Wednesday, May 23, 2018

IAM Search

Final Control Elements - Variable-Speed Motor Controls

An alternative method of flow control in lieu of control valves is to vary the speed of the machine(s) motivating fluid to flow. In the case of liquid flow control, this would take the form of variable-speed pumps. In the case of gas flow control, it would mean varying the speed of compressors or blowers. Flow control by machine speed control makes a lot of sense for some process applications. It is certainly more energy-efficient to vary the speed of the machine pushing fluid to control flow, as opposed to letting the machine run at full speed all the time and adjusting flow rate by throttling the machine’s discharge (outlet) or recycling fluid back to the machine’s suction (inlet). The fact that the system has one less component in it (no control valve) also reduces capital investment and potentially increases system reliability:

Further advantages of machine speed control include the ability to “soft-start” the machine instead of always accelerating rapidly from a full stop to full speed, reduced wear on machines due to less motion over time, and reduced vibration. In applications such as conveyor belt control, robotic machine motion control, and electric vehicle propulsion, variable-speed technology makes perfect sense because the prime mover device is already (in most cases) an electric motor, with precise speed control of that motor providing many practical benefits. In some applications, regenerative braking may be of benefit, where the motor is used as an electrical generator to slow down the machine on command. Regenerative braking transfers kinetic energy within the machine to the power grid where it may be gainfully used in other processes, saving energy and reducing wear on any mechanical (friction) brakes already installed in the machine.

With all these advantages inherent to variable-speed pumps, fans, and compressors (as opposed to using dissipative control valves), one might wonder, “Why would anyone ever use a control valve to regulate flow? Why not just control all fluid flows using variable-speed pumping machines?” Several good answers exist to this question:

  • Variable-speed machines often cannot respond as rapidly as control valves

  • Control valves have the ability to positively halt flow; a stopped pump or blower will not necessarily prevent flow from going through

  • Some process applications must contain a dissipative element in order for the system to function (e.g. let-down valves in closed refrigeration systems)

  • Split-ranging may be difficult or impossible to achieve with multiple machine speed control

  • Limited options for fail-safe status

  • In many cases, there is no machine dedicated to a particular flow path (e.g. a pressure release valve)

 

DC Motor Speed Control

DC electric motors generate torque by a reaction between two magnetic fields: one field established by stationary “field” windings (coils), and the other by windings in the rotating armature. Some DC motors lack field windings, substituting large permanent magnets in their place so that the stationary magnetic field is constant for all operating conditions.

In any case, the operating principle of a DC electric motor is that current passed through the armature creates a magnetic field that tries to align with the stationary magnetic field. This causes the armature to rotate:


However, a set of segmented copper strips called a commutator breaks electrical contact with the now-aligned coil and energizes another coil (or in the simple example shown above, it re-energizes the same loop of wire in the opposite direction) to create another out-of-alignment magnetic field that continues to rotate the armature. Electrical contact between the rotating commutator segments and the stationary power source is made through carbon brushes. These brushes wear over time (as does the commutator itself), and must be periodically replaced.

Most industrial DC motors are built with multiple armature coils, not just one as shown in the simplified illustration above. A photograph of a large (1250 horsepower) DC motor used to propel a ferry ship is shown here, with the field and armature poles clearly seen (appearing much like spokes in a wheel):

 

A close-up of one brush assembly on this large motor shows both the carbon brush, the brush’s spring-loaded holder, and the myriad of commutator bars the brush makes contact with as the armature rotates:

 

DC motors exhibit the following relationships between mechanical and electrical quantities:

Torque:

  • Torque is directly proportional to armature magnetic field strength, which in turn is directly proportional to current through the armature windings

  • Torque is also directly proportional to the stationary pole magnetic field strength, which in turn is directly proportional to current through the field windings (in a motor with non-permanent field magnets)

Speed:

  • Speed is limited by the counter-EMF generated by the armature as it spins through the stationary magnetic field. This counter-EMF is directly proportional to armature speed, and also directly proportional to stationary pole magnetic field strength (which is directly proportional to field winding current in a motor that is not permanent-magnet)

  • Thus, speed is directly proportional to armature voltage

  • Speed is also inversely proportional to stationary magnetic field strength, which is directly proportional to current through the field windings (in a motor with non-permanent field magnets)

