Motors, drives, and HVAC efficiency

Engineers must understand how the components in the HVAC systems they design use power and how they can be optimized without compromising traditional design values. Motors and drives are shown in relation to the design of HVAC systems.

By John Yoon, PE, LEED AP, McGuire Engineers Inc., Chicago January 26, 2016

Learning objectives:

  • Demonstrate how the energy codes directly affect motor and drive design.
  • Compare 6-, 12-, 18-pulse, and active front-end variable frequency drive (VFD) technologies.
  • Understand how VFD selection impacts power quality.
  • Apply products and systems within HVAC design to improve energy efficiency. 

Although variable frequency drives (VFDs) have been used in HVAC applications for a long time, the focus hasn’t extended much beyond basic functional and budgetary concerns. It is generally understood that using a VFD can allow for significant energy savings through the affinity laws. It is also understood that with certain types of fan systems like variable air volume (VAV) systems, VFD use is more or less required. However, ongoing changes in applicable codes and standards have effectively eliminated many other HVAC design options and made the use of VFDs a de facto requirement in many situations.

Code evolution

Design engineers are often caught off guard by energy-code changes. In many cases, the changes seem to add unjustified expense and complexity to HVAC systems that worked fine before. While engineers usually focus on specific aspects of designs, there is a need to take a step back to get a broader perspective on the external influences that will continue to push the designs toward increased energy efficiency.

A country’s energy usage and economy are linked. Although the relationship is changing in highly developed countries where there is a shift away from the industrial sector, a growing economy will generally result in increased electrical usage. Global events like the 1973-74 energy crisis and the two Gulf Wars can destabilize the cost of energy, with major repercussions on a country’s economy. The changes that we are seeing now in our energy codes are actually the result of ongoing federal legislation that can be traced back 40 years.

It should come as no surprise that major federal energy legislation (Energy Policy and Conservation Act of 1975, Energy Policy Act of 1992, Energy Independence and Security Act [EISA] of 2007) immediately followed major global events that impacted the cost of energy. That same energy legislation directly resulted in the energy codes and standards that affect our designs. Although projections from the U.S. Energy Information Administration would seem to indicate that the long-term trend is for economic growth to decouple from energy usage, the basic relationship should remain valid through 2040.

So, what exactly is the goal of energy codes and standards? The U.S. Department of Energy (DOE) is responsible for national energy policy and has federal statutory authority for the evaluation of energy codes and standards. They are given this authority through energy codes’ ability to affect public health and welfare. Energy codes and standards help ensure the public health and welfare by:

  • Reducing dependence on foreign energy by increasing efficiency and promoting alternate sources of energy
  • Protecting consumers through adoption of consistent standards
  • Providing for a more reliable electrical utility gird
  • Promoting economic development.

Efficiency in motors and VFDs is a prime target for any energy code or standard. The International Energy Agency (IEA) estimates that electric motor-driven systems account for 43% to 46% of global electricity consumption (7,108 terrawatt-hours [TWh]/year). Most of this is in the industrial sector. However, commercial building usage still accounts for 1,412 TWh/year.

These are incredibly big numbers and it can be difficult to truly grasp their magnitude—1 TWh is equal to 1 billion kWh. So, while an increase of a few percentage points in efficiency for an individual motor may not seem significant, it should be clear that small changes in industrywide motor-system efficiency standards can dramatically impact global energy consumption.

Delay in adoption of codes

While both are driven by federal energy legislation, there is a distinction between energy codes and industry standards. Standards typically apply to specific types of equipment like ac induction motors or centrifugal chillers. Energy codes apply to how that equipment is integrated into engineered systems.

While new energy-code requirements may seem overwhelming, what many specifying engineers don’t realize is that HVAC equipment manufacturers have been coping with these same problems for a longer period of time. The changes in the minimum federal efficiency standards that are referenced in the energy codes generally take place well in advance of their incorporation into those codes. For example, let’s look at the timeline for general-purpose squirrel-cage induction motors that are governed by the NEMA MG 1 industry standard:

  • EISA 2007 mandates NEMA Premium efficiency levels for motors. Before this, there were both minimum standard-efficiency and premium-efficiency offerings.
  • NEMA MG 1-2009 incorporates these as new minimum efficiency levels. MG 1-2009 became effective Dec. 19, 2010.
  • ASHRAE Standard 90.1-2013 and International Energy Conservation Code (IECC) 2015 incorporates these MG 1-2009 requirements.
  • DOE gets up to 12 months to determine if these energy codes would effectively improve energy efficiency.
  • Each state is required to certify within 2 years that their commercial energy code is in compliance with the new code.

