VFD, motor strategies for energy efficiency

A variable frequency drive (VFD) often is specified to reduce operational cost for pumps, fans, compressors, or any similar equipment with variable load profiles that may be found in a typical building. Here’s how to specify a VFD to meet load conditions while achieving efficiency.
By John Yoon, PE, LEED AP, McGuire Engineers Inc., Chicago January 21, 2015

This article has been peer-reviewed.Learning objectives

  • Review the evolution of efficiency standards and how they relate to the specification of VFDs and motors.
  • Understand how a standard induction motor uses (and wastes) electricity.
  • Introduce some emerging standards and technologies related to VFDs and motors.
  • Examine common pitfalls in specifying motors and VFDs.

Figure 1: One emerging application for permanent magnet ac (PMAC) motors is in cooling towers. This picture shows a 20 hp totally enclosed, air over (TEAO) PMAC motor directly coupled to the fan instead of using a typical belt drive or gearbox; the view is from inside a cooling tower. This particular tower is a two-cell, cooling tower with 3,000 gpm total capacity. The photo is looking up at a low-speed high-volume fan at the top of the tower. . Elimination of the gearbox dramatically improves the efficiency and reduces maintenance requirements. All graphics courtesy: McGuire EngineersMotors get no respect. Although they’re present in many of the systems that we design and they often drive the overall energy usage profile of a building, they usually end up as an afterthought. Traditionally, mechanical, electrical, plumbing (MEP), and fire protection designs for buildings have been focused on system level considerations of health/safety, functionality, and initial capital cost. However, the adoption of more stringent energy codes and standards has put greater emphasis on energy efficiency in our designs.

While this emphasis on energy efficiency may seem like a relatively recent development, it is the direct evolution of federal legislation that was passed almost 40 years ago in response to the 1973-74 oil crisis. The key is to recognize that this is all part of an ongoing progression and that efficiency requirements will only become more stringent. So while many building owners will be more than satisfied with minimal code-compliant designs, as proactive engineers and designers, there is a need to understand how the components in the systems we design use power and how they can be optimized without compromising those traditional design values. In some cases, there may be technological “dead ends” in the individual system components that may require emphasis on different solutions for energy efficiency.

Motors are everywhere

In most design methodologies, you look for the “bang for the buck”—high-impact, low-cost solutions first. The question is, what uses the most electricity in our designs? Although we may not think of motors specifically when considering the energy use of any particular building system, the electrical usage associated with motors is "hidden" in most commercial building energy use categories.

From a broader perspective, electric motor driven systems represent more than a third of the total electricity demand for the United States and between 43% and 46% globally per statistics from the International Energy Agency. Total motor energy usage for the industrial sector outstrips commercial usage by roughly 3:1. Of total power used by motors worldwide, approximately 68% is used by medium-sized motors from 1 to 500 hp, which covers the vast majority of motors used in building systems.

A business concept known as "disruptive innovation" describes an innovation that redefines or replaces a market by offering improved simplicity, functionality, and affordability. They typically are not breakthrough innovations—in fact, they existed in niche markets prior to widespread acceptance. This concept could be applied to the emergence of variable frequency drives (VFDs) and permanent magnetic ac (PMAC) motors for use in building systems. VFDs have been used for a significant amount of time in situations where controllability was as much a driving factor as energy savings, such as in variable air volume (VAV) air handling systems. However, energy codes, such as ASHRAE Standard 90.1: Energy Standard for Buildings Except Low-Rise Residential Buildings and the International Energy Conservation Code (IECC), have redefined design priorities by emphasizing energy-saving aspects of their use. This emphasis only becomes more pronounced with improvements to the baseline energy model in each code revision cycle. The commercial baseline model in ASHRAE Standard 90.1-2010 (IECC-2012) represents a 30% improvement over ASHRAE 90.1-2004 (IECC-2006). The 2013/2015 code update is expected to produce an additional 8.7% energy cost improvement.

