Printer Friendly

Control of underfloor air-distribution systems.

Underfloor air-distribution (UFAD) systems have been designed and built in the United States for more than 20 years with various degrees of success. The system remains controversial, with both advocates and detractors, but has experienced significant penetration in some markets. The most common complaint with these systems, however, is that spaces are chronically over-cooled. (1) Many critical factors have been identified for avoiding this pitfall, but the implementation of effective control strategies is arguably the most important step.

Underfloor air-distribution system typically refers to an HVAC system that delivers conditioning air from an air-handling unit through an access floor plenum to multiple floor-located diffusers or terminals that modulate airflow to individual zones to maintain comfort. Underfloor air is not a universal solution for all office buildings. It is well-suited to open plan, single tenant or owner-occupied buildings. In those buildings, the overall cost of the system, including available economies in systems furniture and cable distribution and certain tax advantages, is competitive with conventional overhead air-distribution systems. For occupancies that require many closed rooms, however, or where construction costs are divided between landlord and tenant, UFAD may be less attractive. Selection of the system should follow a comprehensive review of the usage, goals and configuration of an occupancy and extensive discussion with the occupants and owner of the project.

Differences between this system and a conventional single duct overhead delivery VAV system include:

* Air distribution is primarily through an open plenum under an access floor, rather than through closed ductwork above the ceiling.

* Air delivery from the floor-mounted diffusers is intended to be semi-displacement rather than full mixing and, therefore, the design supply air temperature to the space is much higher (~62[degrees]F vs. ~55[degrees]F [17[degrees]C vs. 13[degrees]C]) and diffuser face velocity is significantly lower than with overhead systems.

* Because floor registers are immediately accessible to occupants, manually adjustable diffusers are often used in interior workstations instead of thermostatically operated diffusers or terminals.

* Air temperature distribution in the space is usually markedly different with a UFAD system than with an overhead mixing system, showing significant stratification.

Many of the parts of an UFAD system are conventional and familiar, although some require some special modifications for UFAD. Primary air-handling units are similar to those of overhead systems, although, in humid climates, air-handling units supplying directly to the plenum will require a coil bypass so that return air may be redirected around the cooling coil to raise the supply air temperature to the space while maintaining the required apparatus dew-point temperature. All humid outside ventilation air is directed across the coil to ensure that the supply airstream has an adequately low dew-point temperature to control space humidity.

Many UFAD systems provide supply air for both interior and perimeter zones from the same source through the same supply plenum. Provision of a separate supply plenum or a separate, often hydronic, cooling source for perimeter zones usually is often ruled out because of operational or first cost considerations. Serving both the perimeter zones and interior zones from the same underfloor supply plenum requires a control sequence that enables comfort control for both types of zones simultaneously.

Supply air temperature degradation due to heat transfer across the access floor into the supply air and across the floor slab from the return air plenum below is a significant issue with UFAD systems. Many strategies have been developed to deal with this issue, but they are beyond the scope of this article. A well-designed underfloor plenum system using all of the known strategies to avoid thermal degradation should have a temperature rise across the plenum ranging from 2[degrees]F (1[degrees]C) to no more than 6[degrees]F (3.4[degrees]C). These strategies include location of supply air insertion points for the plenum to avoid lengthy or circuitous pathways to the most remote outlets and controlling insertion velocity to minimize the generation of large scale vortices under the floor.

The fundamental hypothesis of UFAD systems is that loads in the open plan area served by the system will vary uniformly over time. Control schemes can be applied to the entire distribution system to handle the load variation that does occur in this space. Individual manual control can be applied to the floor diffusers to "trim" air delivery to individual workstations or to handle an extraordinary load "event." Frequent manipulation of the floor diffusers is not considered to be a necessary component for maintaining comfort. A necessary corollary of this hypothesis is that comfort control for spaces not part of the general open plan can occur independently of the control stratagems imposed on the overall air-distribution system. These spaces might include closed offices, conference rooms and perimeter spaces. This corollary has significant implications for the design of the system to avoid conflict between comfort control in these separate spaces. Figure 1 shows typical UFAD system with both interior and perimeter zones served by the same floor supply plenum.

System Control for Maximized Comfort

Historically, UFAD systems have been designed and installed with various arrangements and control strategies with varying levels of success compared to conventional overhead systems. (2) Many times the project's physical form will guide the equipment locations and strategies, but in all cases, engineers should ensure that the systems are arranged to maximize occupant comfort and realize the other benefits possible with UFAD systems. Prior to committing to any control strategy, it is critical that the design team focus on creating system arrangements that minimize thermal decay and air leakage, promote air stratification and facilitate independent control of different space types served by the air-distribution system.