A very simple method for controlling the speed and torque characteristics of a wound-field (nonpermanent magnet) DC motor is to control the amount of current through the field winding:


Decreasing the field control resistor’s resistance allows more current through the field winding, strengthening its magnetic field. This will have two effects on the motor’s operation: first, the motor will generate more torque than it did before (for the same amount of armature current) because there is now a stronger magnetic field for the armature to react against; second, the motor’s speed will decrease because more counter-EMF will be generated by the spinning armature for the same rotational speed, and this counter-EMF naturally attempts to equalize with the applied DC source voltage. Conversely, we may increase a DC motor’s speed (and reduce its torque output) by increasing the field control resistor’s resistance, weakening the stationary magnetic field through which the armature spins.

Regulating field current may alter the balance between speed and torque, but it does little to control total motor power. In order to control the power output of a DC motor, we must also regulate armature voltage and current. Variable resistors may also be used for this task, but this is generally frowned upon in modern times because of the wasted power.

A better solution is to have an electronic power control circuit very rapidly switch transistors on and off, switching power to the motor armature. This is called pulse-width modulation, or PWM.

 

The duty cycle (on time versus on+off time) of the pulse waveform will determine the fraction of total power delivered to the motor:


Such an electronic power-control circuit is generally referred to as a drive. Thus, a variable-speed drive or VSD is a high-power circuit used to control the speed of a DC motor. Motor drives may be manually set to run a motor at a set speed, or accept an electronic control signal to vary the motor speed in the same manner an electronic signal commands a control valve to move. When equipped with remote control signaling, a motor drive functions just like any other final control element: following the command of a process controller in order to stabilize some process variable at setpoint.

An older technology for pulsing power to a DC motor is to use a controlled rectifier circuit, using SCRs instead of regular rectifying diodes to convert AC to DC. Since the main power source of most industrial DC motors is AC anyway, and that AC must be converted into DC at some point in the system, it makes sense to integrate control right at the point of rectification:


Controlled rectifier circuits work on the principle of varying the “trigger” pulse times relative to the AC waveform pulses. The earlier the AC cycle each SCR is triggered on, the longer it will be on to pass current to the motor. The “phase control” circuitry handles all this pulse timing and generation.

A DC motor drive that simply varied power to the motor according to a control signal would be crude and difficult to apply to the control of most processes. What is ideally desired from a variable-speed drive is precise command over the motor’s speed. For this reason, most VSDs are designed to receive feedback from a tachometer mechanically connected to the motor shaft, so the VSD “knows” how fast the motor is turning. The tachometer is typically a small DC generator, producing a DC voltage directly proportional to its shaft speed (0 to 10 volts is a common scale). With this information, the VSD may throttle electrical power to the motor as necessary to achieve whatever speed is being commanded by the control signal. Having a speed-control feedback loop built into the drive makes the VSD a “slave controller” in a cascade control system, the drive receiving a speed setpoint signal from whatever process controller is sending an output signal to it:

 

A photograph of the tachogenerators (dual, for redundancy) mechanically coupled to that large 1250 horsepower ferry ship propulsion motor appears here:


The SCRs switching power to this motor may be seen here, connected via twisted-pair wires to control boards issuing “firing” pulses to each SCR at the appropriate times:


The integrity of the tachogenerator feedback signal to the VSD is extremely important for safety reasons. If the tachogenerator becomes disconnected – whether mechanically or electrically (it doesn’t matter) – from the drive, the drive will “think” the motor is not turning. In its capacity as a speed controller, the drive will then send full power to the DC motor in an attempt to get it up to speed. Thus, loss of tachogenerator feedback causes the motor to immediately “run away” to full speed. This is undesirable at best, and likely dangerous in the case of motors as large as the one powering this ship.

As with all forms of electric power control based on pulse durations and duty cycles, there is a lot of electrical “noise” cast by VSD circuits. Square-edged pulse waveforms created by the rapid on-and-off switching of the semiconductor power devices are equivalent to infinite series of high-frequency sine waves1, some of which may be of high enough frequency to self-propagate through space as electromagnetic waves. This radio-frequency interference or RFI may be quite severe given the high power levels of industrial motor drive circuits. For this reason, it is imperative that neither the motor power conductors nor the conductors feeding AC power to the drive circuit be routed anywhere near small-signal or control wiring, because the induced noise will wreak havoc with whatever systems utilize those low-level signals.

