Reference Publication: Cummings, J.B. 1998. "Standards and Verification Protocols for Commissioning Air Flows in Buildings". Proceedings of the Sixth National Conference on Building Commissioning. Lake Buena Vista, FL. May 1998.
Disclaimer: The views and opinions expressed in this article are solely those of the authors and are not intended to represent the views and opinions of the Florida Solar Energy Center.
and Verification Protocols for
Commissioning Air Flows in Buildings
Florida Solar Energy Center (FSEC)
Commissioning involves three major elements – 1) process, 2) standards, and 3) protocols to verify compliance with the standards. Process is the step-by-step oversight of and involvement in the design, build, and start-up of the facility, to make sure that the design matches the customer’s needs, that the design is consistent with good building science practice, that construction matches the construction documents, and that systems and system interactions are properly tested at start-up. This paper does not address process, rather it focuses on the second and third elements -- standards and compliance protocols. It presents standards related to airtightness, air flows, air distribution system, and pressure differentials against which acceptable building operation can be assessed. It also presents diagnostic methods for assessing building air flows, pressure differentials, the location of the primary air and thermal boundaries of the building, return air restrictions, air distribution system airtightness, and building envelope airtightness, and verifying that these elements meet the standards for a good building.
James B. Cummings is a Principal Research Analyst at the Florida Solar Energy Center where he has worked since 1985. He has a Master of Science degree in Applied Solar Energy from Trinity University. Mr. Cummings has done extensive field research in the areas of building infiltration and ventilation, residential duct leakage, uncontrolled air flows in buildings, moisture storage and transport, and indoor air quality assessment and diagnostics, and has done extensive training in these areas of research.
1. What Do We Mean by Commissioning Air Flows?
For most readers, commissioning of air flows in buildings means Test and Balance -- that is testing and adjusting air flows at designated grills, registers, and hoods to match the flows which have been specified in the design documents. What is being addressed in this paper is much broader. It includes evaluation of uncontrolled air flows such as duct leakage and restricted return air, building envelope airtightness, and building pressure differentials (pressure mapping).
This paper looks at the total building air flow balance, as shown by the design documents and as they actually occur as operated. Reviewing the design documents is straightforward. Design air flows into the building (through the mechanical systems) should be greater than the air flows out of the building (in hot and humid climates), so that the building will operate at positive pressure most of the time. Individual zones should also be at positive pressure with respect to (wrt) outdoors.
Determining the air flow balance as operated is a much more difficult job. Air flows can be measured at the grills, registers, and hoods, and the intake and exhaust air flows can be compared. However, this tells us the building and zone air flow balance only if there are no air leaks in the ducts that convey the air between various grills, registers, and hoods. In actual fact, ducts virtually always leak, and most of the time the leaks are sufficiently large to yield inaccurate air flow balance conclusions.
In addition to duct leakage, building air flows often go out of balance when return air is restricted. Restricted return air occurs under two significant scenarios – 1) when returns are located in the central zone and closed interior doors block return air to the central zone, and 2) when the ceiling space is used as a return plenum and fire walls subdivide the plenum and restrict return air. Restricted return air creates positive pressure in closed rooms and negative pressure in the central zone where the returns are located. Fire wall restrictions, when there are either no return transfer “windows” or undersized “windows”, cause positive pressure in the zone further from the air handler and negative pressure in the zone closer to the air handler. These pressure imbalances drive air out of the building from the positive pressure zones and draw air into the depressurized portions of the building and therefore create ventilation rates that may exceed those indicated by the design.
Second, this paper addresses the airtightness of the building envelope. It is important that the building envelope be reasonably airtight, for two reasons. First, the building envelope should be sufficiently tight so that the conditioned air does not readily escape to outdoors. Secondly, the building envelope should be sufficiently tight to allow positive pressure to be established at a level sufficient to maintain air flow outward through the building shell under most weather conditions.
