VFDs and Motors: Making the right match

Variable frequency drives (VFDs) and electric motors are strange companions: The VFD is a static device, delicate, intolerant of wide variations in environmental conditions; extremely adjustable and controllable by microprocessors; capable of being monitored and controlled from remote locations; and a product of modern electronic engineering and precision—the beauty.


Variable frequency drives (VFDs) and electric motors are strange companions: The VFD is a static device, delicate, intolerant of wide variations in environmental conditions; extremely adjustable and controllable by microprocessors; capable of being monitored and controlled from remote locations; and a product of modern electronic engineering and precision—the beauty.

The motor is a rotating machine, robust and capable of wide variations in environmental conditions; not easily controlled unless assisted by external devices; and a product of nearly two centuries of electrical and mechanical engineering—the beast.

Yet, the two work together successfully to provide a means of speed control of rotating machinery. But before it can happen, there are many issues of compatibility that engineers must carefully evaluate and resolve.

Matching nameplate ratings

The first issue is matching the nameplate ratings. The author has encountered cases where incorrect matching led to incompatibility. Manufacturers nominally rate the VFD in horsepower, but this is only an approximate guide to VFD selection. The actual ratings to be determined are input and output voltage, current, power factor, and frequency.

In one such case at a manufacturing plant, the engineer specified a 25-hp, 480-V, 900-rpm, 3-phase, 60-Hz electric motor to drive a pump at variable speeds of 90 to 900 rpm. The specified VFD had the following ratings:

• Input: 480 V, 3-phase, 60 Hz, 32 amp

• Output: 0 to 460 V, 3-phase, 0 to 66 Hz, 0 to 32 amp.

Facility engineers procured the correct VFD. However, a 25-hp, 480-V, 3-phase, 60-Hz,1,800-rpm motor was available in the warehouse. Believing that the VFD could adjust the motor speed from 90 to 900 rpm, the engineers coupled this motor to the pump. When they started the system, the pump could not be brought up to its full speed. The VFD tripped out at approximately 400 rpm on overload.

At first it was thought that debris in the pump piping caused the overload. The real reason for the tripping was the mismatch in motor and VFD ratings. For an 1,800-rpm, 60-Hz motor to run at 900 rpm, the output frequency of the VFD must be 30 Hz. Because all VFDs are programmed to maintain a constant voltage-to-frequency ratio—the V/f ratio typically is 7.66 for 460-V motors—the output voltage of the VFD must be 230 V. The pump demands a brake horsepower of 25 from the motor at 900 rpm. To produce 25 hp output at 230 V, the motor input current should be 64 amps instead of the rated 32 amps. The VFD input current also should be about 64 amps. The VFD would trip out on overload.

Therefore, a rule to follow is to make sure that at the rated speed of the driven equipment, the nominal ratings of the VFD should be the same as those of the motor.

Harmonic heating

Harmonics in motor currents produce additional heating due to induced high-frequency currents in the rotor bars. Does the motor need to be de-rated when driven by a VFD? In other words, can the motor supply its rated horsepower without overheating? In older current source inverter (CSI) VFDs, which have a large inductor in the dc link, the motor current waveform was a “double-hump” (see Figure 1), rich in the 5th, 7th, and other harmonics. It was common practice to de-rate the motor by 5% to 10%. However, modern pulse-width-modulated (PWM) drives produce almost a sinusoidal motor current except for a small high-frequency component, typically 5 to 10 kHz, because of the PWM carrier switching frequency (see Figure 2). No de-rating is required. To be safe, the VFD manufacturers recommend motors with 1.15 service factor.

Heating due to harmonic currents is not the only consideration when selecting motors to be driven by VFDs. Standard motors that are not inverter duty motors generally are unsuitable because of considerations such as insulation, cooling at low speeds, bearing currents, and voltage spikes.

Variable or constant torque

Manufacturers offer two types of VFDs: variable torque and constant torque. The two are basically the same except that theconstant-torque has a slightly larger kVA rating. The selection depends upon the load driven.

Centrifugal pumps and blowers are variable-torque loads: The torque demanded by the load is reduced as the speed is reduced. In fact, the torque demanded is proportional to the square of the speed. The following proportional relations hold:

Horsepower (hp) a (rpm)3

Torque (ft-lb) a (rpm)2

Flow (gpm) a (rpm)

Therefore, if the pump speed is reduced to 50% of the rated speed, the required torque is reduced to 25% of the rated torque. The torque produced by the motor that drives the pump or fan is proportional to the product of the airgap flux and the stator winding current. The VFD that powers the motor keeps the airgap flux constant at all speeds (constant V/f ratio). Therefore, the motor current also reduces to 25% of the rated current. The motor I2R losses reduce to 6.25% of the losses at the rated current. Motor and VFD would have no problem dissipating the heat.

It's a different story with constant-torque loads, such as rotary and screw compressors, elevator drives, hoists and cranes, and reciprocating pumps. The motor torque, which must be the same as the load torque during steady-state operation, is the same at all speeds, and the motor current remains the same at all speeds. Motor efficiency is lower and heat dissipation is a problem because of the reduced airflow from the shaft-mounted fan.

