This article was originally published in the January/February 1996 issue of Home Energy Magazine. Some formatting inconsistencies may be evident in older archive content.


| Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy | EREN Home Page |

Home Energy Magazine Online January/February 1996

Duct Improvement 
in the Northwest

Part I: New Construction
and Retrofit

by Ted Haskell

Ted Haskell is an extension energy agent with the Oregon State University Extension Service.


The Residential Construction Demonstration Program provides pointers on duct sealing methods and program design.
In the Northwest, energy losses in forced-air systems average 25%-30% in electrically heated homes, with some homes being significantly worse. This appears to be the case in other regions of the United States as well.

The Bonneville Power Administration (BPA) created the Residential Construction Demonstration Program (RCDP) to test innovations in energy efficiency and evaluate their effectiveness. In 1994 RCDP turned its attention to forced-air distribution systems. Under the program, state energy offices in Idaho, Montana, Oregon, and Washington worked with BPA to recruit and train contractors to build tight new systems and to retrofit older ones. Researchers from the energy offices tested each of the 41 new houses and 25 retrofit houses for duct tightness and house operating pressures.

This article summarizes the most valuable results and recommendations of these case studies. The details can be found in the RCDP IV Final Report: Improved Air Distribution Systems for Forced-Air Heating. RCDP was designed as a group of case studies, and the results have not been analyzed for statistical significance. However, the project provided valuable insights for creating other duct-sealing programs.

Findings in New Houses Duct Tightness

Average duct leakage to outside, when pressurized to 50 Pascals (Pa) using a Duct Blaster, was 74 cubic feet per minute (CFM50). This compares to leakage rates of 350-450 CFM50 found in typical systems. The RCDP houses had a large range in leakiness, from 0 CFM50 to 465 CFM50. So, while distribution systems can be made tight, not all contractors get their systems tight on the first try. (RCDP used 50 Pa instead of 25 Pa as a test pressure because the results are less affected by weather conditions at higher pressures.)

Contractors reported duct-sealing costs averaging $355 per house beyond what they normally would spend on installation, with a range from $100 to $900. Generally, the systems that cost more to seal were tighter. For example, the Montana contractors averaged $500, and their systems were very tight (28 CFM50). Washington contractors averaged only $150, but their systems were leakier (81 CFM50). It may well be that the roughly 50 CFM50 reduction is not worth the extra $350.

Here are the highlights of what we learned:

  • Duct mastic works, although cold and wet weather make it difficult to apply.
  • Workers don't like mastic. It ruins clothes and tools, especially when users are novices.
  • Tape-applied mastic seals well. These are the aluminum tapes with 15-mil butyl backing.
  • Air handlers are leaky.
  • Electric air filter housings add large leaks that sometimes cannot be sealed. Ironically, these return leaks are often located in garages or basements, where they pull pollutants into the indoor air!
  • Building cavities used as returns are leaky.

Figure 1. Since most new houses are built with only one return grille, pressure imbalances often occur between bedrooms with closed doors and the rest of the house. Pressure relief should be built into these rooms, and this can be done in many ways. This diagram shows a ducted ceiling bypass into the hallway. With this type of bypass, the duct limits noise transfer. A less expensive option would be a direct vent through the wall into the hallway , or simply cutting the bottom off the door.

Pressure Relief

One source of inefficiency in forced-air systems is differential house pressures caused by not having returns in all rooms. Closed rooms with supplies only have positive air pressures, while rooms with returns have negative pressures. For pressure relief, RCDP contractors either added returns or installed pass-through grilles or ducts between bedrooms and the main body of the house (see Figure 1). Pressure relief grilles worked when they had at least 1 square inch of open (net free) area for each CFM of supply air delivered to the room. Grilles sized significantly below this did not reduce room pressures adequately. Some builders were able to install these grilles for under $200 for a typical three-bedroom house.

