Systems design and performance tips for packaged rooftop units (RTUs)

By Richard L. Kronick, Freelance Writer and Michael Ivanovich, Editor-in-Chief March 1, 2009

A currently fashionable adage is “Where you stand depends upon where you sit.”

That certainly applies to attitudes about direct-expansion unitary rooftop HVAC units (RTUs or “packaged air conditioning units”). Not surprisingly, manufacturers’ representatives speak of RTUs as the workhorses of the industry, as though they are proud Budweiser Clydesdale horses lined up on building roofs.

One senior engineer of a manufacturer extolled the long, consistent improvement in the efficiency ratings of RTUs—and tipped us off that the newest version of ASHRAE 90.1, due in 2010, will raise the bar again. He pointed to the current system of three tiers that allows engineers to balance a unit’s efficiency level with its price. He concluded by claiming that, while it is commonplace that chilled water systems are more economical than RTUs, when you add in the pumping power, the air handler, and the cooling power, chilled water systems are not as efficient as everyone thinks.

Other industry leaders would probably compare RTUs to a different species—say, the goat. David Sellers is a Portland, Ore.-based engineer for Facility Dynamics Engineering with decades of experience in HVAC design, commissioning, and building management. He agrees that you need to look at the whole-building system and not just the efficiency of the RTU, but maintains that the compactness that makes RTUs attractive often results in energy inefficiency.

This article will not pass judgment on whether RTUs should be seen as workhorses or goats. In view of the fact that RTUs account for at least 50% of HVAC tonnage in the United States (some estimates are as high as 60%) and are here to stay for the foreseeable future, the more pressing questions are: What should consulting engineers know about designing with RTUs for optimal lifecycle performance? And what can we expect from manufacturers in the near and not-so-near future?

Of course, “where you stand depends on where you sit” applies to factors such as climate, local contractor expertise for installation and service, what’s first-cost affordable, and a variety of other factors. Based on research and interviews with a wide range of industry leaders, here are some answers

Size matters

When asked for the No. 1 mistake made in specifying RTUs, several experts pointed to oversizing. For example, Glenn Hourahan, vice president of research and technology for the Air Conditioning Contractors of America (ACCA), said, “No one wants to be caught short, so everyone is biased toward bigger units.”

In his article, “Sizing and Selecting Equipment for Proper Humidity Control” in AHRI Magazine (Spring 2003; available at, Hourahan rejects rules of thumb, such as 500 to 600 sq ft/ton or 350 to 450 cfm/ton.

“Those ideas may have worked satisfactorily 30 years ago because buildings were not that tight and not that energy-efficient,” he said. Today, he said, engineers need to do load calculations for energy, IAQ, and humidity. “The programs available from ACCA and from all of the manufacturers deal with the basic issues. How well is the building constructed? What leakage occurs? What activities and processes occur in the building? If you’re building a convenience store, people will constantly be opening the door.

“For example, in a restaurant, cooking loads and ventilation loads need to be removed,” said Hourahan. “Museums and paper mills have stringent humidity requirements. Once you understand how the building is to be operated, you have to define the desired operating conditions: What temperature and humidity are you trying to maintain? What are the outside conditions? Are there special IAQ concerns that need attention? Generally, you design for the worst-case sensible and latent loads, but you also need to look for equipment that can handle the part-load conditions that will exist most of the time.”

A common result of oversizing an RTU, said Hourahan, is a humidity problem. A system may be able to cool a space so rapidly that the latent load is not being managed. In his article, Hourahan states, “The greater the difference between the indoor evaporator coil temperature and the return-air wet bulb temperature, the greater the ability the coil has to remove excess moisture.” Therefore, the solution to excess humidity is to reduce air flow across the coil. Enthalpy recovery units are another option to explore, but be sure to downsize the vapor-compression equipment in an RTU accordingly to account for the moisture removal.

It should be added that, quite often, a consulting engineer cannot know what activity will occur in a space. What begins as a strip mall bookstore with a relatively low load might later be pressed into service as a restaurant with a very high load. Given this all-too-common uncertainty, an engineer can be forgiven for erring on the conservative side and oversizing the equipment. Furthermore, today an engineer can address the sensible/latent heat issue by choosing from a wide range of options, such as variable-speed fans, modulating and staged compressors, energy recovery, demand ventilation, and reheat.

Pat Banse, PE, a senior mechanical engineer with design firm Smith Seckman Reid in Houston, agrees that oversizing is the No. 1 no-no, but Banse adds an additional warning: “One downside of oversizing is that rapid cycling on the refrigeration side can lead to premature failure—you burn out the compressor or a relay.” He said that, in larger systems, the best strategy is often to specify multiple compressors—and depend on more sophisticated controls to manage the situation.