The federal government has acted as the primary catalyst for this emphasis on energy. The Energy Policy and Conservation Act of 1975 (EPCA-1975), with subsequent major amendments by the Energy Policy Act of 1992 (EPAct-1992) and the Energy Independence and Security Act of 2007 (EISA-2007), has given the Dept. of Energy wide-ranging power in mandating energy efficiency standards. While federally mandating energy efficiency may sound intrusive on the surface, the process of adopting new standards does have a reasonable set of formal evaluation criteria, many of which are readily applicable to engineering design. The seven criteria based on EPCA legislation are:

  1. Economic impact on manufacturers and consumers: Will implementing the standard cause unusual hardship for manufacturers or consumers? Is the initial first cost of implementation affordable?
  2. Lifetime operating cost savings compared to increased cost for the product: What is the payback period? What is the lifecycle cost analysis?
  3. Energy savings resulting from implementation: Quantify the magnitude of energy savings if implemented; a national-level consideration.
  4. Lessening of utility or performance of products: Given the function, is this a misapplication of the technology? Am I attempting to tow a trailer with a Ferrari?
  5. Impact of any lessening of competition: Is this simply adopting typical "best in class" products or something proprietary that could cause a monopoly?
  6. Need for national energy conservation: Again, a national-level consideration. Will the standard provide improvements to the security and reliability of the nation’s energy system?
  7. Other factors that the Secretary of Energy considers relevant: General considerations include environmental and employment impact to the country. Will emergence of new technology foster job growth? What is the value of emissions reduction?

Economically, any proposed standard that can demonstrate payback in 3 years is fair game for federal adoption under the authority of the EPCA legislation.

While the development process for motor efficiency standards can seem glacial at times, still we are often caught off guard by the changes that show up in the energy codes. For example, while minimum motor efficiency requirements are just beginning to appear in IECC 2015, they are the result of legislation passed nearly 8 years ago. If we were to examine the progression from legislation to final adoption, it may seem needlessly convoluted. Nevertheless, remember that the effectiveness of the end product hinges on the creation of a universally accepted standard. Standards require robust, repeatable test procedures to be effective. You can’t quantify energy savings when you can’t measure something consistently.

For example, the motor efficiency tables in IECC 2015 are the indirect result of the EISA-2007 legislation. This legislation drove the expansion of the motor industry’s minimum efficiency levels that are reflected in the NEMA MG 1-2011 motor standard for manufacturers. This manufacturer’s standard parallels changes in 2012 to the federal mandated test standards defined in 10 CFR Part 431. These standards, which are just beginning to show up in IECC-2015, won’t be adopted by most states until 2016.

Figure 2: This shows 10 hp secondary hot water circulation pumps for an 8,000 MBH boiler plant installation. VFDs are mounted to stands just outside the picture’s field of view to the right of the pumps.VFD efficiency requirements

While motor efficiency requirements are fairly well documented within IECC-2015, VFD efficiency requirements still remain conspicuously absent. Use of VFDs, while inferred as a potential compliance path in the IECC, still are not specifically mandated as the sole means for code compliance. Note that the code describes only the basic functionality (mechanical or electrical variable speed drive) and not a particular technology defined by specific industry standard to achieve that functionality. The specific omission of a VFD requirement is somewhat odd when you consider that in the typical variable speed applications, VFDs dominate. Systems where VFDs are more or less the de facto means of code compliance include VAV fan systems in complex multi-zone systems, variable speed control for fan motors greater than 7.5 hp, and achieving stringent integrated part load value (IPLV) values for chillers. When trying to satisfy the code’s variable speed drive requirements, there are alternative to VFDs such as 2-speed motors, gearbox-motor combinations, and adjustable inlet guide vanes, but these are exceedingly rare in new construction due to their lower efficiency and greater complexity. The omission of a VFD-only requirement in the energy code could be partially attributed to the fact that until recently, there have not been industrywide accepted test standards for inverter duty motors and VFDs. However, in Europe the International Electrotechnical Commission (IEC) published testing standards for determining the efficiency of VFD controlled induction motors (IEC 60034-2-3) in 2013. Additional standards that will define efficiency classes of variable speed ac motors will be published in 2015; another standard that will define the energy efficiency of adjustable speed electric power drive systems will be published in 2016. The last standard is probably the most important of the three to building design because it addresses combined motor-VFD system efficiency rather than just the efficiency of individual components. With this standard, it eventually may be possible to objectively evaluate any given manufacturer’s motor in combination with another manufacturer’s VFD.

