VARIABLE SPEED ELECTRIC DRIVES - CHARACTERISTICS AND APPLICATIONS

D S Henderson

Napier University, UK.

INTRODUCTION quality and reliability, often arise as a consequence of

the use of a speed control system.

In all of the industrialised nations, the majority of non-

domestic electricity consumption is by electric motors. SPEED CONTROL
In the UK for example, 65% of the electricity used by
industry powers motors, whilst 23% of the electricity
used by commerce is for motors. Any industry which Speed control can save energy in flow control
wishes to improve its profitability by reducing applications as a consequence of the laws which govern
electricity consumption must achieve efficient use of centrifugal fans and pumps. These laws determine that
motors. the flow is proportional to the speed and that the power
absorbed by a centrifugal fan or pump is proportional to
A second force behind the increasing attention being the cube of the speed. Hence a 20% reduction in flow
paid to efficient use of motors is that of the gives rise to (nearly) a 50% reduction in power.
environmental effects from the emissions from
generating powerstations. In a typical fossil-fuel Speed control can be achieved by a number of means.
powered power station the generating efficiency is Coupling control, variable speed motors and electronic
around 35%. Coupled with transmission, distribution speed control are the three main categories. The former
and then the electrical and mechanical losses in a drive of these methods uses a fixed speed motor which drives
system, the net effect is that 100 kWh of fuel input can a coupling which can alter the speed of the load. These
result in only 11 kWh of useful output energy. are described in detail in a separate paper of this
colloquium. The synchronous speed equation provides
The agreements made at the Earth Summit in Rio have two variables, number of poles and frequency, which
resulted in major international commitments to reduce give rise to the other two categories.
CO2 emissions. This has resulted in proposals to
legislate for minimum efficiency standards for motors Variable speed motors are usually induction motors in
and motor driven plant in some countries, e.g. Canada, which the number of motor poles can be altered by
the US and Denmark. external switchgear, and hence change the speed of the
motor. Common pole configurations are 2:4 or 4:6
The majority of installed motors are a.c. induction giving rise to two discrete operating speeds. This is
motors, for many years the first choice due to its low useful for applications where the flow is to be either one
cost, low maintenance and reliability. The most value or another, e.g. day and night flows. This is a
common applications are in the movement of fluids, simple and low cost solution to the energy efficiency
(liquid or gas). Also, over 50% of the motor capacity problem where sophisticated control is unnecessary.
drives either fans or pumps and so the costs of fluid
movement are very high. In the UK alone, these costs Electronic speed control, performed by plant referred to
are estimated at over £2,000 Million per annum. as Variable Speed Drives rely on the control of the
motor supply frequency for their operation. The basic
In many cases, the energy used to move fluids is wasted concept of these drives, Figure 1, is that a rectifier
due to poor design and operation of the system. Most converts the fixed frequency grid supply to d.c. A d.c.
systems are designed to deliver a higher volume of fluid link stage smoothes the rectified output to a stable d.c.
than is actually required. Flow control is used to match voltage (or current). This d.c. is then inverted to provide
the actual demand, traditionally by introducing a a synthesised a.c. waveform at the motor terminals. The
mechanical restriction. Examples are outlet damper fans frequency and power of the a.c. supply delivered to the
or throttle valves on pumps. motor is controlled by a microprocessor built in to the
inverter.
The energy efficient alternative is speed control.
Replacing mechanical methods of flow control with This paper introduces the most common types of
speed control offers an energy and cost effective means variable speed drive systems presently available; and
of controlling flow. Secondary benefits, including for each type reviews the method of operation, lists the
principle characteristics and highlights the principle Natural commutation of current from one switch to
applications. another is relatively slow and so simple forced
commutation circuits are often used. Regenerative
braking is possible through firing angle control of the
INVERTER DRIVES devices in the inverter and the rectifier, without any
additional circuits.

