This article was originally published in the September/October 1994 issue of Home Energy Magazine. Some formatting inconsistencies may be evident in older archive content.
| Home Energy Home Page | Back Issues of Home Energy |
Home Energy Magazine Online September/October 1994
User-Friendly Pressure Diagnostics
by Jim Fitzgerald, Robert Nevitt, and Michael Blasnik
Jim Fitzgerald is a Minneapolis-based trainer and consultant. Robert Nevitt is business manager of the Energy Conservatory in Minneapolis. Michael Blasnik is research director of GRASP, a Philadelphia-based nonprofit, but is joining Proctor Engineering Group in its new Boston office.
Here is an easy-to-understand explanation of pressure diagnostics. The user friendly approach focuses on measuring pressures rather than just leakage to help you quickly determine which weatherization treatments a home needs. Know what you're measuring before you start work.
Using pressure measurements to find and quantify air leakage in houses has been a hot topic over the past two years. Packed sessions at recent conferences are testimony to the intense interest these techniques are generating in the weatherization community. Advanced pressure-diagnostic techniques have greatly furthered our understanding of the complicated nature of air movement through building cavities and chaseways (see In Search of the Missing Leak, HE Nov/Dec '92, p.27). This new understanding is helping answer many questions about why some houses respond well to weatherization treatments while others provide disappointing results.
Unfortunately, transferring these new diagnostic techniques to the field can be a difficult and even intimidating task. The graphs and mathematical formulas associated with the techniques have been known to create math anxiety in many a crew member, while others complain that performing all the tests adds too much time to the house inspection. Yet it has become increasingly clear that we need a better process to guide us in developing treatment plans for houses. In many programs, lost opportunities abound even in fully treated houses.
So how do we choose from the growing list of diagnostic tests and treatment options available? If our goal is improving program results, and not doing research, then not every test procedure and treatment option is appropriate for every house. We want the minimum testing to help make the right choices fast, and the minimum treatment to achieve the desired results.
Here is an easy-to-understand explanation of pressure diagnostics plus some simple examples of pressure tests. With this information, you should have a better understanding of when these techniques may be appropriate and how they can be quickly incorporated into your diagnostic procedures.
Thermal and Pressure Boundaries
One of the most important pieces of information we need to make informed treatment decisions for a house is the location of thermal and pressure boundaries.
Insulation is normally installed to reduce conductive heat loss between conditioned and unconditioned spaces. This surface is called the thermal boundary, or thermal envelope, and is typically used by auditors and crews to determine the intentionally conditioned area of the house. Conventional wisdom tells us that thicker insulation along the thermal boundary or a smaller boundary surface area produces better thermal performance and energy savings.
Plywood, gypsum board, weatherstripping, foams, and other air-barrier materials, on the other hand, all serve to slow the exchange of conditioned air with outside air. While some insulations (such as rigid foam board and dense-packed cellulose) can act as an air barrier, typical insulation applications do little to impede air movement. The location of the actual block between the inside and outside air is called the pressure boundary. Conventional wisdom tells us that sealing holes in the pressure boundary will also produce energy savings.
Blower-door tests are routinely used to measure the airtightness of this pressure boundary and to measure changes in the boundary from airsealing work. Most experts would agree that the advent of blower-door tests has greatly aided crews in directing cost-effective airsealing treatments. Unfortunately, a blower-door test by itself does not tell us the actual location of the pressure boundary. And it is this key piece of information that can make or break a treatment plan.
It is not uncommon for the pressure boundary and the thermal boundary of a house to be in different locations. For instance, the pressure boundary in a house may actually be the roof boards rather than the insulated attic floor. Misalignment of these boundaries can create numerous problems with conventional treatment decisions. First, when the two boundaries are not aligned, the actual conditioned space of the house may be significantly greater than you assumed in your audit. Misalignment can also dramatically increase the surface area over which heat loss can occur. This can happen by unintentional heating of unconditioned spaces like attics, or convective looping occurring within interior walls.
Just as importantly, when the boundaries are misaligned, the house may not respond to treatments as expected. For example, crews may waste time and money by sealing only visible air leaks in houses which are dominated by hidden cavity leakage. This can result in no reduction in house airtightness, despite the crew's plea that several hours were spent sealing attic bypasses.