RFI noise on the AC power conductors may be mitigated by routing the AC power through filter circuits placed near the drive. The filter circuits block high-frequency noise from propagating back to the rest of the AC power distribution wiring where it may influence other electronic equipment. However, there is little that may be done about the RFI noise between the drive and the motor other than to shield the conductors in well-grounded metallic conduit.

 

AC Motor Speed Control

AC induction motors are based on the principle of a rotating magnetic field produced by a set of stationary windings (called stator windings) energized by AC power of different phases. The effect is not unlike a series of blinking light bulbs which appear to “move” in one direction due to intentional sequencing of the blinking. If sets of wire coils (windings) are energized in a like manner – each coil reaching its peak field strength at a different time from its adjacent neighbor – the effect will be a magnetic field that “appears” to move in one direction. If these windings are oriented around the circumference of a circle, the moving magnetic field rotates about the center of the circle, as illustrated by this sequence of images (read left-to-right, top-to-bottom, as if you were reading words in a sentence):

 

 

 

Any magnetized object placed in the center of this circle will attempt to spin at the same rotational speed as the rotating magnetic field. Synchronous AC motors use this principle, where a magnetized rotor precisely follows the magnetic field’s speed.

Any electrically conductive object placed in the center of the circle will experience induction as the magnetic field direction changes around the conductor. This will induce electric currents within the conductive object, which in turn will react against the rotating magnetic field in such a way that the object will be “dragged along” by the field, always lagging a bit in speed. Induction AC motors use this principle, where a non-magnetized (but electrically conductive) rotor rotates at a speed slightly less2 than the synchronous speed of the rotating magnetic field.

The rotational speed of this magnetic field is directly proportional to the frequency of the AC power, and inversely proportional to the number of poles in the stator:


Where,

  S = Synchronous speed of rotating magnetic field, in revolutions per minute (RPM)

  f = Frequency, in cycles per second (Hz)

  n = Total number of stator poles per phase (the simplest possible AC motor design will have 2 poles)

 

While the number of poles in the motor’s stator is a quantity fixed at the time of the motor’s manufacture, the frequency of power we apply may be adjusted with the proper electronic circuitry. A high-power circuit designed to produce varying frequencies for an AC motor to run on is called a variable-frequency drive, or VFD.

A simplified schematic diagram for a VFD is shown here, with a rectifier section on the left (to convert AC input power into DC), a filter capacitor to “smooth” the rectified DC power, and a transistor “bridge” to switch DC into AC at whatever frequency is desired to power the motor. The transistor control circuitry has been omitted from this diagram for the sake of simplicity:

 

 

As with DC motor drives (VSDs), the power transistors in an AC drive (VFD) switch on and off very rapidly with a varying duty cycle. Unlike DC drives, however, the duty cycle of an AC drive’s power transistors must vary rapidly in order to synthesize an AC waveform from the DC “bus” voltage following the rectifier. A DC drive circuit’s PWM duty cycle controls motor power, and so it will remain at a constant value when the desired motor power is constant. Not so for an AC motor drive circuit: its duty cycle must vary from zero to maximum and back to zero repeatedly in order to generate an AC waveform for the motor to run on.

The equivalence between a rapidly-varied pulse-width modulation (PWM) waveform and a sine wave is shown in the following illustration:

 

This concept of rapid PWM transistor switching allows the drive to “carve” any arbitrary waveform out of the filtered DC voltage it receives from the rectifier. Virtually any frequency may be synthesized (up to a maximum limited by the frequency of the PWM pulsing), and any voltage (up to a maximum peak established by the DC bus voltage), giving the VFD the ability to power an induction motor over a wide range of speeds.

While frequency control is the key to synchronous and induction AC motor speed control, it is generally not enough on its own. While the speed of an AC motor is a direct function of frequency (controlling how fast the rotating magnetic field rotates around the circumference of the stator), torque is approximately proportional to stator winding current. Since the stator windings are inductors by nature, their reactance varies with frequency as described by the formula XL = 2πfL. Thus, as frequency is increased, winding reactance increases right along with it. This increase in reactance results in a decreased stator current (assuming the AC voltage is held constant as frequency is increased). This can cause undue torque loss at high speeds, and excessive torque (as well as excessive stator heat!) at low speeds. For this reason, the AC voltage applied to a motor by a VFD is usually made to vary in direct proportion to the applied frequency, so that the stator current will remain within good operating limits throughout the speed range of the VFD. This correspondence is called the voltage-to-frequency ratio, or V/F ratio.