Block loads in the interior zones of office spaces are not constant. Even in an open plan area with uniform work station density, the block load may demonstrate a variation across the day. Following a warm night or weekend, the cooling load will experience a peak during "morning cool-down" as the system overcomes high temperatures that result from the overnight deactivation of the HVAC system. In cooler weather, early morning cooling loads may be almost non-existent as the heat gain from lights, equipment and people must warm up the thermal mass of the space before the heat gain shows up as cooling load. A fundamental requirement for maintaining comfort is accommodation of these basic load profiles, while maintaining flexibility to meet loads in other spaces.

Generally recognized schemes for tracking the block load profile of the open plan interior zone are to reset the positive pressure setpoint of the supply plenum with respect to the space and reset of the supply air temperature in the floor plenum. (3) Each of these alternatives has implications for comfort control in the non-open plan spaces. Reset of the plenum pressure setpoint requires that airflow to the zones that are not interior open-plan be independent of plenum pressure or that the air outlets in those spaces are sized for design airflow at a pressure lower than the maximum setpoint. Reset of supply air temperature implies that the air outlets be sized to meet design cooling loads with a higher supply air temperature than the minimum setpoint. Reset of supply air temperature also implies that the dew-point of the supply air is relatively independent of the supply air dry bulb temperature in order to maintain space humidity control in humid climates. Supply temperature downward reset should also be limited to a minimum of 60[degrees]F (15.6[degrees]C) to avoid thermal asymmetry discomfort (cold feet, warm head) for space occupants.

Figure 2 is a control scheme that has been found successful for several different projects:

* Reset supply plenum static pressure setpoint based on interior space temperature. The logic resets the pressure setpoint to maximum design pressure (e.g., 0.1 in. w.g. [25 Pa]) when the interior spaces are warm down to 0.01 in. w.g. [2.5 Pa] when they are cold.

* Reset supply air temperature to satisfy the perimeter zone that requires the coldest air. The best reset strategy is trim and respond, which easily allows the user to ignore some non-critical zones from the logic. (4)

The two strategies together can help prevent overcooling: as supply air temperature falls when perimeter zone cooling demand increases, the floor pressure falls to reduce airflow to interior spaces to reducing overcooling.

These reset protocols require several temperature sensors mounted in the open office area. (3) The author's experience is that mounting these sensors approximately 6 ft (1.8 m) above the finished floor is an effective strategy. Several sensors, spaced around the open plan area, are used, and they can be averaged to determine whether or not reset is necessary. The setpoint temperature for these sensors, approximately at head height, should be a few degrees warmer than the ideal comfort temperature for seated chest height. These sensors should be identified on the documents as temperature sensors, as opposed to control thermostats, so as not to evoke Americans with Disabilities Act (ADA) requirements for location.

Using differential pressure reset as a control stratagem is dependent upon two factors. The first of these is that the pressure sensors used have the sensitivity and accuracy to measure very low pressures. Inadequate sensors will not be able to deliver sufficiently fine control to modulate capacity in response to load variation. The sensor range should be as low as possible to capture the maximum design pressure. Sensors with accuracy as low as [+ or -]0.5% of full scale are readily available at reasonable cost.

The second requirement is somewhat more subtle and it is that the pressure drop across the floor from the plenum to the space be sufficiently high to allow an adequate control range for re-setting the plenum pressure differential. Airflow through the floor from the plenum to the space is composed of both leakage through the floor and flow through the various diffusers and terminal units that control airflow to the space. Excessive leakage through the floor or deployment of too many floor diffusers can result in lower than anticipated pressure drop through the floor at design airflow. If design airflow is achieved at a much lower pressure differential across the floor than 0.05 in. w.g. (12.5 Pa), then the control range for floor pressure reset may be too small to achieve the required flow modulation to accommodate a varying load profile for the interior zones.

In general, leakage from the supply plenum is classified as Type I, Leakage to Unoccupied Spaces (including outdoors, core and return air plenum), and Type II, Leakage to Occupied Spaces. While Type I leakage may represent energy waste, either fan energy for moving air directly from the supply plenum to the return plenum, or both fan and cooling energy by moving air out of the conditioned area, Type II leakage presents a more subtle controllability problem that may lead to overall occupant dissatisfaction with the building.

Avoidance of this problem requires several different steps. The first is a robust performance specification for air leakage through the floor, accompanied by requirements for verification that the specified measures have been implemented. Recent testing data has indicated that leakage levels, at a pressure differential of 0.05 to 0.06 in. w.g., (12.5 Pa to 15 Pa) of less than 5% for Type I, and less than 7.5% for Type II, may be achieved. (5) Performance specification and testing requirements will enable the owner to require remediation should the floor system fail to comply. The second step is an accurate load calculation to determine the maximum amount of supply airflow that will be required to condition the area served by the underfloor plenum. The third step is to allocate the number of passive floor diffusers such that design flow will only be achieved when plenum pressure is at or above the target pressure differential. Sizing of air terminals and determination of the number and location of passive diffusers should recognize that Type II leakage will contribute a significant amount of uncontrolled conditioning air to the space. The author has often limited passive diffusers to workspace locations, completely eliminating them from transient areas such as passageways and congregation areas, in order to maintain an adequate pressure drop from the plenum to the space. If building commissioning reveals that airflow is achieved at a lower pressure differential than desired, then some of the floor diffusers may be closed off.