The problem is more severe in totally enclosed, fan-cooled motors used in Class 1, Division 2 hazardous locations, particularly when the motor operates at reduced speed for extended periods of time. In such cases, an independently driven blower would be necessary to provide adequate airflow.

Flux-vector drive

Many applications demand unusual operating conditions of the electric motor. For example, an elevator or a crane drive requires operation in both directions. In addition, the motor is required to produce a positive torque while rotating in the negative direction and a negative torque while rotating in the positive direction, required for braking when the load is being raised or lowered. Such drives are said to operate in all four quadrants of the speed-torque plane (see Figure 3).

In some other applications the motor is required to produce the rated torque at zero speed, such as a marine winch motor. An ac induction motor cannot meet these requirements without the help of VFDs, but dc motors can. But dc motors are expensive and require constant maintenance because of the commutator and brushes. A flux-vector drive is used in such applications.

From a control point of view, conventional VFDs for pumps and fans provide open-loop control; the VFD supplies a certain voltage at a certain frequency, and the motor rotates at a speed determined by its characteristics. Speed adjustment is either manual or by a remote signal. In a flux-vector drive, there is a continuous feedback of the motor speed and torque to the VFD. The VFD adjusts the voltage and frequency to produce the required operation. The flux-vector drive gives the ac motor the same capabilities as the dc motor. The direction of rotation can be reversed without actually switching the phases. Earlier flux-vector drives needed feedback from a shaft position sensor. Modern drives are sensor-less drives relying on the current, voltage, and speed measurement to compute the torque.

Voltage doubling, critical cable length, and dv/dt

PWM drives produce an output voltage in the form of high-frequency rectangular pulses of the same magnitude. The frequency of the pulses (known as the PWM carrier frequency) is in the range of 4 to 16 kHz. The width of each pulse is modulated such that the instantaneous average value is a sine wave. Figure 4 shows a line-to-line output voltage of a PWM VFD. For illustration, only seven pulses in one half-cycle are shown. Actually, for a typical carrier frequency of 4 kHz, there would be 33 pulses in each half-cycle. The magnitude of each pulse is equal to the dc link voltage of the VFD (typically 648 V in 480-V 6-pulse drives). The pulses are rectangular with a rate of rise of approximately 5,000 V/microsec. Higher frequency in the range of 4 to 16 kHz makes the motor current more sinusoidal, reduces the noise in the motor, and reduces the switching losses in the VFD transistors, thus increasing the efficiency of the VFD. However, the high-frequency pulses create greater stress in the motor insulation.

The train of pulses invading the motor causes three deleterious effects. First, each pulse nearly doubles its magnitude after it hits the motor winding because of the higher impedance of the motor winding as compared to that of the cable. There is a traveling wave phenomenon here. The pulses travel toward the motor at a velocity of approximately 330 ft/microsec (approximately one-third the speed of light). The peak voltage impressed upon the winding at the leading edge of the pulse depends upon the rate of rise of the voltage and the velocity of propagation. If the length of the cable is more than a certain length—called the “critical length”—then the voltage will be twice the magnitude of the pulse (the full doubling effect). The critical cable length is determined by the following equation:

Lcritical = v*Tr/2, where

v = velocity of propagation, ft/microsec

Tr= rise time, microsec

The velocity of propagation depends upon the size and type of the cable between the VFD and the motor. For example, in a 1/0 1/C unshielded cable in a rigid galvanized steel conduit, the velocity is 333 ft/microsec. The rise time of the pulse depends upon the PWM carrier frequency. Typically it is in the range of 0.2 to 0.8 microsec.

A second injurious effect of PWM pulses is the high-frequency ringing at the leading and trailing edge of each pulse (see Figure 5) due to the stray capacitances of the motor winding, which accelerates bearing wear. Finally, the high rate of rise of voltage stresses the insulation of the motor winding end turns to a greater extent than the other parts of the winding. This may cause partial discharge—a localized corona or arcing that degrades the winding insulation and eventually leads to insulation failure—at the end turns.

The voltage appearing at the motor is not the same as the output voltage of the VFD. Figure 5 shows the doubling effect, which will occur when the length of the cable from the VFD to the motor is greater than the critical length. The figure also shows high-frequency ringing due to stray capacitances.

Because the high rate of rise of the voltage causes additional stress on the winding insulation, many VFD manufacturers offer a dv/dt filter, which is essentially a shunt capacitor that slowly reduces the rate of rise. (The term “dv/dt” refers to a differential in voltage with respect to a differential in time.) An inverter duty motor is designed to withstand the additional harmonic heating and the increased insulation stress due to the PWM VFD.

Bearing currents

There are two mechanisms that cause circulating currents, accelerating wear of the motor bearings: electromagnetic-induced currents in the shaft and the bearings; and electrostatic-induced voltage between the bearing balls or rollers and the stationary surfaces. Both mechanisms are enhanced when the motor is powered by a VFD.