The term “inverter” nearly always implies a static Until recently, the fact that line-frequency thyristors
switching system which synthesises a varying frequency with simple commutation circuits were suitable for the
and controlled amplitude three phase system from a inverter switches was an important asset of CSI drives.
direct current or voltage supply. The simplest from of With the availability of higher power rated controllable
such a system is shown in Figure 2 where six switches switches, nowadays CSI drives are used mostly in very
are arranged in a bridge pattern across a d.c. supply. In large power applications only. Drives are available with
theory it is possible to adopt any switching sequence of converter voltages of up to 10 kV, rated currents up to
these devices, but in practice there are constraints 2000 A and rated powers of up to 60 MW.
necessary to avoid a short-circuit on the d.c. side. In
addition, as the inverter load is usually an inductive Voltage source inverters (VSI’s)
(motor) one, freewheel diodes are required in parallel
with each device to protect against overvoltages. Method of operation
In the voltage source inverter, a firm d.c. link voltage is
A common switching sequence results in the so-called switched between the motor windings, and the load
“quasi-square” waveform shown in Figure 3. Any currents establish themselves depending on the load
desired three-phase output frequency can be obtained impedance of the motor. The d.c. link voltage is
within the switching rate capabilities of the devices maintained by the presence of a capacitor in the d.c.
used, however control of the output amplitude requires link, Figure 4b.
further control action.
Characteristics and applications
The source of the d.c. input to the inverter is the starting The polarity of the d.c. link voltage can not be reversed,
point for describing the different inverter types. If an and so regeneration requires the reversal of the link
uncontrolled diode bridge rectifier is used, the d.c. current. In order to recover such regenerative power, it
output is a substantially constant voltage. This has the is thus necessary to deploy a four quadrant converter
advantage that a battery can be used as a standby power (switch-mode or back-to-back) at the input bridge stage.
source. The disadvantage is that if control of the
inverter output voltage is required, then this has to be The majority of inverter drives used nowadays are
done separately. If a controlled rectifier is used, then the fundamentally of the VSI type. Different methods of
d.c. voltage can be varied by firing angle control of the control of the output voltage and different speed control
input bridge, and reverse power flow is also possible (if techniques give rise to a multitude of overall drive
fully controlled). topologies.

There are two approaches to the control of the d.c. link; The simplest of these is the quasi-square wave or six-
either voltage control or current control. These step inverter. Each inverter switch is on for half a cycle
approaches give rise to the two fundamental inverter and a total of three switches are on at any one time.
types; current-source inverters and voltage-source Control of the output voltage is achieved by varying the
inverters. d.c. link voltage. This can be done either with a d.c
chopper circuit in the link itself, or through controlled
Current source inverters (CSI’s) rectification at the input bridge. Low-order harmonics
are a feature of this drive which gives rise to operational
Method of operation problems such as torque pulsations and noise at low
In the current source inverter, a controlled d.c. link speeds. This type of drive has been superseded in
current is switched between the motor windings in a performance several times over and is rarely applied
sequential pattern, and the load voltages establish nowadays.
themselves depending on the particular speed, torque
and rotational direction of the motor. The “hard” d.c.
link current is maintained by the presence of a large
inductor in the d.c. link, Figure 4c.

Characteristics and applications

PWM inverters To emulate the field orientation process of the d.c.
motor, the flux vector drive needs to know the spatial
Method of operation angular position of the rotor flux inside the a.c.
A better approach to the control of the output voltage is induction motor. This sophisticated control is available
the PWM (Pulse Width Modulation) technique. In this in the form of the vector or field orientated control and
system, the output voltage is controlled within the has effectively become the present day “industry
inverter, resulting in the deployment of a cheaper and standard” in respect of inverter drives.
simpler uncontrolled diode rectifier at the front end,
Figure 4a. The inverter switches are controlled in such a Method of operation
way as to produce a series of short positive pulses of The underlying principle of vector control, Figure 6, is
constant amplitude followed by a series of short to separate out the component of the motor current
negative pulses. The mark-space ratio of these pulses is responsible for producing the torque and the component
arranged to produce an average which approaches a sine responsible for producing the flux in such a way that
wave of variable peak amplitude, Figure 5. The duration they are magnetically decoupled, and then control each
of the positive and negative sequences can be controlled independently. Information about the rotor status is
and hence the output frequency and voltage can thus be obtained by feeding back rotor speed and angular
varied independently. position relative to the stator field by means of a pulse
encoder, a closed-loop speed control system. In
Characteristics and applications addition, the motor’s electrical characteristics are
The pulse width and spacing are designed so as to mathematically modelled within the microprocessor
eliminate the low order harmonics. Consequently control system of the drive. The outputs from this
operation at low speeds is much improved when control system are the controlling variables of voltage,
compared with the quasi-square wave inverter. This current and frequency which are fed via a modulator to
method does not use feedback in respect of speed or the motor. It is noted that the torque is therefore
rotor position and is effectively open-loop control. controlled indirectly.
Speed holding accuracy is typically in the order of 1 - 3
%. Characteristics and applications
Vector control offers good torque response, achieving
The PWM inverter drive is available for three phase full torque at zero speed and a performance approaching
drives from 5 kW ratings up to several thousand kW. that of a d.c. drive. Speed holding accuracy can be in
Converter voltage and output current are available at the the order of 0.02% depending on the encoder selection.
660 V, 2000 A or at the 6600 V, 1200 A levels. As This high-level performance is achieved at the expense
such, this drive offers a low cost and simple solution to and the complexity of a feedback device and the
control of a.c. motors. By implication it is best suited to modulator circuitry.
applications which do not require high levels of
accuracy or precision, such as fans or pumps. Power outputs from 5 kW up to 890 kW are typically
available at converter voltages of up to 480 V and rated
currents of up to 1130 A. Paper mills, steel mills,
VECTOR CONTROL DRIVES cranes/hoists etc. represent the type of applications to
which this sophisticated drive is best suited.