All new motors currently in the marketplace should meet the minimum efficiency standards in MG 1-2009, but most states, as of December 2015, still haven’t adopted either ASHRAE 90.1-2013 or IECC 2015, which mandate their use in new construction.

Motors, VFDs, and the energy code

Motors are present in almost every piece of HVAC equipment that we specify and drive their overall energy usage. However, barring unforeseen technological innovations, continuing increases in NEMA MG 1 efficiency requirements are not sustainable. Motors are commodity products, and recent increases in efficiency can be attributed to adopting previously "premium" product requirements and making those the base-level standard. With each increase in motor efficiency, the associated manufacturing cost increases and, at some point, the improved efficiency no longer justifies the additional costs.

Given the diminishing returns associated with further increasing motor efficiency, the next obvious target is to focus how those motors are integrated and controlled within systems. With most HVAC systems, the peak-load/design-day condition represents a small percentage of the overall operating hours for that system. When considered in conjunction with the affinity laws, being able to operate motors at reduced speed/horsepower to meet actual load requirements clearly has the potential for dramatic energy savings.

The increasing importance of part-load efficiency is not only reflected in the energy codes, but also in evolving industry standards for several types of HVAC equipment. The Air-Conditioning, Heating, and Refrigeration Institute (AHRI) is the manufacturers’ trade association that sets standards for most HVAC, refrigeration, and water-heating equipment. AHRI has set efficiency standards for many types of air conditioning equipment including chillers, condensing units, air-cooled heat pumps, etc. These standards often mandate minimum efficiency requirements for not just full-load but also for part-load conditions.

These part-load efficiency requirements will take a weighted average of the equipment’s relative efficiency at multiple different load levels (25%, 50%, 75%, and 100% load). This weighted average is more representative of how the equipment will be used in a real-life application. Case in point: Chillers have different efficiency requirements for full load (FL) versus integrated part load values (IPLV). Large air conditioners have similar requirements where the full load energy-efficiency ratio (ERR) will be different than the integrated energy-efficiency ratio (IEER). Achieving these part-load efficiency requirements is generally accomplished by controlling the motors within that equipment with VFDs.

It is interesting to note that while there are minimum industry efficiency standards for motors per NEMA MG 1, there are no corresponding industry standards for VFDs yet.

ASHRAE 90.1 and IECC requirements

This shifting emphasis from equipment efficiency to system efficiency is reflected in each subsequent edition of the energy code. Again, efficiency requirements for specific pieces of equipment are somewhat transparent in that most HVAC equipment and components already meet the applicable NEMA and AHRI efficiency standards.

Let’s take a look at ASHRAE 90.1/IECC requirements as they directly relate to motors and VFDs. While there are some differences, both codes typically require some type of control to reduce flow under part-load conditions for both hydronic and air systems. Here are some examples:

  • Fan systems serving multiple zones shall be VAV systems (IECC 2012 and 2015).
  • VAV supply fans that control space temperature by airflow must have the capability to reduce fan motor demand to 30% power at 50% design air volume and one-third of total design static pressure. This is applicable to both Standard 90.1 and IECC, but each code has different thresholds for the minimum motor horsepower where this applies.
  • Multizone systems that adjust cooling capacity based on space, temperature, at least two stages of fan control (66% and 40% of full speed), are required for systems larger than 65,000 Btu/h, effective Jan. 1, 2016, under both ASRHAE 90.1-2013 and IECC 2015.
  • Pumps in hydronic systems with a combined motor capacity of 10 hp or larger need to automatically vary fluid flow by not less than 50%. Typical to both Standard 90.1 and IECC.
  • Under Standard 90.1 (both 2010 and 2013), the individual motor ≥5 hp in hydronic variable-flow systems with a total system power exceeding 10 hp must also meet the 30% power at 50% flow requirement.
  • Under Standard 90.1-2010, single-zone VAV with fans 5-hp and greater must be able to reduce their speed to the larger of 50% or the minimum speed necessary to meet ventilation requirements. Direct-expansion (DX) units with a capacity of 100,000 Btu/h must be able to reduce speed to two-thirds or the minimum speed necessary to meet ventilation requirements. Standard 90.1-2013 was revised and makes no distinction between multizone and single-zone VAV.
  • Heat-rejection equipment fan control—fan motors ≥7.5 hp shall have the capability to operate at two-thirds full speed with associated automatic controls (per both Standard 90.1-2010/2013 and IECC 2012/2015).