Use of plenum pressure reset as a means of tracking the block load of the interior space means that other types of zones must be able to track their individual loads independently of plenum pressurization. For enclosed private offices or small conference rooms this may mean the use of thermostatically controlled floor diffusers that are sized to deliver design airflow at lower than design pressure. Thermostatic controls can restrict flow through the diffuser during periods of lower part loads in the space or of higher pressurization of the supply plenum. Areas with more intense cooling loads such as large interior conference rooms and perimeter zones require thermostatically controlled fan forced air supply to those zones. Variable speed fan terminals convey air from the plenum to the space independently of plenum pressurization, fully isolating perimeter zone temperature control from load tracking in the interior zone. Ideally, the heating mechanism for the perimeter zones is completely separated from the underfloor air system, for example, under-window convectors, but rarely is this solution architecturally acceptable. As a result, the fan terminals usually incorporate hydronic coils or electric resistance coils to provide heat to the perimeter zones. The fan terminal control scheme in Figure 2 recommended to be compliant with ASHRAE/IES Standard 90.1 restrictions on reheat of previously cooled air. (6)

Large conference room variable speed fan terminals follow a similar control scheme except that the fan does not shut off in the deadband in order to fully comply with ASHRAE Standard 62.1. C[O.sub.2] sensors can be used to dynamically reset the minimum airflow setpoint. Because C[O.sub.2] emission from occupants can cause C[O.sub.2] to rise faster than occupant heat gain causes space temperature to rise, reheat coils may be required to maintain the room within the required temperature range.

If reduction of supply plenum differential pressure proves inadequate to avoid overcooling the interior zones of the space, the second stage of capacity control for the interior zones is raising the supply air temperature setpoint. Unfortunately, upward reset of supply air necessarily impacts system cooling capacity for the perimeter zones. This strategy should be avoided except during periods when perimeter or conference room loads are very unlikely to be at design levels, such as during nighttime partial occupancy. In most cases, occupied periods with the lowest internal zone cooling loads, possibly required supply air temperature reset, are the same periods that will likely have lower conference room and perimeter zone loads.


Many projects have demonstrated that UFAD systems are an appropriate and successful HVAC system selection for some office building applications. Successful design of UFAD systems requires reconciling passive comfort control in the interior open-plan zones with active comfort control in perimeter and enclosed zones. The most common comfort complaint in UFAD systems is overcooling in the open plan interior areas. Successful temperature control in these areas requires control schemes that allow the system to track interior zone load profiles without inordinately curtailing system capacity at the perimeter zones. Achievement of this goal can be accomplished through the following control measures:

* Use plenum pressure control as the primary means of tracking interior zone cooling loads.

* Use sensors that are sufficiently sensitive and accurate, precisely to control plenum pressurization.

* Ensure that supply air temperature reset does not compromise required cooling capacity at exterior zones, private offices or conference rooms.

* Use capacity modulation methods in perimeter and enclosed spaces that are relatively independent of supply plenum pressure.

These goals can be achieved with either a common or a separate cooling source for perimeter and exterior zones. Success will be determined by rigorous recognition of how the control sequences interact to maintain comfort in both types of zones.


(1.) Lee, E.S., et al. 2013. 'A Post-Occupancy Monitored Evaluation of the Dimmable Lighting, Automated Shading, and Underfloor Air Distribution System in The New York Times Building." Lawrence Berkeley National Laboratory, pp. 49-50.

(2.) Woods, J. 2004. "What real-world experience says about the UFAD alternative." ASHRAEJournal 46(2).

(3.) Megerson, J.E., et al. 2013. UFAD Guide:Design, Construction and Operation of Underfloor Air Distribution Systems. Atlanta: ASHRAE.

(4.) Hydeman, M., et al. 2014. "Final Report: ASHRAE RP-1455 Advanced Control Sequences for HVAC Systems, Phase I."

(5.) Anticknap, S., M. Opalka 2011. "Testing for leaks in underfloor plenums." ASHRAEJournal 53(12).

(6.) ASHRAE/IES Standard 90.1-2013, Energy Standard for Buildings Except Low-Rise Residential Buildings, p 52.

(7.) Lee, K. H., et. 2011. "Lessons Learned In Modeling Underfloor Air Distribution Systems." Center for the Built Environment.


Daniel H. Nall, P.E., FAIA, is vice president at Syska Hennessy Group, New York.

Caption: FIGURE 1 Configuration of the underfloor air-distribution system.

Caption: FIGURE 2 Fan operation and airflow for perimeter fan terminals. (7)
COPYRIGHT 2015 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Nall, Daniel H.
Publication:ASHRAE Journal
Geographic Code:1USA
Date:May 1, 2015
Previous Article:A beacon: for urban waters.
Next Article:Vitruvius does veneers.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters