HVAC: The Next Generation

Variable-air-volume systems came into favor primarily as a result of the energy concerns of the early 1970s. While they did wonders for that period, the 21st Century needs a new technology that overcomes the limitations of the old. That new technology may already be in place. As an HVAC strategy incorporating a dedicated outside-air source, radiant cooling and heat recovery promises to not on...

By Stanley A. Mumma, Ph.D., P.E., ASHRAE Fellow, Professor of Architectural Engineering, Penn State University, University Park, Pa. January 1, 2003

Variable-air-volume systems came into favor primarily as a result of the energy concerns of the early 1970s. While they did wonders for that period, the 21st Century needs a new technology that overcomes the limitations of the old. That new technology may already be in place. As an HVAC strategy incorporating a dedicated outside-air source, radiant cooling and heat recovery promises to not only solve the problems inherent in VAV, but may well revolutionize the industry.

What must be overcome

VAV clearly offers advantages over other all-air systems—terminal-reheat, dual-duct and multizone—including reduced fan energy consumption and greatly reduced bucking, which is simultaneous cooling and heating at part-load conditions. However, VAV systems have problems that the industry has been unable to solve over the past 30 years:

  • Poor air distribution.

  • Poor humidity control.

  • Poor acoustical qualities.

  • Poor use of plenum and mechanical shaft space. Large ducts and VAV boxes require deep plenums, increasing the first cost of the building envelope and everything else that moves vertically through the structure from columns and vertical circulation equipment to roof drains and ductwork.

  • Control problems. This is particularly the case with tracking return fan systems.

  • Poor energy transport. The density-specific heat characteristics of air make it one of the worst energy transport media imaginable.

  • Poor resistance to the threat of biological and chemical terrorism. VAV systems blend huge quantities of recirculated air—often over 80%—with outside air (OA) at the air-handling unit.

  • Poor and unpredictable ventilation performance.

Many have attempted to overcome these problems without sufficient success, but others in the industry are in denial about these problems. It is time to consider the next generation of HVAC.

Hope for a new generation

Before delving into the particulars of the new system, let’s first make a wishlist of what a next-generation HVAC system should do:

  • Solve the inherent problems of VAV with special emphasis on the ventilation performance.

  • Be available at equal or lower first cost.

  • Demonstrate 30% to 40% lower operating cost than VAV.

  • Exhibit superior humidity control, virtually eliminating microbial problems and associated sick-building syndrome.

  • Demonstrate superior thermal and acoustical comfort.

  • Be resistant to biological and chemical terrorism.

  • Be capable of generating a higher number of LEED Green Building Rating Points than VAV systems.

Okay, now let’s look at a next generation HVAC system on the books and in action at Penn State—the DOAS—and see how it meets our wish list criteria. The DOAS—dedicated outside air system—delivers 100% OA with ventilation air delivered to each conditioned space by a constant-volume OA unit utilizing total energy recovery (see Figure 1, p.40). By delivering ventilation air directly to each space, the OA flow rate is 20% to 70% less than that required by VAV systems, yet always meets the requirements of ASHRAE Standard 62.

Further, the OA unit removes all of the OA latent load and dries it sufficiently to remove the entire space latent load. This OA conditioning leads to a decoupling of the space sensible and latent loads, and to superior humidity control in the space—regardless of the sensible loads—thus essentially eliminating sick-building problems.

Finally, the OA unit is equipped with a total energy recovery device, specifically an enthalpy wheel, to cool and dehumidify OA in the summer. This reduces the OA load on the cooling coil by up to 80%. Furthermore, it significantly reduces the required size of the refrigeration plant for the building, often by more than 40%.

The enthalpy wheel also reduces the size of the winter heating and humidification plants and substantially reduces the energy consumed in many geographic locations.

By delegating all of the latent loads to the DOAS, or decoupling the space sensible and latent loads, the parallel system need only respond to the space sensible loads.

Of course, there is the question of parallel sensible cooling control. Fortunately, there are a host of options:

  • Fan coil units operating with dry coils, which eliminate septic sources throughout the facility.

  • VAV system operating with dry coils and only conditioning return air.

  • Unit ventilators operating with dry coils and with the OA path sealed.

  • Unitary air conditioning units operating with dry coils and with the OA path sealed.