Inevitably there must be comparison with the

performance of the d.c. drive. Initially d.c. drives were DIRECT TORQUE CONTROL DRIVES
used for variable speed control because they could
easily achieve good torque and speed response with
high accuracy. These parameters are controlled directly Method of operation
through control of the armature current. In addition, the Torque vector controlled drives are capable of
d.c. motor generates maximum torque through field controlling the stator flux and torque more accurately
orientation whereby the field is always maintained (by than vector controlled drives, while the controller
the commutator) at right angles to the field created by complexity is reduced considerably. Field orientation is
the armature winding. achieved without rotor speed or position feedback using
advanced motor theory to calculate the motor torque
The a.c. drive suffers from the fact that the controlling directly without using modulation. The controlling
parameters are voltage and frequency. Consequently it variables are motor magnetising flux and motor torque.
has relatively poor transient performance, especially in
applications where rapid response to changes in speed The measured motor current and voltage are inputs to
or load are required, e.g. in rolling mills. an adaptive motor model which produces torque and
flux values which are used as feedback signals for the
controller, Figure 7. The states of the power switches
are determined directly by the measured and the frequency, and does not suffer from slip. Hence
reference torque and flux signals. Motor torque and flux accurate speed control is achieved without speed
comparitors compare the actual values to the reference feedback. These drives compete well with large d.c.
values. The outputs from these comparitors are fed to an drives in applications calling for high speeds and in
optimum pulse selector which determines the inverter hazardous area applications (no brushes).
switch positions. In so far as the voltage and current are
subsequently effected by the new switch positions,
closed loop control is operative, however in terms of CYCLOCONVERTERS
speed control, the basic system is an open-loop one.

Characteristics and applications This drive is suitable for use with both induction and
Open loop speed control matches that of closed loop synchronous type a.c. motors. The cycloconverter
flux vector control. Through the use of a feedback employs a line commutation phase control method first
encoder, closed loop speed control can be provided with developed for producing 16.66 Hertz for traction
improved accuracy, a value of 0.01% being quoted by purposes. Within its operational limits, it is an infinitely
the manufacturers. variable frequency drive with regenerative capabilities.

Drives with rated output powers in the range 2.2 kW to Method of operation
315 kW are presently available. Again this type of drive The basic operation of the cycloconverter is illustrated
is best suited to high performance applications such as by the fact that 9 controllable switches enable any
mixers, conveyors, lifts, cranes and winders, as well as portion of a 3-phase supply to be connected to any
fans and pumps. phase of a 3-phase load. Consequently a 3-phase
variable voltage and frequency output can by
synthesised directly, i.e. without the intermediate d.c.
SYNCHRONOUS DRIVES link. In practice, each phase requires a “positive” bridge
for operation when the supply voltage is positive and a
“negative” bridge for operation when the supply voltage
All of the drives previously described have been is negative, Figure 9. Hence each phase of the motor
designed for operation with a standard cage induction requires its own double bridge converter, consisting of
motor. The drive described in this section is designed 12 thyristors, and so the three phase drive requires a
specifically for operation with a synchronous motor. total of 36 devices. Although this represents a high
device count, the advantage is that natural commutation
Method of operation means that standard converter grade devices can be used
The basic drive, Figure 8, is similar to the CSI drive. and there are no forced commutation circuits.
The input bridge is a fully controlled rectifier, the d.c.
link has a smoothing inductor and the output bridge is a Characteristics and applications
standard thyristor inverter. There are however, two Cycloconverters are only capable of producing
major differences. Firstly the switching points of the acceptable output waveforms at frequencies well below
inverter devices are set by a shaft encoder on the motor. the 50 Hz supply system value, typically one third of
Secondly, commutation of the inverter devices is mains frequency. Again, power ratings are theoretically
achieved by the use of the back EMF generated by the unlimited. Drives with converter voltages of up to 18
motor field. Natural commutation allows the use of kV and rated currents of up to 2500 A are available.
standard converter grade devices.
Typical applications are in traction and haulage. The
Characteristics and applications combination of multi-pole, slow speed motors and the
The d.c. link voltage is reversible, and so regeneration infinitely variable low frequency drive makes this a
is possible without the use of any additional converters. good option as a very low speed, direct drive for mine-
High speeds up to 10000 rev/min are possible and high winding, kiln and crusher applications.
power ratings are also possible, theoretically unlimited.
Converter voltages of up to 18 kV and rated currents of
up to 2000 A are available. SLIP-ENERGY RECOVERY DRIVES