So, how could a new set of seemingly obscure European testing standards possibly impact an engineer half a world away in the United States? Again, the development of new testing standards telegraphs the potential for future code changes. Development of these European VFD efficiency testing standards doesn’t guarantee that parallel standards will be developed and adopted in the United States. However, with today’s manufacturers emphasizing standardized products for global markets, there are already strong efforts to harmonize regional test standards. For example, IEEE Standard 112: Standard Test Procedures for Polyphase Induction Motors and Generators, last updated in 2004 prior to the release of the 2006 version of NEMA MG 1, is currently in the workgroup session phase for new revisions. Among the revisions being considered is inclusion of certain IEC 60034 test procedures. It would seem that development of these standards is only a glimpse of things to come.

Value of energy efficiency

In most applications, operating cost dramatically outweighs the cost of the motor itself during its useful service life. Unfortunately, in typical building systems we typically size motors to accommodate the worst-case loads, which generally represent a very small percentage of total operating hours. If that motor were to run at full speed, regardless of the load, an incredible amount of power would be wasted. However, through the affinity laws, we know that power varies in relationship to speed by the following formula for centrifugal loads:

hp2 = hp1 (rpm2/rpm1)3

Table 1: In this particular example, the return on investment (ROI) would be less than a year. Courtesy: McGuire EngineersBased on this formula, if a load can be accommodated by a slower speed, you can dramatically reduce the power that the motor needs to produce.

To illustrate this relationship, let’s apply this concept to a simplified load profile using a 7.5 hp, 1800 rpm motor (see Table 1). If the average speed required by our theoretical load was of 75% of full load speed, the average power required would be 3.16 hp. Using equipment costs from an online electrical distributor specializing in induction motors and VFDs, and defining a few variables including electrical cost per kWh and hours of operation, we can get a rough comparison of the cost of the equipment compared to the energy cost and the return on investment.

These variables are generic, and your mileage may vary depending on the exact operational characteristics of your load and cost of electricity in your area. Because the affinity laws tell us that motor horsepower has a nonlinear relationship with speed, it is misleading to state that a calculation based on the average speed of your load profile will accurately reflect the overall energy usage (unless the load spends the vast majority of its time at that average speed). However, the point of this example is to illustrate that energy costs are usually the determining factor when variable speed motor operation is anticipated.

ac induction motor efficiency

Multi-phase ac induction motors have traditionally dominated the electric motor industry. The industry has improved overall motor efficiency in each subsequent NEMA-MG 1 motor standard revision. To make sense of motor efficiency, one must understand what efficiency losses are attributed to. Motor losses can be assigned to four major categories:

  1. Electrical losses: I2R conduction losses in the stator and rotor that increase dramatically with increased current. Can be improved by reducing the resistance of the stator windings and rotor squirrel cage.
  2. Magnetic losses: Hysteresis/eddy current in the steel laminations. Can be improved by improving the metallurgy of the steel.
  3. Mechanical losses: Friction in bearing system, parasitic loads like cooling fans in totally enclosed fan-cooled (TEFC) motors, etc.
  4. Stray load losses: Flux leakage/irregularities in the rotor/stator air gap. Can be improved by improved precision in manufacturing

Of these losses, I2R losses dramatically outweigh the other categories. I2R losses manifest themselves as increased heat, so by addressing these, other losses such as cooling fan motor losses in TEFC designs can be indirectly affected.

All induction motors have a base speed (synchronous speed) that is directly proportional to the quantity of magnetic poles in the motor’s design (generally between 2 and 8) and the frequency of the electrical source (60 Hz in North America):

ac motor synchronous speeds (60 Hz)

Poles    Speed (rpm)
2           3600
4           1800
6           1200
8             900

Table 2: High inertia loads with greater starting torque requirements generally necessitate NEMA motor designs with lower efficiency. Courtesy: McGuire EngineersHowever, the speed of the rotor always slightly lags the speed of the rotating magnetic field in the stator—it’s always trying to catch up. That difference between the synchronous speed and that actual speed of the rotor is known as “slip.” Reductions in the resistance of the stator and rotor reduce I2R losses but also result in less slip. Unfortunately, slip is directly related to the amount of torque that a motor of a given design can produce. As such, high inertia loads with greater starting torque requirements generally necessitate NEMA motor designs with lower efficiency. (See Table 2 for NEMA design designations.) Stator designs with lower I2R losses also typically have higher locked rotor current.