There is a small electromagnetic-induced voltage in all motor shafts in the axial direction caused by the leakage of magnetic flux linking the rotor and the shaft outside the airgap. Because of the slight static and dynamic eccentricity of the shaft, the leakage flux induces a voltage that establishes a circulating current through the shaft, the two bearings, and the frame (see Figure 6). The circulating current is small and normally is not a concern. Its path is broken by insulating one of the two bearings, typically the non%%MDASSML%%drive-end bearing. However, if there are harmonics in the motor current, such as in the case of CSI VFDs, the circulating currents increase and can cause local heating in the bearing. Because most PWM drives provide a nearly sinusoidal current to the motor, they only slightly aggravate the circulating current problem.

It is the electrostatic-induced voltage that has led to several bearing failures. The stray capacitances, which normally have negligible effect, come into play when high-frequency pulses are impressed upon the winding. Figure 7 shows the relevant capacitors and the voltage divider formed by these capacitances. A high-frequency voltage VB appears across the film of lubricant between the bearing balls and the stationary surface. The ratio VB/V is called the bearing voltage ratio (BVR). The BVR typically is 3% to 10%. If it is high enough, the voltage VB causes a dielectric breakdown of the lubricant film creating a current called the electric discharge machining, or EDM, current. The breakdown is in the form of a tiny arc between the rolling elements and the bearing race. The threshold voltage for the arcing to occur is approximately 5 to 30 V. The arcing causes bearing erosion at an accelerated rate, and surface irregularities (called fluting) appear in the bearing race. The irregularities become progressive until the bearing fails.

The question is how to minimize the damage to the bearing. A low-PWM carrier frequency and suppression of the ringing helps. An output filter or a dv/dt filter reduces the rate of rise of voltage of each pulse and also reduces the capacitive-induced voltage in the shaft. Shielded cables between the VFD and the motor also assists in reducing the dv/dt. Insulating both the drive-end bearing and the NDE bearing reduces the BVR. Finally, a shaft grounding brush is an economical way of preventing the arcing in the bearing.

How to ensure compatibility

A few suggestions on how to engineer a successfully matched VFD-motor pair are:

  • Know the type of load and its torque-speed characteristic to correctly size the motor. Be aware of the heat dissipation problem in constant torque loads at low speeds. Specify the motor to have Class F insulation (239 F rise) but to have Class B temperature rise (194 F rise) at the rated full load.

  • Specify inverter-duty motors and reference NEMA MG-1 Section IV Part 31.

  • Estimate the length of the cable from the VFD to the motor and supply this information to the VFD vendor.

  • Specify an output filter and/or a dv/dt filter, particularly for medium-voltage drives.

  • Specify insulated motor bearings and a shaft grounding brush arrangement, particularly for medium-voltage drives.

Some manufacturers offer an integrated VFD-motor package. This product transfers the responsibility of matching to the manufacturer. Use this product wherever it is possible to locate the VFD close to the motor.

Author Information

Peeran has more than 20 years of experience in the design of electrical distribution systems. For several years he was an adjunct professor at Northeastern University, Boston, and is a member of the CSE editorial advisory board.

No comments
Consulting-Specifying Engineer's Product of the Year (POY) contest is the premier award for new products in the HVAC, fire, electrical, and...
Consulting-Specifying Engineer magazine is dedicated to encouraging and recognizing the most talented young individuals...
The MEP Giants program lists the top mechanical, electrical, plumbing, and fire protection engineering firms in the United States.
40 Under 40; Performance-based design; Clean agent fire suppression; NFPA 92; Future of commissioning; Successful project management principles
BIM coordination; MEP projects; NFPA 13; Data center Q&A; Networked lighting controls; 2017 Product of the Year finalists
Emergency lighting; NFPA 3 and 4; Integrated building systems; Smart lighting, HVAC design
Commissioning electrical systems; Designing emergency and standby generator systems; VFDs in high-performance buildings
Tying a microgrid to the smart grid; Paralleling generator systems; Previewing NEC 2017 changes
Driving motor efficiency; Preventing Arc Flash in mission critical facilities; Integrating alternative power and existing electrical systems
As brand protection manager for Eaton’s Electrical Sector, Tom Grace oversees counterfeit awareness...
Amara Rozgus is chief editor and content manager of Consulting-Specifier Engineer magazine.
IEEE power industry experts bring their combined experience in the electrical power industry...
Michael Heinsdorf, P.E., LEED AP, CDT is an Engineering Specification Writer at ARCOM MasterSpec.
Automation Engineer; Wood Group
System Integrator; Cross Integrated Systems Group
Fire & Life Safety Engineer; Technip USA Inc.
This course focuses on climate analysis, appropriateness of cooling system selection, and combining cooling systems.
This course will help identify and reveal electrical hazards and identify the solutions to implementing and maintaining a safe work environment.
This course explains how maintaining power and communication systems through emergency power-generation systems is critical.
click me