Finally, since most residential insulations require still air to function properly, houses tend to perform better when the pressure boundary lines up with the thermal boundary. When the boundaries are aligned, we are less likely to have air movement through insulation, which causes R-value degradation.
Finding the Pressure Boundary
Simple pressure tests can help us determine which houses will respond more predictably to treatments and can also reinforce training goals. We do this by learning which part of the house structure is really inside the pressure boundary, which part is outside, and which parts are a mixture of the two. We can no longer assume every house behaves like a simple box with holes. Using the right pressure test, fast production houses can be quickly differentiated from the complicated house having multiple interconnected leaks with connections to ducts, side attics or porches.
Let's look at three examples of simple pressure measurements and how they help us decide on treatment options. Before we begin, we need to set some ground rules. First, our pressure tests are always conducted with a blower door depressurizing the house by 50 Pa with respect to (WRT) outside. This means that all exterior windows and doors in the house are closed, interior doors open, and the blower door is blowing air out of the house, creating a lower pressure inside the house than outside.
If the house has a forced-air duct system, we need to decide whether the duct system should be considered outside the house, with high savings potential, or inside the house, with low savings potential. Set up the house for the test accordingly (see Figure 2). Clearly decide, and leave no in-between semi-conditioned areas. Safety testing must be done after the exterior-interior boundary is repaired, because HVAC systems cannot create pressure imbalances if no effective pressure boundary exists before treatment. Also, if there is no effective pressure boundary between the basement and the upstairs, or between rooms, then no pressure problems are created by duct leaks to the inside that remain after treatment.
Our simple tests will include two or more of the following pressure measurements taken with the blower door operating:
Before turning on the blower door, place a pressure hose through the attic hatch so that one end is in the attic and the other can be reached from inside. Place it 2-3 ft past any direct leaks. If you have a digital gauge, use it for all the readings, or use the same gauge to measure house-to-attic pressure and house-to-outside pressure. Detach the blower door gauges and use the top (0-60 Pa) gauge for all the measurements; leave the fan running to maintain an even -50 for all the tests.
Pressure and Air Leakage Rates
Before we go any further, let's take a quick look at the relationship between leakage rates and pressures.
Figure 1 shows 7 examples of how leakage rates between a house and attic and between the attic and outside relate to simple pressure readings taken with the house being depressurized by a blower door by 50 Pascals (Pa).
For example, if you measured a house to attic pressure of -12 Pa and an attic to outside pressure of -38 Pa, the leakage rate from the house to the attic would be twice the leakage rate from the attic to outside. As the ceiling air pressure boundary is tightened (for instance by sealing attic bypasses), the house-to-attic pressure moves closer to -50 Pa, and the attic-to-outside pressure gets closer to zero. By the time most of the attic leaks have been sealed and the house to attic pressure reaches -48 Pa, the leakage rate between the house and the attic is 1/8th the attic-to-outside leakage rate.
Importantly, Figure 1 only gives the relative size of the leakage rates. You need to know the actual size of one side of the leakage path (house-to-attic or attic-to-outside) to determine if the leakage path is significant. Fortunately, attic vents can often be used to estimate the attic-to-outside leakage rate. For example, let's consider a house with a tight roof and one single attic vent installed. A single common attic vent has an effective leakage area of about 25 in2. Therefore, if we measure a house to attic pressure of -12 Pa, we can estimate from Figure 1 that the house-to-attic leakage area is about 50 in2 (twice the attic to outside leakage area). If we seal the attic until the house-to-attic pressure equals -48 Pa, we estimate that only 3 in2 of house-to-attic leakage remains (one eighth the attic-to-outside leakage area). Our attic pressure measurement tells us that this bypass job will now pass any inspection.
On the other hand, if you are working on a house with loads of attic venting, an attic pressure of -48 or -49 before work begins may still mean plenty of unsealed attic bypasses. For a house with 600 in2 of attic venting, a house-to-attic pressure of -48 Pa means there is roughly 75 in2 of bypass area unsealed; plenty of work left to be done--about 750 CFM50.