It is also important that the building not be too tight. If the building is too tight, air flow imbalances that result from duct leakage, restricted return air, and exhaust air/intake air imbalance may produce extreme pressure differentials (sometimes +20 to 50 Pa). These pressures can lead to moisture damage to walls, backdrafting or flame roll-out from combustion appliances, and suction of sewer gases into buildings. In one fairly tight restaurant, for example, large exhaust fans and undersized make-up air caused the entire building to go to -63 Pa, and a sign was posted on the exterior doors saying “Please Pull Hard”. In another tight restaurant, outdoor air (OA) filters and make-up air filters had not been cleaned for 8 months, so they were caked with dust and dirt and were restricting air flow into the building, causing -28 Pa. As a consequence, the instantaneous gas water heater pilot light was repeatedly extinguished and sewer gases were being drawn into the building where a toilet was not properly seated. When the filters were cleaned, building pressure went from -28 to -2 Pa.
Third, this paper also addresses building pressure differentials (dPs). dPs are important because they move air across the building envelope and change the building ventilation rate. They also determine the direction of air flow through exterior walls and may affect combustion appliance operation, entry of moisture into walls, and entry of sewer or soil gases into the building.
1.1 Recent Research Has Found Widespread UAF Problems
Recent commercial building research has found that uncontrolled air flow (UAF) problems are widespread and cause a range of problems, many of them severe (Cummings et al., 1996). What is UAF? It is defined as air flow across the building envelope or between building zones, where the pathways of flow, direction of flow, and origin of the air is unknown, unspecified, or unintended. UAF falls into four categories; 1) duct leakage, 2) restricted return air, 3) unbalanced exhaust air, and 4) excessive ventilation caused by oversized OA or exhaust air, or excessive building envelope leakage. Problem: if air flows are not controlled, we cannot assure control over the quality of the indoor environment in terms of temperature, relative humidity (RH), and air quality.
UAF can cause a variety of problems such as energy waste, occupant discomfort, high or low RH, moisture damage to walls, soil or sewer gas entry, combustion equipment backdrafting, and poor indoor air quality (IAQ) (Cummings et al., 1996). These are caused primarily by three mechanisms; 1) air flows from zones with undesirable temperature, RH, or air quality, 2) excessive air flows from outdoors, and 3) zone depressurization. Consider examples of consequences from these three mechanisms. 1) If air moves into the building from hot and/or humid spaces (attic or unconditioned ceiling spaces), then cooling energy use will increase and the cooling load may exceed capacity causing comfort problems. Or if air is drawn from contaminated zones (e.g., crawl spaces, loading docks), then IAQ will deteriorate. 2) If UAF causes excessive ventilation, then RH will increase (hot/humid weather) and mold/mildew contamination problems may result. 3) If the building or building zones are depressurized, then a range of problems may result; wall moisture damage, mold growth behind vinyl wallpaper, radon intrusion, and backdrafting of combustion appliances.
1.2 Catastrophic Moisture Failures
Many failures occur in buildings but most do not create major problems.
The thrust of this paper is that UAF deserves focused attention in
the commissioning community because it has the potential to cause major
and even catastrophic building failures. In hot and humid climates,
one of the common and often most catastrophic results of UAF is moisture
accumulation in walls (and other building cavities, as well). This
often happens in hot and humid climates when the building or zones
of the building are depressurized by duct leaks, restricted return
air, or exhaust air imbalance. Space depressurization draws humid air
into the wall from outdoors where it comes into contact with cool wall
materials where it is cooled and goes to a high RH, which drives moisture
into the gypsum board by adsorption and possibly condensation.
If high moisture content occurs in the gypsum board, then mold and mildew growth is likely, and with further wetting the wall materials may become soft and lose their integrity. Wall moisture accumulation depends not only on the rate of adsorption in the wall interior but also on the rate of drying through the wall board to indoors. It is possible that no problem will exist even when a high moisture adsorption rate exists inside the wall because of a high drying rate to the room. However, if the wall has vapor resistant paint or vinyl wallpaper, then destructive moisture accumulation can occur even within a matter of months with catastrophic results. Instances of buildings having 30% or more of their gypsum board replaced even before the buildings open have been reported. (Note that the colder the indoor temperature, the greater the potential for moisture accumulation.) A critical factor is persistence of depressurization. Pressures of -1 or -2 Pa can create moisture problems when that pressure and the resulting air flow direction are persistent over months of hot and humid weather (outdoor dewpoint of 70oF or higher). On the other hand, if the pressure reverses itself on a daily or even weekly basis, then dry indoor air will pass through the wall cavities and may dry materials out before significant damage occurs.
