps short-circuit protection. If earth current is limited by other means, then the coupled reactors may be helpful.

29.6.7.2 Sine-Wave Filter

This mechanism filters the PWM carrier frequency; thus, the converter output voltages are sinusoidal. This type of filter is best suited for low-performance drives and/or retrofit applications (old or standard motors). Ref. [13] and Table 29.17 illustrate the filtering options for high-power VSDs.

Employing a filter at the inverter output has some practical consequences:

• Cost and weight of filter

• Filter power losses and voltage drop

• A small derating of power switches due to circulating current between filter L, C, and DC link capacitor

• Reduced torque response due to time delay in the filter, sine-wave type

• Potential oscillations that have to be electronically dampened

• Potential induction motor self-excitation

29.6.7.3 PWM (dV/dt) Filter

This reduces the dV/dt seen by the motor to a level, which does not compromise the motor or EMC. It is ideal for high-performance drives with custom-built motors.

29.6.7.4 RC Filter at Motor Terminals

A simple RC network is used at the motor terminal; the capacitor would represent a short circuit for the high-frequency components (sharp dV/dt). Wave reflection will not happen if the resistor value is similar to the cable characteristic impedance. Resistor losses are generally small, as current flow will only occur at the rising and falling edges of the PWM waveform.

29.6.7.5 Common Mode Reactor

The presence of capacitive current due to the high dV/dt can be improved by employing a common-mode reactor. It is well established that such a choke is not effective to reduce the RMS and mean values of the leakage current but only effective to reduce the peak value. The presence of such a choke in the circuit increases the inductance and resistance of the zero sequence impedance.

29.6.8.1 Regenerative Braking

Several techniques are usually used for regenerative braking.

A simple diode front-end supply bridge will operate in two quadrants (positive and negative speeds). There is no regenerative power capability as any regeneration of power would result in an increase in the DC link voltage, and the drive will trip on overvoltage.

If a small amount of regeneration is required, during stopping, or speed reversal, then a dynamic brake chopper may be used. This is a simple chopper with a dynamic brake resistor. The size of the resistor is very much dependent on the regenerative brake energy, its magnitude, and repetition rate.

Full power regeneration is possible by employing a fully controlled antiparallel thyristor front end. This is similar to that used on DC drives or cycloconverters.

A more modern approach is to use pulse converter front end (fully controlled bridge). This is a four-quadrant converter with the ability to control the power factor and the DC link. Such an option necessitates the use of a precharge circuit for the DC link and smoothing inductance on the AC side.

For fully regenerative drives, the supply needs to be receptive.

By using a PWM rectifier as a primary converter in this composite structure, both the problems of regeneration and line current distortion are successfully solved—with the penalty of having a much more complicated converter structure and control system.

With the modern PWM-VSI VSD controller, the supply bridge can be fully controlled. Such an option offers the following benefits:

• Fully regenerative drive

• Unity power factor all time

• Sine-wave input voltage and current

• Can operate with controlled power factor (e.g., leading power factor)

• Can operate as an active filter while supplying power to the load. Possible elimination of low-order supply harmonics (fifth and seventh)

• Output voltage equals the input voltage

29.7.2 Applications by Industry

29.7.2.1 Deep Mining

Ref. [14] lists various high-power MV VSI inverter drive applications for the mining industry. In deep mine conveyor belt applications, a PWM-VSI drive offers significant advantages over other conventional alternatives. The following benefits have been identified for deep mine conveyor belts applications:

• Improved drive starting and stopping

• Improved reliability

• Matching belt speed to production

• Easier belt inspection

• Reduced belt wear and increased belt life

• Lower specification belt material may be used

• Low speed running to reduce coal removal by windage

• Manpower saving—less coal spillage

• Unity power factor with low harmonic content

• Reduced AC supplies disturbances

For hoist applications, the PWM-VSI drives can also be used to replace DC and cycloconverter drives for mine hoist applications. The benefits are the following:

• Improved drive control, with 100% continuous stall torque available with induction motors

• Reduced AC supplies disturbances

• Very unlikely to need reactive MVAR correction even at high ratings

• Improved immunity to AC supplies dips

The use of electrically coupled mine hoist systems has many advantages especially for deep mines and set to become an essential feature of many new mine shaft systems. The circuit is shown in Fig. 29.8.

The power flows naturally from motor 1 to motor 2 such that at the point of balance, the AC supply current is virtually zero and at near-unity power factor.