Variable-frequency motor drives are manufactured for industrial motor control in a wide range of sizes and horsepower capabilities. Some VFDs are small enough to hold in your hand, while others are large enough to require a freight train for transport. The following photograph shows a pair of moderately-sized Allen-Bradley VFDs (about 100 horsepower each, standing about 4 feet high), used to control pumps at a wastewater treatment plant:

 

 

Variable-frequency AC motor drives do not require motor speed feedback the way variable-speed DC motor drives do. The reason for this is quite simple: the controlled variable in an AC drive is the frequency of power sent to the motor, and rotating-magnetic-field AC motors are frequency-controlled machines by their very nature. For example, a 4-pole AC induction motor powered by 60 Hz has a base speed of 1728 RPM (assuming 4% slip). If a VFD sends 30 Hz AC power to this motor, its speed will be approximately half its base-speed value, or 864 RPM. There is really no need for speed-sensing feedback in an AC drive, because the motor’s real speed will always be limited by the drive’s output frequency. To control frequency is to control motor speed for AC synchronous and induction motors, so no tachogenerator feedback is necessary for an AC drive to “know” approximately3 how fast the motor is turning. The non-necessity of speed feedback for AC drives eliminates a potential safety hazard common to DC drives: the possibility of a “runaway” event where the drive loses its speed feedback signal and sends full power to the motor.

As with DC motor drives, there is a lot of electrical “noise” cast by VFD circuits. Square-edged pulse waveforms created by the rapid on-and-off switching of the power transistors are equivalent to infinite series of high-frequency sine waves4, some of which may be of high enough frequency to self-propagate through space as electromagnetic waves. This radio-frequency interference or RFI may be quite severe given the high power levels of industrial motor drive circuits. For this reason, it is imperative that neither the motor power conductors nor the conductors feeding AC power to the drive circuit be routed anywhere near small-signal or control wiring, because the induced noise will wreak havoc with whatever systems utilize those low-level signals.

RFI noise on the AC power conductors may be mitigated by routing the AC power through filter circuits placed near the drive. The filter circuits block high-frequency noise from propagating back to the rest of the AC power distribution wiring where it may influence other electronic equipment. However, there is little that may be done about the RFI noise between the drive and the motor other than to shield the conductors in well-grounded metallic conduit.

 

Motor Drive Features

Modern DC and AC motor drives provide features useful when using electric motors as final control elements. Some common features seen in both VSDs and VFDs are listed here:

  • Speed limiting

  • Torque limiting

  • Torque profile curves (used to regulate the amount of torque available at different motor speeds)

  • Acceleration (speed rate-of-change) limiting

  • Deceleration (speed rate-of-change) limiting

  • Dynamic braking (turning the motor into an electromagnetic brake5)

  • Plugging (applying reverse-direction power to a motor to quickly stop it)

  • Regenerative braking (turning the motor into a generator to recover kinetic energy)

  • Overcurrent monitoring and automatic shut-down

  • Overvoltage monitoring and automatic shut-down

  • PWM frequency adjustment (may be helpful in reducing electromagnetic interference with some equipment)

Not only are some of these limiting parameters useful in extending the life of the motor, but they may also help extend the operating life of the mechanical equipment powered by the motor. It is certainly advantageous, for example, to have torque limiting on a conveyor belt motor, so that the motor does not apply full rated torque (i.e. stretching force) to the belt during start-up.

If a motor drive is equipped with digital network communication capability (e.g. Modbus), it is usually possible for a host system such as a PLC or DCS to update these control parameters as the motor is running.

 

Metering Pumps

A very common method for directly controlling low flow rates of fluids is to use a device known as a metering pump. A “metering pump” is a pump mechanism, motor, and drive electronics contained in a monolithic package. Simply supply 120 VAC power and a control signal to a metering pump, and it is ready to use.

Metering pumps are commonly used in water treatment processes to inject small quantities of treatment chemicals (e.g. coagulants, disinfectants, acid or caustic liquids for pH neutralization, corrosion-control chemicals) into the water flowstream, as is the Milton-Roy unit shown in this photograph:

 

 

Adjustment knobs on the front of the pump establish the maximum flow rate at a control signal value of 100%:

 

 

 

While some metering pumps use rotary motor and pump mechanisms, many use a “plunger” style mechanism operated by a solenoid at variable intervals. Thus, the latter type of metering pump does not provide continuous flow control, but rather a flow consisting of discrete pulse events distributed over a period of time. The “plunger” metering pumps are quite simple and reliable, and are entirely appropriate if non-continuous flow is permissible for the process.