Leaky Interior with Tight Exterior
House #1 has an insulated attic with large attic bypasses, a very tight roof with no attic venting, and leaky supply ducts running in the attic. With the house depressurized by 50 Pa (for example, the digital gauge reads -50 Pa of house pressure WRT outside), we measure the pressure of the house WRT the attic. Our digital gauge reads close to 0. Now we take a pressure pan reading on a ceiling supply register. The pressure pan reading is also close to zero, even though we know there are leaky ducts in the attic. What do these readings tell us?
In this case, the pressure boundary (the roof) is above and outside of the thermal boundary (the attic floor). Since the house ceiling is very leaky to the attic, the ceiling can not hold pressure and act as a pressure boundary. The roof is the tightest air barrier so becomes the pressure boundary by default. Conditioned air moves through and above the insulation in the attic floor lowering the effective attic R-value and increasing the envelope surface area and overall heat loss.
The leaky supply ducts in the attic do not show up in our pressure pan readings, then, because the duct leaks are located within the house pressure boundary. The misaligned pressure boundary effectively hides the duct leaks from the pressure pan.
Because the pressure boundary is the roof and not the attic floor, the cost and savings potential for this house will not be well-correlated with before-treatment CFM50 measurements. Although we think this house badly needs attic air-sealing, the blower door by itself will not provide good crew feedback on air-sealing effectiveness, because it is measuring the airtightness of the roof and not of the attic floor. As a result, although sealing the attic bypasses will result in little CFM50 reduction, it should produce significant energy savings. Instead of relying on the CFM50 measurement alone for feedback, then, crews will also need to look at the house-to-attic pressure.
First install one roof vent and re-measure house-to-attic pressure, to estimate the leakage rate to the attic using Figure 1. If the house-to-attic pressure becomes much closer to -50 Pa with the addition of one small roof vent, then we know that the house-to-attic leakage is relatively small and not worth a lot of time. Assuming that the house-to-attic pressure doesn't increase greatly, seal attic bypasses until house-to-attic pressure is close to -50 Pa, then check attic ducts with a pressure pan and continue duct-sealing until pressure pan readings are less than -1 Pa. Now add the remainder of attic vents needed and re-insulate the attic as necessary. By sealing the attic bypasses and venting the attic, we re-align the pressure boundary with the thermal boundary. You can expect an increase in energy use and CFM50 reading for the house by adding roof vents without airsealing the attic well.
Leaky Exterior with Tight Interior
House #2 was heavily vented and air-sealed years ago using the old technique of sealing all interior baseboards, wall outlets, and other wall/ceiling joints. With the house depressurized to 50 Pa, we measure a house-to-attic pressure of -50, and a house-to-interior wall cavity of -50. The interior wall cavity pressure of -50 tells us that a significant part of the pressure boundary is below and inside of the thermal boundary. The pressure boundary is the interior sheetrock, and not the attic floor. Interior wall volumes, including the plumbing wall and fireplace chase, appear to be open at the top to the vented attic and are actually acting as if they were outside (greatly increasing heat loss surface and energy use through convective looping). If there were ducts running through interior wall cavities, duct leakage to the outside would be significantly greater than if the pressure boundary was aligned properly.
Once again this house is a good candidate for air-sealing. But because of the misaligned boundaries, air-sealing may have little effect on blower door CFM50 readings, even though large energy savings are available. The energy-savings potential for this house may not be well-correlated with before-treatment CFM50 measurements. As with House #1, crews may need to use pressure to get feedback on their progress, rather than CFM50 alone. If there are ducts running through interior cavities, a reduction in duct leakage (indicated by the pressure pan) may show up with attic air-sealing, because the interior cavities will be sealed off from the outside.
Seal the tops of walls and chaseways until the house-to-interior chaseway pressure moves closer to zero. If ducts are running through chaseways, check the new pressure pan readings and continue to seal exterior leaks to the chase until the pressure pan reads close to zero (exterior duct sealing merits the extra detail). By sealing the attic-floor level, we re-align the pressure boundary with the thermal boundary.
Interconnected Leakage Paths
This one-and-a-half-story house has interconnections between the basement, sidewalls, and interior wall cavities to the first-floor ceiling joists, kneewall attics, and the peak attic. This house also uses a first-floor ceiling joist as a return duct for the upstairs. With the house depressurized to 50 Pa, we measure -5 Pa at a second floor register using a pressure pan and -20 Pa between the house and a first-floor interior building cavity.