The use of the synchronous motor tends to determine

that this drive combination is not economically viable Through the use of a variable rotor resistance, it has
below ratings of approximately 250 kW. This is partly always been possible to effect speed control of the
as a result of the fact that synchronous motors are not as wound-rotor induction motor. The disadvantage of this
readily available at low power ratings. This drive has an method has been the dissipation of the rotor power in
advantage over the induction motor drive because the the external rotor resistance. The slip energy recovery
running speed is a direct function of the applied drive (or Kramer Drive) achieves the same controlling
function whilst recovering the rotor power and feeding control systems relying on a shaft position and speed
it back into the supply system. feedback signal, in addition to a motor model, to
generate a modulated output. The latest development to
Method of operation emerge is the so called Direct Torque Control Drive
The slip-frequency a.c. rotor power is rectified into d.c. which is an open-loop speed control system which uses
and a smoothed d.c. link voltage established, Figure 10. advanced motor theory to calculate the torque directly,
A three phase inverter bridge converts the rotor power without using modulation. This type of drive offers
to a.c. and then the power is returned to the a.c. source, even better dynamic speed performance than the vector
usually via a transformer. The firing angle of the control drive.
inverter devices is used to control the d.c. link voltage.
In turn, this controls the slip, and hence the speed of the Application specific drives are also available, such as
motor. This method is very efficient as the rotor power the current source inverter, the synchronous motor
is returned to the supply. drive, the cycloconverter and the slip energy recovery
scheme. Each of these has a fairly limited range of,
Characteristics and applications usually high power, applications to which they are best
This method of speed control is economical when suited.
compared with a standard inverter drive because the
converters only carry the slip power of the rotor, not the This paper has reviewed the principal types of VSD
fully rated stator power. Consequently the converters available today. It has sought to explain the principles
are much smaller than if they were placed in the stator of operation of each and to identify their key
circuit. The proviso here is that the speed range of the characteristics. The principal applications of each drive
application is limited to half speed to full speed (or less, have also been given. It is acknowledged that this is not
often 80 -100% speed) otherwise the rotor power will an exhaustive summary of all of the available drives.
begin to approach the rated stator power. Such drives New control schemes are emerging all of the time
are available in the 1 - 20 MW power range with which offer a plethora of variations on the main themes.
converter voltages and currents to suit the application. Future developments will undoubtedly raise the
standards of performance ever upwards in a constantly
By far the most common application for this drive is in changing industry.
the field of moving fluids, where the cube law relating
power to speed applies. Hence typical applications are BIBLIOGRAPHY
large centrifugal fans and pumps.
ETSU, 1994, “Introduction and Review of VSD’s to
Improve Energy Efficiency”, COMETT short course
CONCLUSION notes.

Proceedings of the IEE Colloquium on “Advances in

The use of a Variable Speed Drive for a speed control Control Systems for Electric Drives”, IEE Digest No
application usually offers an energy efficient and 1995/114, London UK.
environmentally friendly solution. The best
opportunities for energy savings, with subsequent Hughes A, 1993, “Electric Motors and Drives” 2nd
economic savings, arise through the laws which govern Edition, Butterworth-Heinemann Ltd , ISBN 0 7506
the operation of centrifugal fans and pumps. Simple and 1741 1.
straightforward VSD’s, such as the PWM inverter
drives, are available for applications where the speed Vithayathil J, 1995, “Power Electronics - principles and
control accuracy is not critical. applications”, McGraw-Hill, ISBN 0 07 067555 4.

More onerous speed control applications may require Cegelec Drives, Standard products and systems drives:
the deployment of the, now standard, vector control application guide.
drive. Such drives compare favourably with their d.c.
counterpart in respect of their dynamic and speed ABB, 1995, Technical Guide No. 1, Direct Torque
holding accuracy. Largely these are closed-loop speed Control.
 Figure 1 - Variable Speed Drive block diagram

Figure 2 - Inverter bridge circuit Figure 3 - Quasi-square wave

Figure 4 - Inverter arrangements; a) PWM, b) VSI, c) CSI.

 Figure 5 - PWM voltage waveform

Figure 6 - Vector Control block diagram

Figure 7 - Direct Torque Control block diagram

 Figure 8 - Synchronous motor drive

Figure 9 - Cycloconverter motor drive

Figure 10 - Slip Energy Recovery drive