Pressure relief serves several purposes:

  • It prevents depressurization of combustion zones.
  • It reduces heat loss due to air leakage driven by the heating system.
  • It reduces moisture loading of building cavities caused by air leakage.
Pressure relief is especially important in houses with combustion devices in the heated space (in this study, wood stoves and fireplaces were commonly found). Without pressure relief, rooms often become depressurized enough to backdraft these devices. Recommendations to Contractors Don't use building cavities as ducts. Panned floor joists (cavities that have been made into ducts by attaching sheet metal to span the opening between two joists), chases, under-stair cavities, between-floor joist spaces, and other building cavities are almost always leaky. In the RCDP, a couple of systems used panned floor joists because the contractor was sure he could make them tight. He couldn't-his systems were the leakiest in the study.

Locate as much of the system inside as possible. When all ducts and the air handler are located inside heated spaces, distribution systems are much more efficient-sometimes over 95% total useful heat delivered. This requires early cooperation among the HVAC contractor, the general contractor, and the designer. It also requires space in the living area, which many builders are unwilling to provide (See Researchers Approach Builders on Duct Location, HE Nov/Dec '95, p. 6).

Plan ahead for access to ducts. Leaks are often not sealed because it's hard to reach them. This is especially true near the air handler. Leave enough space to get to critical leaks.

Provide a pressure-balanced system. In rooms with no returns, provide pass-through grilles or other means of allowing air to get back to returns. Grilles should have free area equal to at least 1 square inch per CFM of supply to room.

Use high-quality sealing materials. Water-based mastics and tape-applied mastics (foil tape with at least a 15-mil sealant) appear to work best. For leaks wider than 18 inch, reinforce mastic with mesh tape.

Figure 2. The proper procedure for making flex duct connections.

Don't confuse sealing with mechanical fastening. Tapes and sealants are not designed to hold connections together. Solid mechanical connections are crucial to effective duct sealing. Use sheet metal screws for metal-to-metal connections. Use compression straps for flex duct (see Figure 2). Plastic compression straps require a special tool to achieve adequate tightness.

Use a duct tester. A duct tester (such as a Duct Blaster) makes it easier to evaluate progress and to identify the techniques that work best. It also helps to prevent backsliding or loss of quality as new employees are brought on board. Finally, it is one of the few ways to demonstrate the quality of the job to customers.

Recommendations to Utilities The RCDP demonstration showed the importance of getting new forced-air systems tight. The systems were made twice as tight for two-thirds of the cost of the retrofit program. Regardless of how contractors are encouraged to build tight systems (approved contractor status, exemplary home programs, financial incentives, and so on), the utility will need to set program standards and provide some training in how to meet those standards.

Setting Standards

System tightness. Set a performance target verified by testing. Testing duct leakage is also good training-contractors get better at sealing ducts when they can see and measure what's working. The target for RCDP was no more than 50 CFM50 leakage to outside or 0.02 CFM50 per square foot of living space, whichever was greater. The average leakage rate was 80 CFM50. This average included some notable failures-most systems actually came close to the target.

Unfortunately, duct leakage to outside is probably not a very practical test for a new construction program. It requires the use of both a blower door and a duct tester, and house construction must be completed. Instead, utilities could base their standards on total duct leakage, which requires only the use of a duct tester (the system must be complete, with the air handler hooked up, but the house can be at any stage of completion). This leakage rate averaged twice the duct leakage to outside in RCDP. Based on this study, a standard of 150 CFM50 or 0.06 CFM50 per ft2 of conditioned space would be a good starting point for a utility program.

Materials. Requiring the use of sealants that meet Underwriter's Laboratory (UL) standards would be a good starting point. (The State of Florida now requires duct sealants to satisfy UL requirements.) Another option would be to require the use of duct mastics or tapes manufactured for sealing the type of duct material being used. Require the use of reinforcing mesh tape for gaps over 18 inch or where joints are under stress. Sealants for flex duct should be those approved by the duct manufacturer.

Application of sealants. Sealants should be applied according to the manufacturer's instructions. These instructions may include surface cleaning and preparation, temperature and pressure requirements, and other details. When sealing a flex-to-metal connection, the inner liner must be sealed to the metal.