Another cause of catastrophic failure is persistent high indoor RH leading to mold and mildew growth and contamination. High RH can result from a variety of factors, including oversized cooling equipment, continuous air handler fan operation (evaporation from the coil during compressor-off cycles), short-cycling of the compressor, three-way by-pass valves on chilled water systems (causing a warm coil), and large sources of indoor moisture generation. However, UAF can also be a significant and sometimes overwhelming factor. It can cause excessive ventilation and latent cooling loads, which can cause persistently high indoor RH which can lead to serious IAQ contamination problems, health problems, sick buildings, evacuated buildings, and lawsuits. Consequently, it is important that standards exist for managing air flows in buildings and that protocols be available to verify compliance with those standards.
2. Standards for Managing Air Flows
To successfully manage air flows in buildings requires standards for building airtightness, building and zone pressures, air flows, and air distribution.
2.1 Standards for Building Airtightness
Typically, there have not been standards for building airtightness. Rather building airtightness was a by-product of the particular construction details used. The most important variable in commercial building airtightness is the presence of suspended, t-bar ceilings which are very leaky (about 10 to 20 times more leaky than typical gypsum board ceilings). Because the ceilings are so leaky, building airtightness depends, in large part, upon the airtightness of the space above the ceiling wrt outdoors. If it is well ventilated, then the building will be very leaky. If tight, the building may be tight.
How tight should commercial buildings be? The building envelope needs to be sufficiently airtight to control natural infiltration and maintain positive building pressure, but not so tight as to create excessive dPs in response to unbalanced air flows.
2.2 Standards for Air Pressure
2.3 Standards for Air Flows
Air flows should be as small as possible while still fulfilling the purpose of each specific HVAC system. This is especially important for exhaust fans. In buildings such as restaurants, schools, and sports facilities, exhaust air is often “the tail that wags the dog”. Very large exhaust flow rates create: 1) the need for large make-up air flows (both conditioned and unconditioned), 2) the potential for serious space depressurization (all 9 restaurants in UAF study were moderately or seriously depressurized; Cummings et al.), 3) the potential for humidity control problems, 4) the need for larger heating and cooling systems, and 5) increased heating and cooling energy use. Therefore, exhaust systems should be designed with the smallest air flows consistent with effective capture of heat, odors, and vapors. An important part of minimizing exhaust air flow is optimization of hood design in kitchens, and proper location of the cooking appliances under the hood.
OA should be sized to be consistent with the ventilation and make-up air requirements of the building. Where possible, try to meet make-up air requirements with unconditioned air. Excessive ventilation beyond what is required wastes energy and raises RH during hot/humid weather, and wastes energy and creates low RH during cold weather.
Not only is it important to properly size exhaust air, make-up air, and OA, it is important that actual “as operated” flow rates match design. Deviation from these air flow rates often occurs because of duct leaks, failure to maintain filters (restrict air flow), or tampering with equipment or damper settings. In one school building, for example, facility staff changed kitchen exhaust hood filters from screen type to grease extraction type (though no grease cooking was being done), and the exhaust flow rate decreased from 8100 cfm to 5000 cfm in a kitchen .
Exhaust fans (and make-up air) in kitchens, locker rooms, etc. should be on a schedule so that they do not run when not needed. OA, also, should be brought into the building only as needed. This can be achieved by means of occupancy sensors (motioned detectors with delay on make relays, or carbon dioxide controllers) or time-clock schedule combined with dampers on the OA ducts. In hot and humid climates, this sort of control over OA not only saves energy but also reduces the latent cooling load and indoor RH.