 

1This equivalence was mathematically proven by Jean Baptiste Joseph Fourier (1768-1830), and is known as a Fourier series.

2The difference between the synchronous speed and the rotor’s actual speed is called the motor’s slip speed.

3For more precise control of AC motor speed (especially at low speeds where slip speed becomes a greater percentage of actual speed), speed sensors may indeed be necessary.

4This equivalence was mathematically proven by Jean Baptiste Joseph Fourier (1768-1830), and is known as a Fourier series.

5This is accomplished in very different ways for DC versus AC motors. To dynamically brake a DC motor, the field winding must be kept energized while a high-power load resistor is connected to the armature. As the motor turns, the armature will push current through the resistor, generating a braking torque as it does. To dynamically brake an AC motor, a relatively small DC current is passed through the stator windings, causing large braking currents to be induced in the rotor.

 

References

Baumann, Hans D., Control Valve Primer, A User’s Guide, Second Edition, Instrument Society of America, Research Triangle Park, NC, 1994.

“Cavitation in Control Valves”, document L351 EN, Samson AG, Frankfurt, Germany.

Control Valve Handbook, Third Edition, Fisher Controls International, Inc., Marshalltown, IA, 1999.

Control Valve Sourcebook – Power & Severe Service, Fisher Controls International, Inc., Marshalltown, IA, 1988.

“Design ED, EAD, and EDR Sliding-Stem Control Valves”, Product Bulletin 51.1:ED, Fisher, Marshalltown, IA, 2006.

Grumstrup, Bruce, Considerations in the Design and Selection of Bellows Seal Equipment Valves, Technical Monograph 37, Fisher Controls International Inc., Marshalltown, IA, 1991.

Hutchison, J.W., ISA Handbook of Control Valves, Second Edition, Instrument Society of America, Research Triangle Park, NC, 1976.

Irwin, J. David, The Industrial Electronics Handbook, CRC Press, Boca Raton, FL, 1997.

Jury, Floyd D., Fundamentals of Aerodynamic Noise in Control Valves, Technical Monograph 43, Fisher Controls International Inc., Marshalltown, IA, 1999.

Lipt´ak, B´ela G., Instrument Engineers’ Handbook – Process Control Volume II, Third Edition, CRC Press, Boca Raton, FL, 1999.

“Micro Trims for Globe and Angle Valve Applications”, Product Bulletin 80.4:010, Emerson Process Management, Marshalltown, IA, 2005.

“Packing Selection Guidelines for Sliding-Stem Valves”, Product Bulletin 59.1:062, Emerson Process Management, Marshalltown, IA, 2007.

“Pipeline Accident Report – Pipeline Rupture and Subsequent Fire in Bellingham, Washington June 10, 1999”, NTSB/PAR-02/02, PB2002-916502, Notation 7264A, National Transportation Safety Board, Washington DC, 2002.

Richardson, Jonathan W., Primary Seat Shutoff, Technical Monograph 47, Fisher Controls International LLC, Marshalltown, IA, 2005.

Riveland, Marc, Fundamentals of Valve Sizing for Liquids, Technical Monograph 30, Fisher Controls International Inc., Marshalltown, IA, 1985.

Schafbuch, Paul, Fundamentals of Flow Characterization, Technical Monograph 29, Fisher Controls International Inc., Marshalltown, IA, 1985.

Warnett, Chris, Using Valve Actuators as Predictive Maintenance Tools for MOVs, Rotork Controls, Inc., Rochester, NY, 2000.

“Valve Sizing Technical Bulletin”, document MS-06-84-E, revision 3, Swagelok Company, MI, 2002.

Go Back to Lessons in Instrumentation Table of Contents


Comments (0)Add Comment

Write comment

security code
Write the displayed characters


busy

Related Articles

Promotions

  • ...more

Disclaimer

Important: All images are copyrighted to their respective owners. All content cited is derived from their respective sources.

Contact us for information and your inquiries. IAMechatronics is open to link exchanges.

IAMechatronics Login