How do you spell energy efficiency?

As shown in Table 1 at the bottom of the page, efficiency ratings for RTUs are expressed in EER, IEER, and IPLV. Minimum efficiency ratings for RTUs first appeared in ASHRAE standards in 1975. Since then the numbers have increased steadily—and also have been the subject of much debate. Table 1 lists the three methods for stating RTU energy efficiency that appear in the most recently published version of the Air-Conditioning, Heating, and Refrigeration Institute’s (AHRI) 340/360 Standard for Performance Rating of Commercial and Industrial Unitary Air-Conditioning and Heat Pump Equipment.

Beginning Jan. 1, 2010, IEER will replace IPLV in ASHRAE 90.1 as the efficiency metric for RTUs. It will take into account the various add-ons that allow shifting gears from part load to full load, such as variable-speed fans and staged compressors. The new standard will take into account that, on commercial units, the indoor fan is always on, which was not considered in the IPLV calculation. Also on Jan. 1, 2010, AHRI’s certification program for RTUs will be expanded to cover machines up to 760 Btu/hr.

How efficient does efficient have to be? Table 2, shown below, presents the federal government’s current minimum EERs.

Having stated all of the above, manufacturers offer many units that far exceed the government’s minimums, however, the highest efficiencies may have costs outside the boundaries of most budgets. There is ground between the minimum and maximum, such as the voluntary minimum performance levels specified by the Consortium for Energy Efficiency (CEE). CEE provides voluntary minimum efficiency levels for RTUs for utilities that manage rebate programs that help buy down the costs for higher-efficiency units.

Afroz Khan, CEE commercial HVAC program manager said, “That while we recognize that efficiency at the box only addresses one aspect of a broader issue, it’s still very much to your advantage to find out what incentives or programs your local utility is able to offer to encourage the installation of higher efficiency rated equipment.” A table of RTU efficiency specifications by size and type are in the “commercial” section of the CEE website at

Regarding refrigerants: Chill out!

Speaking of Jan. 1, 2010, that also is the date after which manufacturers may not produce new HVAC equipment containing HCFC-22 refrigerant (sometimes referred to as R-22). HCFC-22 is one of the hydrofluorocarbon (HCFC) refrigerants developed as a transitional replacement to chlorofluorocarbons (CFC), which damage the ozone layer of the earth’s atmosphere. The section of the U.S. Clean Air Act that addresses ozone depletion mandates that all HVAC equipment sold from Jan. 1, 2010, onward must contain an alternative to HCFC-22.

Though the law does not specify alternatives, the most widely recognized one-for-one replacement for HCFC-22 is 410A. While 410A has no effect on ozone, it is a “greenhouse gas” and therefore contributes to global warming if vented to the atmosphere. However, when the Montreal Protocol mandated the phaseout of HCFC-22 to address the ozone depletion potential, global warming potential was not on the table.

Understandably, this change has generated some anxiety in the industry, which must implement another refrigerant phaseout. However, the aggregated opinion of the experts interviewed for this article was “chill out.” Smith Seckman Reid’s Banse said, “I’ve had conversations with owners who were concerned [about switching to 410A]. But the equipment manufacturers are telling me they’re still going to support HCFC-22 for 10 years. I think you should have a plan for replacing your equipment, but you don’t necessarily have to do it today.”

Matt Muhlada, Trane’s North American product manager for large RTUs, agrees. “Some people are saying, ‘I’m just going to go ahead and buy a 410A system today.’ But they don’t need to do that. Now that refrigerant reclamation is mandated, HCFC-22 will be available for use in older systems,” he said. The Clean Air Act approves the use of recycled HCFC-22 in existing machines.

In addition to requiring an alternative to HCFC-22 in new equipment, the Clean Air Act mandates a gradual phaseout of HCFC-22. According to the Act, effective Jan. 1, 2015, no one may produce HCFC-22 (or any other “class II substance”) in an annual quantity greater than the amount that same person (which of course can mean a corporation) produced in “the baseline year.” Though the baseline year has not yet been announced, reliable sources suggest that it is likely to be retroactively set to 2007 or 2008 and that, over a several-year period, there will be a gradual reduction in the amount that can be produced compared to the baseline.