  • Ceiling radiant cooling panels.

HVAC system designers in the United States are familiar with the first four options. Radiant cooling, however, is something that is not commonly employed here—not yet, anyway. However, in combination with DOAS, it will eventually replace VAV (see “Radiant Ceiling Panels,” p. 40).

Back to meeting the goals of our wish list… The biological and chemical resistance of a DOAS/radiant system cannot be overlooked. Since the system does not use any recirculated air, biological or chemical agents—released accidentally or intentionally inside the building—are not transported to other parts of the building. Rather, they are diluted and exhausted from each individual space.

Moreover, since the quantity of OA that must be treated is generally 20% less than that of conventional VAV systems, extensive biological and chemical agent filtration/treatment of the entering OA is far more practical with the DOAS/radiant system in terms of both first and operating costs.

DOE affirms DOAS

Besides these facts, consider the finding of the government. In July 2002, the U.S. Dept. of Energy issued a report entitled, “Energy Consumption Characteristics of Commercial Building HVAC Systems: Volume III, Energy Savings Potential.” In that report, 55 of the most promising technologies were ranked, with the top 15 receiving detailed analyses (see Table, p.41). The top energy-savings technology is radiant ceiling cooling, because water—an excellent heat transfer media—is used as the medium instead of air. The second energy saving technology is total energy recovery, and the third is DOAS. Imagine what the integration of these three technologies will produce. For one, their combined energy savings is 1.6 quads—almost as much as the other 12 technologies combined.

Another significant finding of the DOE report is presented in Figure 2, also on p. 41. On the basis of simple payback, both DOAS and radiant cooling have zero payback periods. This is certainly an attractive characteristic for the next-generation HVAC system.

Great, how do you do it?

About two years ago, planning began for the installation of the “proof of concept” project on the Penn State University campus in University Park, Pa. The building was selected, the design completed and financial and in-kind support raised. Installation began in the fall of 2001, with the mechanical system completed by January 2002. However, the web-based controls were not completed until early August 2002. Operating experience with the hot, humid summer conditions, as well as the cold winter conditions, are now in hand.

The project is located in a 40 ft. x 80 ft. architecture studio—the building was built in the early 1900s and the envelope had never been updated. It houses 40 students equipped with computers and task and overhead illumination.

The 1,200-cfm DOAS unit includes an enthalpy wheel. Air then enters the cooling and dehumidification coil, where it is cooled to a dew-point temperature (DPT) of 52is where condensation would be concentrated by gravity and trigger the condensate switch. The system also includes the chiller, primary chilled-water pumps, a radiant-panel pump and a pair of three-way control valves to regulate both the temperature of the air leaving the cooling coil and the temperature of the water entering the radiant panels. The panel control valve is spring-loaded and hard wired through the condensate sensor. So, if the direct digital control system fails to maintain the panel water temperature above the space, resulting in DPT and condensation, the condensate sensor switch opens and the panels are immediately isolated from the source of cooling, stopping condensation immediately before any damage can occur.

OK, how does it perform?

Realtime performance data shows that when rather hostile OA of 85penditure of chiller energy. At the cooling and dehumidification coil, another 4.5 tons of cooling was done, this time by the chiller. The OA leaving the high induction diffusers and entering the conditioned space had a DPT of 52.4ºF. The supply air was able to remove a combined sensible and latentload in the space of 3.3 tons as its enthalpy increased from 21.85 to 29.64 BTU/lbm. The radiant panels extracted the balance of the space sensible load, 4.1 tons, with water entering at 55ºF and leaving at 60ºF. At these design conditions, the chiller provided 8.6 tons of cooling—the chiller is rated at 10 tons, but was downsized due to the 4.7-ton capacity of the enthalpy wheel—while supplying 49.5ºF chiller water and receiving 60ºF chilled water in return.

Operating observations

For a further sense of how the system has operated since startup, consider the following five points.

1. Operation of the panel control valve at startup . For the first month of operation, the system was operated manually for a complete understanding of how it would behave. When the system was finally started in the automatic control mode, the space and the chilled-water temperatures were both 85ng below the space DPT. As the chilled-water temperature continued to drop and the DOAS supply air DPT dropped, the space DPT also dropped. As it did, the panel control valve PID loop allowed the panel water temperature to drop. This process continued until the space operative temperature set pointwas satisfied. Never was there any sign of condensation during automatic control operation, at start up or since.