While this isn’t intended to be a comprehensive review of the energy-code requirements, it should be noted that there’s a reoccurring theme to have the ability to reduce flow to 50% and that the power required at half speed should be no more than 30% of that required at full speed. In fan systems, both codes allow for multiple means of compliance with these control requirements. These include mechanical or electrical variable speed drives (VSDs), vane-axial fans with variable-pitch blades, or "other controls and devices."

So, if we’re not specifically required to use VFDs, why are we forced to use them? If we look at means of adjusting flow in VAV systems, we can see that many of the traditional ways to accomplish this don’t meet the 30% power consumption at 50% flow requirement. The following is a summary from the ASHRAE 90.1 User’s Manual of various VAV-control methodologies:

  • Air foil fan with discharge dampers—does not meet efficiency requirement.
  • Forward curved fan with discharge dampers—does not meet efficiency requirement. 
  • Air foil fan with inlet guide vanes—does not meet efficiency requirement.
  • Forward curved fan with inlet guide vanes-does not meet efficiency requirement.
  • Vane-axial fan with variable-pitch blades—may meet the requirement in some circumstances.
  • Any fan with variable-speed control—generally meets the minimum efficiency requirement.

Actual fan performance will vary depending on multiple factors like static pressure sensor location, fan selection, etc. But, in general, variable-speed control with VFDs becomes the de facto means of compliance with this 50% flow/30% power requirement.

Changes to control requirements

VFDs are inherently capable of smoothly modulating motor speed. However, there are currently no specific mandates in the energy code for use of VFDs or continuously variable/modulating flow control to meet the power-reduction requirements. VFDs usually represent the best path to code compliance, but the current speed-control requirements in the code only list discrete percentages of full-load speed (40%, 50%, or 66%, depending on the system controlled). When you consider the power-speed relationship defined by the affinity laws, the gaps in these steps are dramatic. They leave an incredible amount of potential energy savings on the table. Ultimately, this gap can be attributed to two factors: inconsistent application/design of control systems and a lack of an industry standard for efficiency of VFDs.

There are currently no North American efficiency standards for VFDs similar to NEMA MG 1 for induction motors. It’s impossible to mandate efficiency for a particular product when there are no industry standards for how to consistently measure the efficiency of that product.

While direct digital control (DDC) systems generally are accepted in commercial HVAC system, the primary challenge aside from initial installation cost for DDC systems is having uniform design guidelines that can produce consistent, quantifiable improvements over existing baseline systems. Seemingly innocuous guidelines in the code, like defining the proper location of static pressure sensors in VAV systems, are laying the groundwork for this eventuality.

While there are multiple variables that can affect overall efficiency in complex HVAC systems, the basic affinity-law relationships remain valid. It is anticipated that fully modulating control will eventually be required as the cost of DDC and VFDs continue to fall and the industry agrees on VFD efficiency standards.

Harmonic distortion

With the changes in the energy code, it is inevitable that the use of VFDs will increase. As VFDs start to represent a disproportionately larger percentage of the electrical-load profile for a facility, the design process has to consider not just the need to ensure the reliability of the motor and VFD, but also what impact their use will have on the rest of the facility.

VFDs are nonlinear electrical loads that can create significant electrical harmonic distortion. To change the speed of a motor, VFDs change the frequency of the ac power wave-form that is supplied to that motor. This is done by converting 60-Hz ac utility power to dc power, which, in turn, is used to synthesize a new ac power sine wave. This wave-form can be changed in frequency based on the motor’s speed requirements. For this first ac-to-dc conversion step, VFDs and nonlinear loads generally draw power in short, high-amplitude bursts. For simple 6-pulse VFD designs, these bursts occur when the rectifier section of the VFD fires at the positive and negative peaks of the ac sine wave. They don’t draw power consistently through the entire sine wave—only at specific points, thus distorting/notching the wave-form at those peaks.