One new and “chilling” development is that on Dec. 11, 2008, the U.S. Environmental Protection Agency issued a proposed rule to ban the sale or distribution of air conditioning and refrigeration products containing HCFC-22, HCFC-142b, or blends containing one or both of these substances, starting Jan. 1, 2010. According to a transcript of a hearing held to obtain verbal comments from industry stakeholders, the proposed rule, once finalized, would “restrict the sale and distribution in interstate commerce, including the export and import of appliances that have been pre-charged with HCFC-22, HCFC-142B, or a mixture of those substances.”

The EPA currently is considering legislation to further regulate HCFC-22 regarding the shipping of pre-charged equipment. Stay tuned to the EPA for the announcement of further regulatory changes. No matter how the HCFC-22 phaseout is eventually structured, effective Jan. 1, 2030, no one may produce any HCFC-22 (or any other Class II substance).

The takeaway lesson here is: Keep one eye on the prices of 410A refrigerant and equipment and the other on the price of HCFC-22.

Noise and vibration

To reduce the transmission of noise from an RTU into a building, Trane’s Muhlada recommends an integral plenum curb/isolation rail. He said this blends the structural vibration reduction advantages of an isolation rail with the airborne acoustical advantages of a plenum and keeps duct breakout noise above the roofline. The cost of this option can be offset by the ability to use a common shaft for both supply and return ductwork.

However, Sellers points out that, on and near the West Coast, engineers face an additional issue—earthquakes—and a standard isolation rail will not suffice. He points to a “working paper” published by Mason Industries entitled “A Pictorial Study of Seismic Damage and the Use of Proper Safeguards.”

Muhlada said another way to avoid noise transmission is to use horizontal supply and return ductwork connections to the unit. Running ductwork above the roofline dissipates sound via breakout above the roofline. Muhlada describes a system of ducts with 90-deg turns to take advantage of end-reflection in the lower octave bands before entering the building. He said many consultants are using this strategy to successfully avoid sound transmission, though he warns that it will not eliminate structural vibration transmitted through the isolation rail. He also acknowledges that the raised equipment creates a line-of-sight issue and an access problem since you need to attach a ladder and a platform to the unit.

Meanwhile, Sellers warns against the potential inefficiency introduced by 90-deg duct angles. He provides two examples to underscore his point. Figure 1 below shows that what might seem to be small differences in duct geometry can make a big difference in pressure loss. Figure 2 below demonstrates that two closely spaced elbows can have significantly more pressure drop than the individual loss coefficients would predict.

The long and winding road

So what are we left with? Not surprisingly, the answer is a fair amount of disagreement among experts. Where some see the regulatory scene as a tangled web, others see it as an orderly process that has shown consistent progress driven by federal control.

Perhaps the balance point is struck by Sellers, who emphasizes personal responsibility: “We all need to do the best we can with the available and emerging technology,” he said. “There are a lot of resources out there that we can use to help educate designers, owners, and end users, which will help guide and even drive the industry toward more sustainable equipment and systems.”

Table 1: AHRI energy efficiency standards for RTUs

Methodology Symbol Description Notes
Energy efficiency ratio EER Ratio of Btu/hr to wattage of all power input At full-load only
Integrated energy efficiency ratio IEER An expression of efficiency at part-load See the standard for full details
Integrated part-load value IPLV Not an efficiency ratio; referred to as a “figure of merit” See the standard for full details
Source: Air-Conditioning, Heating, and Refrigeration Institute

Table 2: U.S. federal standards

Minimum energy efficiency for RTUs
Unit capacity (Btu/hr) Current minimum EER requirement Minimum EER as of Jan. 1, 2010
65,000-135,000 10.0 11.2
135,000-240,000 9.7 11.0
Source: U.S. Code Title 42, section 6313: Standards. Accessed online at on Feb. 26, 2009.

Figure 1: Pressure drop with two configurations of ductwork is shown. The method with the expansion after the elbow has a much lower pressure drop, which would be quieter and save more energy. Source: David Sellers, PE, Facility Dynamics Engineering, based on ASHRAE data.

Figure 2 : Pressure drop with three configurations of ductwork is shown. The method with 90-deg turns (close-coupled elblows) would have greater pressure drop, resulting in more noise and energy use. Source: David Sellers, based on ASHRAE data and data from AMCA-200-95 (R2007) Air Systems.

Author Information
Kronick is a Minneapolis-based freelance writer specializing in engineering and architecture. He is also a writing trainer who has presented more than 1,000 business writing and technical writing seminars on four continents. Ivanovich has been the editor-in-chief of Consulting-Specifying Engineer since 2007 and has a master’s degree in building systems engineering from the University of Colorado at Boulder.