In one tested building, OA (about 750 cfm on a five ton system) was being drawn from a ceiling space, and this ceiling space was inside the building air boundary (only slightly vented to outdoors) but was outside the building thermal boundary (insulation on the ceiling tiles). Essentially, OA was being drawn from a hot and poorly ventilated attic. Therefore, this “outdoor air” was transporting heat but no fresh air into the building. In similar manner, OA ducts that run through attic or crawl spaces may have significant leaks drawing undesirable air into the building while providing minimal ventilation. In other cases, outdoor intakes are located near pollution sources, such as exhaust discharge, parking garages, or loading docks, and therefore denigrate IAQ.
2.4 Standards for Air Distribution
Duct system airtightness is often expressed in terms of CFM25, or the air flow rate through the leak sites when the ducts are at -25 Pa. SMACNA (Sheet Metal and Air Conditioning Contractors National Association) standards for Class 6 ducts (sealed ducts) is 1.3 CFM25 per 100 square feet of duct surface area (SMACNA, 1985). In a typical commercial building, duct surface area might be about 35% of the floor area. So in a 5000 square foot building with 1750 square feet of duct surface area, duct leakage should be about 23 CFM25, based on this standard. In actual practice, duct systems are much more leaky. In a group of 46 buildings with average floor area of about 5000 square feet, average CFM25 was 1212 or 53 times more leaky than the SMACNA standard (Cummings et. al., 1996). A substantial portion of the duct leakage that exists in these buildings occurs in building cavities that are used as part of the air distribution system.
Several Home Energy Rating programs around the country use a standard of 30 CFM25 per1000 square feet of floor area. If this standard were applied to 5000 square foot (commercial) buildings, then CFM25 would be 150, or eight times tighter than actually exists. Since cooling systems in small commercial buildings have approximately twice the capacity per 1000 square feet compared to residences, it is reasonable to expect greater duct leakage per 1000 square feet in commercial buildings because the ducts are larger.
3. Protocols for Verifying Compliance with Air Flow Management Standards
Section 2 presented standards for building airtightness, building air boundary location, air pressures, and air distribution systems. Section 3 presents protocols to verify compliance with these standards.
3.1 Building Airtightness Test Protocol
A blower door test is performed using one or more blower doors installed in exterior doorways of the building to obtain CFM50, the air flow rate through the building leak sites when the building is at 50 Pa wrt outdoors. CFM50 actually represents the cumulative hole size of all leaks in the building envelope. In a multi-point test, the blower door fan speed is varied to achieve a range of building pressures. It is not necessary, however, to reach 50 Pa to have a successful test. If wind induced pressure fluctuations are not large, then a multi-point test even in the range of 5 to 10 Pa can yield a successful test. The air flow and dP data are input to a computer program, and a projected CFM50 is determined. Alternatively, a single point test can be performed at 50 Pa, yielding a useful but somewhat less accurate test result compared to the multi-point test.
The author feels that the test should be done in the depressurization mode. If the building is pressurized, exhaust fan dampers, sky lights, jalousie windows, awning windows, and suspended ceiling tiles will tend to be pushed open and create artificial leakiness, and will therefore overestimate building leakiness. Prepare the building for testing by closing exterior doors and windows, opening interior doorways, turning off HVAC systems, and masking off exhaust air, make-up air, and OA openings. Switch vented combustion appliances to “off” or “pilot” so that they cannot operate during the test. Add water to dry traps in floor drains to prevent sewer gas entry into the building. Alternatively, building airtightness can be approximated using one or more HVAC fans. For example, in a restaurant, the kitchen exhaust air and make-up air can be turned on. It is common for the imbalance between exhaust and make up air flows to produce -10 to -50 Pa in the building. By measuring both air flows, the net air flow can be obtained, and a value for CFM50 can be estimated.