Nonlinear electrical loads cause "dirty" power. Harmonic content changes the sinusoidal characteristics of ac power and can result in a wave-form that looks more like a sawtooth profile. Excessive distortion/notching may result in additional zero crossings in the wave-form, thus affecting electronic equipment that depends on a consistent 60-Hz electrical-source frequency for proper operation.

We refer to certain "order harmonics" when we refer to the type of wave-form distortion—but what exactly are they? They are defined by an "integer multiple" of the base frequency of the source. So for example, the 3rd harmonic in a 60-Hz electrical distribution system would be 180 Hz, the 7th would be 420 Hz, etc.

Certain order harmonics are usually more problematic than others due to the fact that they can arithmetically reinforce/resonate with each other. The ones that are of greatest concern are commonly referred to as "triplens." Triplens are odd multiples of the 3rd harmonic (3, 9, 15, 21, etc.). Although predominant in single-phase nonlinear loads, these aren’t an issue with balanced 3-phase loads like motors. However, the 5th, 7th, 11th, and 13th are an issue with balanced 3-phase loads, with the 5th being the greatest concern. The 5th is a negative-sequence harmonic and, when supplied to a standard induction motor, can produce negative torque (i.e., slow it down from synchronous speed). Higher-order harmonics beyond these usually aren’t as much of a concern due to their general lower overall magnitude.

How much harmonic distortion is acceptable? IEEE Standard 519: Recommended Practice and Requirements for Harmonic Control in Electric Power Systems is the industry-accepted standard in this regard. This standard does not state how to minimize harmonic distortion—it is not a design manual. The latest version, the 2014 edition, has been significantly abbreviated and deletes most of the specific electrical-design references and appendixes that were present in the previous version (1992) to emphasize this point. Rather, the new focus of the standard is to only define the total harmonic distortion (THD) level that is allowable within an electrical distribution system. The means and methods to achieve these THD limits are at the designer’s discretion. It also notes that the threshold values given are only recommendations and should not be considered binding in all cases. In fact, the THD requirements have been relaxed in comparison to the previous version of the standard.

To quantify the harmonic distortion in a sensible manner, you need a consistent point in the electrical distribution system to measure/calculate it. This point in the electrical distribution system where the harmonic content is analyzed is called the point of common coupling (PCC). This is typically the low-voltage side of the utility transformer. Tables 1 and 2 summarize the most pertinent IEEE 519-2014 requirements for voltage and current distortion.

The key takeaway from these tables is that, although the allowable current distortion can increase in certain situations, voltage-distortion thresholds are fixed. This can be attributed to the fact that current distortion is pathway-dependent; it generally only goes between the source of electricity and the nonlinear load. Although the flow of excessive harmonic current may cause overheating of conductors and nuisance tripping of overcurrent protection devices, it generally won’t directly affect any other loads that are connected to the same electrical distribution. Voltage distortion, on the other hand, is not path-dependent and can affect anything else that is connected to that common electrical distribution bus. IEEE 519 also has guidance for acceptable current-distortion levels for specific individual harmonic orders. For the sake of clarity, that information isn’t listed here. However, in general, lower-order harmonics (i.e., 3rd through 11th) generally have higher acceptable limits.

It is recognized that lower-order harmonics are more difficult to address and also have a greater potential for power-quality issues. As such, IEEE 519-2014 has new provisions where, if a user can reduce lower-order harmonics, the threshold values for higher orders of harmonics for current distortion can be increased by a multiplying factor. Typically, if lower-order harmonics are limited to 25% of the values given in the standard, a multiplier can be applied to the remaining higher harmonic orders. The multiplier is directly proportional to the pulse order of the 3-phase rectifier being used. The equation is:

If a 6-pulse VFD was used, the multiplier would be 1.0x. With a 12-pulse VFD, the multiplier would be 1.41x, with an 18-pulse the multiplier would be 1.73x, etc. As we examine VFD topologies next, it will make more sense why the IEEE 519 standard rewards the use of 12- and 18-pulse VFD.

VFD design topologies

There are four primary VFD design types: 6-, 12-, and 18-pulse, and active front-end. Most mechanical engineers have a general understanding of what a VFD does, but if push comes to shove, a significant number can’t clearly describe the relative pros and cons for each of these VFD design topologies. Because the VFDs are often specified and purchased as part of a package with the associated MasterFormat Division 23 HVAC equipment, details on the specifications and performance requirements for those VFDS can often slip through the cracks.

The 6-pulse VFD design dominates the HVAC industry for very good reasons (see Figure 2). They are:

  • Cheap
  • Small
  • Efficient
  • Simple.

However, their simplicity does have a flipside. The rectifier section, the part that turns ac power to dc power, draws power only six times (two per phase, at the positive and negative peaks for each of the three phases) per each complete sine wave cycle. This results in unusually high levels of harmonic distortion. This harmonic distortion for a 6-pulse VFD can be up to 80%. Even after adding some type of filtering like line reactors, the harmonic distortion levels can still be in the 30% range. Due to the 6-pulse design of the rectifier section, the 5th-, 7th-, 11th-, 13th-, 17th-, and 19th-order harmonics are generated. Six-pulse VFDs seldom meet IEEE 519 requirements if the total VFD load represents a high percentage of the overall service size.

Line reactors can be added to help mitigate the harmonic distortion. These work by adding an impedance, which reduces the rate of change in the input, thus helping smooth the pulsing characteristics associated with harmonics. However, it also slightly reduces overall efficiency and causes voltage drop. So while high-impedance line reactors are theoretically more effective in mitigating harmonics, a 5%-impedance line reactor is typically the highest level used.

The 12-pulse VFD is essentially two 6-pulse VFDs grafted together (see Figure 3). This design also uses a 30-degree phase-shift transformer with two outputs. Each output feeds one of the 6-pulse rectifiers, which in turn are connected to a common dc bus. The VFD draws power 12 times (four times per phase for each of the three phases) per each complete sine wave cycle. While each 6-pulse rectifier still generates the same harmonics mentioned previously, what makes this design special is the transformer.

North American electrical power is delivered at a base frequency of 60 Hz. The transformer shifts the two 60-Hz outputs 30% out of phase. How does that same 30% phase shift for the 60-Hz base frequency translate to a 5th-order harmonic with a 300-Hz base frequency (60-Hz base frequency times five for the 5th harmonic)? The end result is that most of the major harmonic frequencies generated by each rectifier are shifted 180 degrees out of phase with each other. This effectively cancels out most harmonics and brings down the THD to around 12% to 18%. While the design is effective in mitigating 5th and 7th order harmonics, the 11th, 13th, 23rd, and 25th orders are still present to a lesser degree.

While a 12-pulse VFD goes a long way toward addressing harmonic distortion, there are significant downsides. Because of the additional components (two rectifiers and a transformer, etc.), the overall cost is significantly higher and the physical size is much larger. The cost of a 12-pulse VFD will typically be twice that of a 6-pulse VFD with line reactors. While it is common to have wall-mounted 6-pulse VFDs, most 12-pulse VFDs are freestanding enclosures. While not necessarily as great a consideration, 12-pulse VFDs are also slightly less efficient due to transformer losses and additional switching losses associated with the extra rectifier.

If 12-pulse designs are good, 18-pulse must be better—right? An 18-pulse VFD takes the concept one step further and uses a 3-output, 20-degree phase-shift transformer to feed three parallel 6-pulse rectifiers (see Figure 4). This design draws power 18 times (six times per phase for each of the three phases) per each complete sine wave cycle. This design is also extremely effective in addressing harmonic distortion.

All harmonics below the 17th order are dramatically reduced. Typically, THD will be in the 3% to 6% range. However, with this type of performance, there are significant increases in cost and complexity. They are typically only used for larger motor loads where the cost can be justified based on the potential impact that a large nonlinear load would otherwise have on a facility. The cost for an 18-pulse VFD will typically be 50% higher than a 12-pulse VFD.