Mechanical fasteners. Require sheet metal screws for metal-to-metal joints. Flex connections require compression straps on both the inner liner and the outer liner. Plastic compression straps must be tightened using the appropriate tool.

House Operating Pressures

Rooms with supplies only should be pressure relieved so that they are not pressurized more than 3 Pa with reference to the main body of the house. Pressure relief openings (pass-through grilles, door undercuts, passive ducts, or combinations) should total 1 square inch of net free area for each CFM delivered to the room.

The effectiveness of pressure relief should be tested after the house is completed, using a digital micromanometer, with all doors closed.

Findings in Retrofit Houses Retrofit duct sealing was done primarily by weatherization contractors. The procedure was for state energy office researchers to test the system in the morning, direct the contractors' sealing efforts, and then retest the system to measure any reductions in leakiness.

Sealing the right leaks, and sealing them the right way, is crucial. Here, the air-handler-to-plenum and plenum-to-ducts connections are the center of attention, and mastic is the sealant of choice.

Leakage Reductions

Sealing reduced the average leakiness from 340 CFM50 to 160 CFM50. Because supply leaks tend to be more costly than return leaks, it was important to know which side of the system the leakage reductions were coming from. In the 15 houses where the supply and returns were measured separately, on average half of the leakage was on the supply side. The percentage varied a great deal from house to house.

Retrofit duct sealing costs averaged $335 per house, with a range of $120 to $630. A more interesting figure is the cost per CFM50 in leakiness reductions. The average cost per CFM50 was $3, but the range was extreme: from 50 to $11.

Advance Screening

Some houses already have relatively tight distribution systems, so it might not make much sense to spend time and money sealing them. The fastest test for screening houses is the pressure pan test, which requires a blower door and a digital manometer. It's an easy screening procedure in programs where a blower door is already in use.

John Tooley of Natural Florida Retrofit has developed pressure pan screening criteria whereby systems are categorized as tight, gray, or loose. When the RCDP retrofit systems are categorized using these criteria, it becomes clear that systems categorized as loose were sealed more cost-effectively. The loose systems averaged 450 CFM50 leakiness before sealing, and 190 CFM50 after sealing. This was roughly five times the reduction in the houses that were categorized as tight or gray. Furthermore, the average cost per CFM50 of leakage reduction for loose systems was only $1.60, while the average for the gray and tight systems was over $5.

Know your enemy-the panned floor joist. A joist cavity makes a very poor duct, as it is extremely difficult to make airtight.
Recommendations to Contractors Much of what applies to new construction also applies to improving existing distribution systems. Solid mechanical connections, high-quality sealing materials, and testing to assure quality are as important in retrofit as they are in new construction. Some recommendations that apply only to retrofit are listed below:

Screen using the pressure pan test. This establishes whether the duct system is leaky, and it helps to locate leaks. This test requires a blower door rather than a duct tester.

Screen houses for safety. Before ducts are sealed, the house should be tested for combustion zone depressurization under typical and worst-case operating conditions (combinations of furnace blower, exhaust fans, and door closure) to make sure that combustion devices are not backdrafting. Duct sealing may worsen backdrafting under some conditions, posing health and safety risks to the occupants. Homes with safety problems should not have work done on them unless the occupants are willing to have pressure relief measures installed.

Screen houses for other cost-effectiveness factors. Some systems may not be cost-effective to seal even if they're leaky. Examples include homes with inaccessible ducts (the definition of inaccessible changes with experience) and homes with low heating bills (very small homes or homes with wood stoves).

Prioritize sealing efforts. Besides using the pressure pan test to locate leaks, there are a couple of other general rules that will increase the effectiveness of duct sealing:

  • Start with the leaks that are closest to the air handler, where the pressure difference between the duct and outside is highest.
  • Prioritize supplies, since these leaks are more costly.
Sometimes sealing the building shell is most effective. Ducts that are inaccessible are occasionally located in building cavities that can be sealed. For example, when a duct travels through a chase into an attic, it may be easier to seal the top and bottom of the chase rather than the duct.