This technology is the natural successor to the DC electrically coupled winders and totally solves the poor AC power factor that would result if twin drive cyclo motors were used.

29.7.2.2 Industrial Processes

In this industry, there are a number of viable drive solutions available for the major market power ranges, from LCIs to FCI. However, there is a developing market for MV variable speed drives. The PWM-VSI using new high-power IGBTs or IGCTs appears to be the best solution for the future. Benefits include better power factor, no limit on frequency, and higher voltages.

PWM-VSI converter cost is likely to be higher than equivalent other well-established technologies (e.g., LCI). Hence, the flexibility in choice of motors and improved control must be exploited. The advantage of offering an MV solution may prove significant. Possible means of reducing motor costs are the following:

1. Higher frequencies are achievable, allowing the use of high-speed motors and gearboxes.

2. Higher pole number machines can be used, giving a cost saving.

3. Better power factor over the speed range giving power supply saving.

4. Induction motors with rotors adapted for use with VSDs can be used with resultant cost savings over standard DOL, fixed speed motors.

5. Higher voltages and smaller conductors.

6. However, at low powers, the relative cost of the machines is less significant versus the cost of the converters. Hence, the viability of this technology in this market requires close examination.

The cycloconverter drive with a synchronous motor is used when four-quadrant operation is required. Particularly for high power rating with high torque at low speed and at standstill but with a rather low maximum speed are drive requirements. Gearless cement mill drives were the first applications of cycloconverters. The mill tube is driven from low-speed wraparound motor with a high number of poles.

29.7.2.3 Metal Industry

The majority of installed hot mill drives are cycloconverters. A few LCI drives have been used, but applications are limited for such technology on mill main drives. Most early-generation plants are equipped with DC drives. The trend is to replace DC with AC.

Direct current drive applications were universally used in the first generation of rolling mills. The market for new mills requiring this technology is declining as the steel industry moves to AC as a preference. On early-generation mills where motors are retained, DC drives are likely to be required. Customers in their inquiries, some requesting AC alternatives, are still requesting DC drive solutions.

DC drives are probably still the most economic for the power range of 750–1500 kW. The number of manufacturers however, producing DC motors, is declining, particularly in the case of large DC motor manufacture. The lower price of DC solutions is offset by the advantage of the use of AC motors in AC solutions, making AC the more popular choice.

Current-source inverter LCI can be applied to roughing stands of rod and bar mills. Technical limitations include the risk of torque pulsation and a minimum drive output frequency of 8 Hz.

Cycloconverter is the solution most often used. However, it is relatively expensive compared with alternative technology. Major cost penalties arise from supply transformers, cabling, and bridge configuration. In some cases, active power supply compensation equipment may be required, taking the costs even higher. Cycloconverter solutions will still be cost-effective for medium-to-high power, low-speed, low-frequency (say below 21 Hz maximum operating frequency) applications. This would include hot reversing mills, with direct drive, for the primary rolling processes, albeit a declining application area, and possibly for direct-drive, low-speed, high-torque roller-table applications. Technical limitations include limited output frequency (typically 29 Hz for 12-pulse at 60 Hz supplies), which can necessitate the use of two-pole motors to reach application speeds.

The high-power PWM-VSI using new power devices (IGBT/IGCT) appears to be the best solution for the future. Benefits include better power factor, no limit on frequency, and higher voltages. Potentially, either the two-level or the multilevel solution will meet the market requirements.

In some applications, like coilers/uncoilers, the system is composed of several drives, which have different power cycles; when some drives are furnishing power, others are braking. A common DC bus system will allow that the energy fed from drives operating in the regenerative braking mode will be utilized by other drives connected to the same DC bus but operating in the motoring mode. The supply bridge, that is, rectifier, feeding the DC bus system will only be rated for the total system power.

The benefits of the DC bus systems include the following:

• Good operating power factor

• Low harmonics (lowest when using 12-pulse or 18-pulse front ends)

• Possibility of energy transfer on the common DC link solutions (reducing front-end converter and transformer sizes with attendant energy saving, possibility of using kinetic energy to allow controlled stopping)

29.7.2.4 Marine and Offshore

Drive powers are commonly in the range of 0.75–5.8 W for thrusters and 6–24 MW for propulsion. The evolution in the commercial market is toward powers from 1 to 10 MW for propulsion. Higher powers are required for naval applications with package drive efficiency better than 96%.

The PWM inverters at these powers would allow the use of induction machines, rather than the more expensive synchronous alternatives required for LCI drive. This could give savings in the price of the motor.