2. Balancing the DOAS airflows . At startup, the constant volume supply airflow was set at 1,200 cfm, and the return airflow was set at 1,150 cfm. The 50 cfm difference was enough to pressurize the space and cause a slight exiting flow around the doors and windows. Because this is a constant-volume system, the balance is not disturbed by load changes, as is the case with VAV systems.

3. Condensation experiences . In order to test the condensate sensor, the DOAS cooling coil was shut down and the chilled water in the panel loop reduced to 45turn panel control valve closed, isolating the panel from the cooling source. It should be noted that the pump is programmed to continue operating for 15 minutes after this event, thus quickly warming up the water through the system and terminating further condensation. Further, it should be noted that no water dripped from the horizontal water piping or the radiant panels in this test.

4. Concern about moisture migration thru the very porous space enclosure . The space, with 1900s-era 12.34-ft. x 10-ft. single-glazed movable sashes, gave cause for moisture migration concerns. Remarkably, such migration has not been observed in the data, even when the outside DPT is 20

5. Heating/energy recovery . The proof of concept project was intended to primarily provide summer cooling and dehumidification information, so no provision for heating was provided. The existing space does have cast-iron wall-steam radiators, but their valves have been shut off, and the space operative temperature is maintained with an enthalpy wheel. Controls are in place to predict when the enthalpy wheel will frost, in which case the DOAS shuts down. Otherwise, the system is able to maintain 68

Time for a change

Clearly, such results indicate it is time for the U.S. HVAC industry to move forward and adopt this next-generation system. Not only does it provide superior humidity control and thermal comfort without the noise, draft or chemical or biological agent distribution problems inherent in VAV systems, it can be employed with equal or lower first cost, yet still deliver 30% to 40% lower energy operating costs. Case closed.

Table – U.S. Dept of Energy Top 15 Energy Saving Technologies

Technology Status Energy Savings in Quadrillion BTUs (quads)
Adaptive/fuzzy logic controls new 0.23
Dedicated outdoor air systems (DOAS) current 0.45
Displacement ventilation current 0.20
Electronically-commutated current 0.15
Permanent magnet motors
Enthalpy/energy recovery heat exchangers for ventilation current 0.55
Heat pumps for cold climates (zero-degree heat pump) advanced 0.10
Improved duct sealing current/new 0.23
Liquid desiccant air conditioners advanced 0.2/0.06
Microenvironments/occupancy-based control current 0.07
Microchannel heat exchanger new 0.11
Novel cool storage current 0.2/0.03
Radiant ceiling cooling/chilled beam current 0.6
Smaller centrifugal compressors advanced 0.15
System/component diagnostics new 0.45
Variable refrigerant volume/flow current 0.3

Radiant Ceiling Panels

There are a number of radiant ceiling systems available for heating and cooling applications. One type is comprised of panels that are an architectural finished product consisting of an aluminum absorbtion plate with copper tubes thermally and mechanically attached to the plate. The panels can be perforated to provide acoustical qualities. Panels are about the same weight as the acoustical tile they replace and are therefore responsive to load changes and have time constants of only minutes.

There are three drawbacks—condensation, lack of capacity and first cost. The former problems are addressable and cost can be offset with associated mechanical system reductions.

Although susceptible to condensation , the good news is that the phenomena is very slow forming. In fact, even a doubling of an occupancy will only form a film on the panels about the thickness of a human hair even after 14 hours. Further, a direct digital control system can be programmed to hold the panel fluid temperature above the space dew point temperature—about 55

Radiant panels lack capacity when held up to the 300-400 sq. ft. per ton of cooling rule of thumb typical of many facilities. In other words, about 40 BTU/hr-ft

1 cfm/ft

First cost is the final concern, but the DOAS/radiant approach results in reductions in the following systems compared to a typical VAV system:

Chiller plant size

Chilled water and condensing water pumps size


VAV box elimination

AHU size (An 80%)

Electrical service size

Plenum depth, reducing the envelope area and vertical elements in the building

Mechanical shafts, translating into less lost rentable space

As a result, the net effect is that the first cost of the mechanical system can be reduced by approximately $2 per sq. ft. of building floor area.