Set tightness targets. In every job, there's a point of diminishing returns where extra effort saves so little energy that it's no longer worthwhile. Unfortunately, RCDP and other research has not yet provided us with a simple description of this point. Until this information is available, contractors should watch their pressure pan test numbers as they work, and stop working when the numbers change very little. With experience, pressure pan targets can be set for various house types.

to Utilities Specifications for materials, mechanical fasteners, and application should be essentially the same as in new construction. However, unlike new homes, where duct systems should all be built tight, retrofit homes need to be selected for utility programs.

Selection Criteria

Safety. Houses that haven't been screened for combustion safety should not be part of a utility-sponsored program. Houses should be tested for existing carbon monoxide problems, existing backdrafting, worst-case combustion zone depressurization, and the impact of any duct or house sealing. When homeowners are unwilling to have these problems corrected, no further work should be done.

Pressure pan test. This test is to determine whether the ducts are leaky enough to justify utility expenditures. Houses with fewer than three pressure pan readings above 2 Pa are unlikely to be cost-effective to seal.

Heating and cooling bills. Even if the ducts test leaky, it doesn't make economic sense to seal them if the utility bills are low. This may be the case when there's a wood stove in use.

Accessibility. When ducts are difficult to get to, sealing costs go up and effectiveness goes down. Start with simple houses where the ducts are accessible.

Setting Standards

Performance specifications are more difficult to set for retrofit. Some systems just can't be made as tight as others, through no fault of the retrofit contractor. Here are some considerations for setting targets:

Tightness targets by house type. The RCDP experience indicates that 125-150 CFM50 leakage to outside is a reasonable target for simple houses. For complex houses, doubling this target may be appropriate. Measuring leakage to outside requires both a duct tester and a blower door, which may be too time-consuming for utility programs. Using pressure pan targets would require only the blower door. For simple houses, pressure pan readings should mostly be brought below 1 Pa, with a few at 1.5 Pa.

Fleet averages. Not all systems will be able to meet the target. To avoid contractor frustration, targets should be fleet averages. Some houses should be brought below the target to make up for those that can't.

Leakage reduction per hour. It can be tough deciding when to quit work on a system that won't meet the tightness target. By testing periodically, contractors can track leakage reductions per hour (either CFM50 or Pascals). For these houses, set a target for leakage reductions per hour of labor invested. When the reductions get too small, work should stop.

Testing Out

Key safety tests should be repeated after completing work. Specifically, combustion appliance zone depressurization should be tested under worst-case conditions. If combustion appliance zones exceed depressurization limits, they must be pressure relieved.


Overall Utility Program Recommendations

Begin with New Houses

It's easier to get tight ducts in new construction. There are fewer technical problems and it gives everyone involved a chance to learn about materials, suppliers, and techniques.

Getting Started in Retrofit

Start with a pilot project. Retrofitting is more complex than new construction. Start the program with only one or two contractors. Make sure each contractor will have enough work to make it worth the initial investment in time and money.

Start with simple houses. Start with single-story houses where the ducts and the air handler are in crawlspaces, attics, or garages. Most of the ducts are accessible, and most leak to the outside. Sealing is simpler and more likely to bring savings.

Collect data for future decision making. We're often asked by utilities for more detail on how much energy duct programs will save. Unfortunately there's not much data out there-utilities need to collect their own data on their pilot programs. For most utilities, analyzing heating and cooling bills with PRISM will provide valuable information (see Advancing the Art of PRISM Analysis, HE July/Aug '95, p. 19).


Train contractors in the field. Trainers in all four states in RCDP felt that the most valuable training took place on the job site. Classroom training is valuable in getting the support of company owners, but to get to the workers, you've got to get on site.

Use field measurement to make training more effective. Duct testers and blower doors give instant feedback on what's working and what isn't. This feedback quickly focuses contractors and workers on key points, and it helps get new employees up to speed.

| Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy | EREN Home Page |


Home Energy can be reached at:
Home Energy magazine -- Please read our Copyright Notice


  • 1
  • NEXT
  • LAST
SPONSORED CONTENT What is Home Performance? Learn about the largest association dedicated to home performance and weatherization contractors. Learn more! Watch Video