Current source inverter (CSI) drive LCI is used for all applications except for icebreakers where cycloconverter drives are used. The PWM (voltage source) inverter using new force-commutated drive appears to be the best solution for the future. Benefits include better power factor, no limit on frequency, and higher voltages. Many icebreakers and some other ships are equipped with diesel-generator-fed cycloconverter synchronous motors with power ratings up to about 20 MW per unit.

29.7.3 Examples of Modern VSD Systems

29.7.3.1 Integrated Power System for All Electric Ship

This is a full-scale main propulsion drive for the US Navy [15]. It consists of a main propulsion 19 MW induction motor drive system. The power converter consists of three 6-pulse rectifier stages, three 6 kV DC links, and 15 IGBT-based H-bridges feeding a 15-phase induction motor (Fig. 29.9).

This drive demonstrates the potential of modern power electronics over more traditional solutions such as cycloconverter and LCI. The volumetric power density of the new converter is reported to be 905 kW/m³, compared with 455 kW/m³ for cyclo and 313 kW/m³ for LCI.

29.7.3.2 Subsea Separation and Injection System

This is a full-scale pilot plant developed to increase recovery and improve the economics of offshore oil and gas fields. The system comprises several VSD units, typically 500 kW oil pump, 1 MW multiphase booster and 1–2.5 MW water injection drive unit, and such a system is called “SUBSIS” [16]. The main task for such a system is to separate the bulk water from the well stream and treat it either for discharge into the sea or reinjection into the reservoir.

This system employs subsea-based rotating machinery for pumping, boosting, and compression. The subsea electric power distribution system (SEPDIS) is an innovative and cost-effective subsea processing (Fig. 29.10). The pump motors are mounted in a pressurized vessel and positioned on the seabed. Ref. [16] identifies the benefits of subsea drives as follows:

• 3%–6% increase in oil and gas recovery

• Improved pipeline transportation conditions by removing water from the well stream

• Reduced environmental impact due to lower energy consumption and reduction in chemicals used to inhibit corrosion

• Reduced size and cost of new platforms

• Cost-effective development of marginal fuels through reuse of existing infrastructure

29.7.3.3 Shaft-Generator for Marine Application

During cruising at sea, up to 3.5 MW of electric power is extracted from the ship's main diesel engine/propeller shaft (90,000 hp per ship) via a salient pole shaft generator, which is fitted to the main propeller shaft. The converter output voltage (set at 60 Hz) is stepped up to 6.6 kV [17].

The converter is based on 24-pulse converter (LCI) technology, which is traditionally used as main ship propulsion. The shaft generator output frequency varied between 14 and 25.7 Hz; six-phase generator and the output stage are configured as 24-pulse, via a step-up transformer. At the output, a passive LC filter is employed, with a synchronous condenser started by a pony motor.

This type of application is likely to significantly benefit from the higher volumetric power density of the PWM-VSI with fully controlled front-end. Such a system will eliminate the need for a passive filter at the output stage and use a standard three-phase generator.

The same technology is also applicable to high-speed generators and windmill energy. The ability of the active front end to sustain fixed DC link voltage over a relatively wide shaft speed range results in a very good control of the output voltage, irrespective of the shaft speed.

In wind power plants, the optimal efficiency of the wind turbine depends on the speed when the wind conditions change. It is, therefore, advantageous to vary the speed of the generator and link it via a frequency converter to the AC system.

For high-speed generators, driven by diesel engine or gas turbine, the fully controlled converters enable direct power conversion from AC high frequency (hundreds of Hz) to fixed power frequency (50/60) Hz fixed output voltage. The magnitude of the output voltage is kept constant irrespective of speed variation of the generator (Fig. 29.11).

29.7.3.4 Linear Motor Drive for Roller Coaster

This drive involves a fully regenerative PWM-VSI. The supply front end is made of antiparallel thyristor front end, while the machine bridge is based on PWM IGBT VSI. “Superman” has been developed for Six Flags California at Magic Mountain. The inverter output frequency is 0–230 Hz and 525 V AC RMS. The power rating is 1.8 MW. The duty cycle is 1.8 MW for 7 s, followed by 16 s at zero power, and 1.3 MW for 5 s and a stop period of 32 s. This ride involves acceleration at 4.5 g, speed and free fall (6.5 s of weightlessness, during which a height of 415 ft is achieved [18]).

The same concept employed in this application could be used for aeroplane launcher on aircraft carriers, instead of the conventional catapult.