The overall efficiency improvements to this point in time have been limited to optimizations and not revolutionary changes in existing induction motor designs. The greatest recent jump in industry motor efficiency is attributed to EISA-2007 mandating NEMA Premium motor standards for all general-purpose, 1- to 200-hp, 3-phase motors rated to 600 V. While even greater efficiencies are possible, the improvements up to this point have been incremental and, if anything, will only start to level off without dramatic changes in the basic motor design. For example, the difference in efficiency between EPAct compliant and NEMA Table 3: Based on this data from the U.S. Dept. of Energy, MotorMaster software, and NEMA 60 Hz database (2010), other motor technologies should be entertained or other elements of the system, such as control by VFDs, should be more greatly emphasized. Courtesy: McGuire EngineersPremium motors is generally only 1% to 3% (see Table 3). These diminishing returns would suggest that if major increases in system efficiency are desired, other motor technologies should be entertained, or more importantly, other elements of the system such as control by VFDs should be more greatly emphasized.

How do you improve efficiency in VFDs?

That’s a trick question. But going back to an earlier discussion of standards in this article, you can’t consistently quantify efficiency if there’s no industrywide standard to measure it. Losses in VFDs are generally attributed to conduction (electrical current flowing through the device) and switching losses (the power lost by switching the transistors on/off during operation of the VFD’s input rectifier and output inverter sections). Theoretically, VFD manufacturers can address efficiency by optimizing these aspects of the VFD’s design. In general, most manufacturers will quote mid-90% efficiency for their newer VFD designs.

While VFD designs continue to evolve—the introduction of sixth-generation insulated-gate bipolar transistor (IGBT), which can offer roughly 20% relative reduction in overall switching losses compared to earlier IBGT designs—it’s difficult for specifying engineers to quantify exactly how these designs affect our overall system efficiency. Direct manufacturer-to-manufacturer comparisons are academic without a true testing standard.

Although a VFD manufacturer’s quoted efficiency can allow for a ballpark efficiency estimate of motor-VFD system, numerous variables can impact actual system efficiency. Examples include variables such as the carrier frequency at which a motor is operated or the NEMA design type of the motor being used. Ultimately, until the appropriate standards are developed, increasing system efficiency is best addressed by focusing on the fundamentals:

  • Understand the speed and torque characteristics of the load
  • Understand overall load profile/duty cycle
  • Specify right-sized equipment that can reliably meet those project parameters
  • Control it so it performs only the amount of work needed and no more.

The energy savings associated with properly addressing these load management concepts overshadow the few percentage points of efficiency that may exist between different manufacturers’ VFDs. However, it’s inevitable that VFD standards will eventually be developed similar to those currently in place for induction motors.

Figure 3: Harmonic current mitigation is typically accomplished with line reactors or 12-pulse rectifiers. This is the internal view of a 12-pulse rectifier with the 30-deg phase shift transformer typical of this type of design clearly visible in the lower left-hand corner. Note the cooling fans in the side of the enclosure with generous room to allow for proper heat dissipation.Reinventing the wheel with PMAC motors

While new efficiency standards have pushed the development of ever more efficient induction motors, there have been diminishing returns with the release of every new standard. While additional efficiency improvements are still possible, at some point it may not be economically justifiable per the EPCA 3-year payback criteria without some type of significant technological innovation. Induction motors may be approaching an evolutionary dead-end in this regard. However, it is possible that other similar technologies could supersede induction motors, similar to how VFDs acted as a “disruptive innovation” in the HVAC industry.

One of the more recent developments in motors are permanent magnetic alternating current (PMAC) motors. PMAC motors are also called synchronous ac or brushless dc motors. These motors, while not necessarily new, were not used in the HVAC industry until very recently. PMAC motors were traditionally used only where precise low speed and torque control were required, thus functionally overlapping with the induction motors using vector control VFDs. Their higher cost compared to a typical induction motor/VFD combination has limited the penetration of PMAC motors into the building systems market. However, the functional benefits of PMAC motors have led at least one cooling tower manufacturer to use them for large-diameter, draw-through cooling tower fans in its premium cooling tower model. The manufacturer exploited the low-speed performance characteristics of PMAC motors and eliminated problematic fan belt drive/reduction gearboxes typical in such designs. This not only simplifies the mechanical design of the tower, the elimination of the mechanical speed reduction mechanism also dramatically improves the overall efficiency of the design. This type of application, where the operational speed of the load is dramatically less than the standard base speed of an induction motor (1800 rpm, 3600 rpm, etc.) and belt drives or mechanical gearboxes end up being required, exemplifies where this type of motor shines.