3.2 Air Boundary Location Protocol
The primary building air boundary (the ceiling or the roof deck) is identified during the blower door test. With the occupied space at -50 Pa, pressure in the ceiling space is measured. Whichever plane shows the greater dP is the primary air boundary. If the ceiling space pressure is -40 Pa wrt outdoors, for example, then the roof is the primary air boundary. We would like the primary ceiling/roof air boundary to have airtightness of about 0.5 CFM/square foot (required in the Florida Energy Code), suspended ceilings generally have airtightness of about 5 CFM50/square foot of ceiling, or 10 times more leaky than acceptable. When checking ceiling space pressure, we would like that dP to be at least -45 Pa wrt outdoors, indicating that the roof deck is about 10 times tighter than the ceiling.
3.3 Air Pressure Measurement Protocols
Pressure mapping is performed to observe building and zone pressure response to a variety of typical HVAC control states. A tube (1/8" i.d.) should run two building heights distance from the building, and away from structures, trees, etc. Terminate the tube with a static pressure probe in dry sandy or loose soil to provide shielding from the wind. Connect the other end of the tube to the “reference” or “low” side of a manometer with resolution of 0.1 Pa or better. A tube attached to the “high” tap can be run to any zone, and the dP measurement will be expressed as “zone pressure wrt outdoors”. Using the long-term average feature (available on some brands), it will display a running cumulative average, and by watching this display one can observe oscillation due to wind dampen and disappear.
Prepare for the test by opening interior doors and closing windows and exterior doors. Testing could proceed as follows (all pressures wrt outdoors unless otherwise indicated). First, measure building pressure with all HVAC equipment off. Then turn on exhaust fans (and make-up air) and record building dP. Then turn off the exhaust fans, turn on the AHUs, and record building dP. Then close interior doors and measure dP in the central zone (hallway) wrt outdoors, and dP in the closed rooms wrt the hallway. Open interior doors, turn on exhaust and make-up air fans, and record building dP. Also measure dP in various rooms and building cavities. For example, measure dP in the mechanical rooms wrt the hallway. Pay special attention to zones where combustion devices are located. dP beyond -5 Pa can backdraft atmospherically vented combustion devices. dP beyond -15 or -20 Pa can cause flame rollout from atmospherically vented water heaters. Measure pressure in wall cavities where you suspect connection of the wall to the return plenum. If a ceiling space is used as a return plenum, measure plenum pressure wrt the hallway. Measure pressure in attic space or crawl space.
3.4 Air Distribution Measurement Protocols
Duct system airtightness can be measured by a duct tester which is like a small blower door. The test is done with the system off and registers masked. Normally, one duct tester is attached to a return grill and a second to a supply register. The duct testers draw air from the duct system to achieve -25 Pa in both sides of the system. The flow rate through the two duct testers is called CFM25. The test can also be done with one duct tester. With the duct system split (sealed) at the air handler, each side of the system is tested separately.
Managing air flows in buildings is very important, because not managing building air flows can produce a range of problems, including energy waste, comfort problems, building materials damage, mold/mildew growth, high or low RH, combustion safety problems, and poor IAQ. Unless air flow is under control, we cannot insure the quality of our indoor environment. In some buildings in hot and humid climates, moisture accumulation in walls or high indoor RH caused by UAF can lead to catastrophic consequences, including sick buildings, evacuated buildings, and lawsuits.
Because UAF is a widespread problem with potentially catastrophic consequences, it is important that proper management of air flows and dPs be a major focus of commissioning both new and existing buildings. In order to properly address UAF issues, it is necessary to establish air flow management standards. This report contains 17 recommended standards by which to evaluate whether air flows will be properly controlled. This report also contains a variety of diagnostic protocols by which to verify that the building is in compliance with these standards. These protocols include testing for building and duct system airtightness, building air boundary location, measurement of HVAC system air flow rates, building air flow balance, and pressure mapping.
Cummings, J.B., C.R. Withers, N. Moyer, P. Fairey, and B. McKendry. 1996. "Uncontrolled Air Flow in Non-Residential Buildings; Final Report" FSEC-CR-878-96 Florida Solar Energy Center, Cocoa, FL.
Sheet Metal and Air Conditioning Contractors' National Association, Inc. (SMACNA). 1985. "HVAC Duct Construction Standards -- Metal and Flexible", first edition.