Active front-end VFDs

VFDs are "double conversion" devices. They take ac power, convert that in the rectifier section to dc power, and, finally, the inverter section uses that dc power to resynthesize an ac wave-form (ac to dc to ac). This basic concept is closely paralleled in another type of equipment, uninterruptible power supplies (UPSs). Many of the harmonics and power-quality issues discussed in this article regarding VFDs are an even greater issue in the mission critical UPS marketplace. Standby generators used to provide backup power to those mission critical facilities were extremely vulnerable to harmonic distortion. High levels of harmonic distortion could wreak havoc on a generator’s ability to properly regulate its output voltage, therefore jeopardizing the critical load.

While the adoption of technologies like permanent magnet generators, excitation in generators has significantly reduced the impact of harmonic distortion, there remains a heightened sensitivity in the mission critical industry. As such, transformer-based 12-pulse UPS designs and input filters were fairly common until recently. However, given the nature of the industry, most UPS manufacturers were motivated to develop new, more cost-effective designs with similar or better performance. The solution was to swap new insulated-gate bipolar transistors (IGBT) for the silicon control rectifiers (SCR) traditionally used in the front-end rectifiers, and add active filtering to that. The active filter within the UPS eliminates harmonic distortion by inserting equal and opposite current into the line—somewhat similar to how noise-canceling headphones function. In comparison to the SCRs traditionally found in rectifiers, IGBTs have dramatically increased speed and controllability. In combination with active filtering, THD can be reduced to the 3% to 5% range while still increasing overall efficiency. This approach was so effective that it is now the industry standard in the mid- to large-capacity UPS market.

The most current evolution of VFDs reflects this design approach with IGBT-based rectifiers and active filtering. However, adoption of this new technology still lags far behind the UPS industry. While an active front-end VFD might be comparable in cost to an 18-pulse VFD, it is dramatically more expensive than a simple 6-pulse VFD. As such, current applications are typically limited to larger motors where the cost can be justified. Another limitation of the technology is that IGBTs aren’t as robust as traditional SCRs. While UPS manufacturers are starting to explore the substitution of IGBT in semiconductors with more robust silicon carbide (SiC) and gallium nitride (GaN), that technology is still in its infancy. It will be a long time before we will see that technology trickle down to the cost-sensitive VFD marketplace.

Fractional horsepower motors

Small fractional-horsepower motors (motors below 1 hp) are everywhere—fan coils, fan-powered VAV boxes, condensing unit fans, induced draft blowers, etc. The 2015 version of the International Energy Conservation Code (IECC) and ASHRAE Standard 90.1 are also beginning to address smaller fractional-horsepower motors (smaller than 1 hp) in fan motor applications. While there is a dramatically greater quantity of installed fractional-horsepower motorsworldwide, larger motors between 1 and 200 hp still use more power overall. However, with efficiency improvements for larger motors starting to level off, these smaller motors are now becoming the next target for new efficiency standards. IECC 2015 and ASHRAE 90.1-2013 now state that fan motors between 1/12 and 1 hp have to be electronically commutated motors (ECM) or have a minimum efficiency of 70% at full speed.

The fractional-horsepower market still is dominated by permanent-split-capacitor (PSC) type motors. In general, PSC motors don’t meet this minimum 70%-efficiency requirement. However, there is a loophole in that motors that comply with minimum federal energy efficiency requirements for PSC motors are acceptable per ASHRAE 90.1-2013 and IECC 2015. The barrier for federal adoption of this 70%-efficiency requirement could be attributed to the dramatic price difference between ECM and PSC motors.

However, ECMs are continuing to drop in price. Air-Conditioning, Heating, and Refrigeration Institute (AHRI) efficiency standards for typical small residential split-system air conditioners is a seasonal energy efficiency ratio (SEER) of 13. This efficiency level is incredibly difficult to achieve without the use of ECMs. In fact, early ECMs from one manufacturer were branded "X13" motors in reference to the ability to help air conditioner manufacturers meet the 13.0 SEER requirement. Increased acceptance in the residential air conditioner market may eventually translate to reduced product costs and facilitate further penetration into the commercial market. The fact that, in addition to being more efficient, ECMs also are inherently capable of variable-speed control suggests that they will at some point become the de facto means for compliance with the code.

John Yoon is the senior staff electrical engineer for McGuire Engineers. He has more than 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, laboratories, and industrial facilities. He is a member of the Consulting-Specifying Engineer editorial advisory board.