Figure 4: This photo shows a “fanwall” air handler with 6 hp totally enclosed air over (TEAO) fan motors in a 6x12 array rated for 50,000 cfm. In this particular application, each fan was provided with a dedicated VFD. However, in this type of application it is common for multiple small fan motors to be connected to a single VFD.In smaller sizes (10 hp and smaller), the efficiency improvements inherent in the PMAC design offset its increased cost compared to induction motors. The overall efficiency improvement can be 2% higher than a NEMA Premium induction motor. However, PMAC motor efficiency really shines at slower speeds where the difference in overall efficiency can be greater than 20%. In combination with the 1000:1 or 2000:1 speed turndown ratios typical for PMAC motors used in vector control methodologies, it makes them an excellent choice for applications where regular low-speed operation is unavoidable.

The basic operational concepts are the same for induction motors and PMAC motors. A rotating magnetic field in the stator interacts with the rotor, and the speed of the rotor is dictated by the quantity of poles and frequency of the electrical source applied to the stator. In fact, the stator for a PMAC motor is virtually identical to that for an induction motor. The key difference lies in the rotor. Instead of the squirrel cage and the stacked steel laminations in an ac induction motor’s rotor, the PMAC rotor substitutes strong rare-earth magnets affixed to the steel laminations.

This substitution creates a permanent magnetic field of consistent strength in the rotor instead of the induced magnetic field characteristic of an induction motor. This is a doubled-edge sword in that the strength of the rotor’s magnetic field remains constant regardless of the speed of the motor. Going back to the earlier discussion of the types of losses in an induction motor, I2R losses are the greatest. While there are still I2R losses associated with the stator windings for a PMAC motor, the stator’s rotating magnetic field interacts directly with magnets in the rotor so there are negligible I2R (heat) losses associated with the induced current in the rotor. More or less eliminating the I2R rotor losses by use of permanent magnets dramatically reduces the overall losses. However, magnets can also make it very hard to start the motor and, once moving, act as a generator that causes “back electromotive force (EMF),” which we will discuss later.

Controlling a PMAC motor can be tricky. The magnets in the rotor will readily align in one position and resist it in others. So while the rotor is creating torque aligning with a particular stator winding, the other windings will not contribute the same magnitude of useful torque. In fact, depending on the location of the rotor rotation angle in relationship to a particular winding, the torque contributed by that winding can become negative and fight the forward rotation of the motor. The solution is to momentarily "turn off" that winding when the relative rotor position will create negative torque. Also, when the motor is started, as the rotating electromagnetic field in the stator builds, it may be insufficient to overcome the permanent magnets’ attraction to the stator’s steel structure without jerking. As such, a PMAC motor requires a VFD to control the input current to the motor stator windings. By extension, this means that a PMAC motor is not self-staring—it is impossible to start it with a standard across-the-line starter.

One important consideration with PMAC motors is that because of the presence of the permanent magnets in the rotor, the motor will act as an electrical generator. As the rotor turns faster, the voltage generated will increase and oppose the external voltage being applied to the stator. This concept is known as back EMF, and it can be both a blessing and a curse. Because the back EMF is proportional to speed, if you can somehow measure it, it can give you an indication of rotor speed without the need for a shaft encoder. If you know rotor speed, you should be able to precisely govern motor speed and torque (i.e., sensorless vector control). However, with faster speeds, the voltage that is generated will continue to increase and limit the maximum theoretical speed of the motor. Another caution is that as a motor spins down without input power, the voltage generated will energize the motor terminal and potentially cause a shock hazard.


John Yoon is the senior staff electrical engineer for McGuire Engineers. He has 20 years of experience in the design of electrical distribution systems. His project experience covers a broad spectrum, including high-rise building infrastructure renewal programs, tenant build-outs, mission critical data centers, GMP laboratories, and industrial facilities.

Want this article on your website? Click here to see how ContentStream® can make that happen.