Based on the construction details of this house, we suspect that many of the leakage paths may be interconnected so we conduct one more set of tests. We open an access door between the house and the side attic. Then we readjust the blower door to depressurize by 50 Pa and re-measure the duct and interior cavity pressures. We now find the duct pressure is -3 Pa and the interior wall cavity pressure is -11 Pa.
Because the duct and cavity pressures changed after we opened the attic access, we know that the interior cavity and ceiling duct chases are interconnected to the side attic. The pressure boundary in this house is complex--it's partially above the thermal boundary at the roof, partially below the thermal boundary at the interior finish level, and partially located in the resistance to air flow through the spaces between floors and walls. It is everywhere except where it belongs. Neither the attic or the interior wall and floor volumes function as completely inside or outside. Weather, wind direction, and duct operation all affect the mix of conditioned and unconditioned air in these volumes. Increased heat loss and comfort problems in this house may be more important than air leakage.
Importantly, when sealing leaks on interconnected complicated houses, all pathways for the leak must be sealed before you will see a reduction in measured house airtightness using a blower door. Sealing off only one of multiple pathways to the outside simply redirects air flow; it does not reduce air flow. Only when the last part of the interconnected leak is sealed does the blower door respond. In this type of house it is not uncommon for 50 ft2 of air barrier to be installed to reduce measured leakage by 500-1,000 CFM50 (roughly 50-100 in2 of equivalent leakage area).
Install dense-pack cellulose under all roof cavities that contain open walls, and all kneewalls, to effectively seal off side attics from interior cavities. This will reduce conduction losses by reducing surface area to the outside, reduce convective looping in interior floor and wall cavities, and reduce duct leakage to the outside by sealing leaky floor cavities (used as a return run) from the outside. Measured duct and interior cavity pressures should be close to zero when the job is complete.
This type of house often presents the largest savings potential, because one well-placed treatment can fix multiple problems. It is also the riskiest type of house, because failure to completely seal all leakage pathways will result in the house showing little or no savings.
Be careful interpreting single-point pressures. Simple pressure measurements are an aid to the blower door and other diagnostic tests. They are not a stand-alone substitute for infrared scanning, visual inspections, and plain common sense. Cost-effective work requires looking at actual utility, structural, and treatment costs. No one measurement can tell us everything.
We recommend against testing single closed cavities like walls or individual floor joists, unless they are obviously connected by a lowered ceiling, ductwork, or plumbing to a pathway with significant air flow. Larger volumes like attics, garages, and duct chases are less likely to be perfectly airtight, so measurements are easier to interpret. Good luck! n
Figure 1. The table below gives the relative size of the leakage rates. However, you need to know the actual size of one side of the leakage path (house-to-attic or attic-to-outside) to determine if the leakage path is significant.
Relative Size of Leakage Rates Measured Pressures*
House-to-Attic Attic-to-Outside House-to-Attic Attic-to-Outside
2 1 -12 -38
1 1 -25 -25
1/2 1 -37 -13
1/3 1 -41 -9
1/4 1 -45 -5
1/8 1 -48 -2
1/13 1 -49 -1
* with the house depressurized by 50 Pa.
Figure 2. Duct Location
Leaky Interior with Tight Exterior
Leaky Exterior with Tight Interior
Interconnected Leakage Paths
Related ArticlesIntroduction to Blower Doors (Keefe) Two Favorite Test Methods, By the Book (Modera) Beauty and the Beast Upstairs (Legg) Discovering Ducts: An Introduction Duct Fixing in America (Penn) Infiltration: Just ACH50 Divided by 20? (Meier) Leak Detectors: Experts Explain the Techniques (Proctor, Blasnik, Davis, Downey, Modera, Nelson, and Tooley) Mobile Homes: Small Zones, Big Problems (Kinney) New Group Hunts Bad Ducts (Obst) The New Monster in the Basement (Treidler) Selecting an Infrared Imaging System (Snell) Sizing Up Skylights (Warner) Telecommuting: An Alternative Route to Work (Quaid)
- FIRST PAGE
- PREVIOUS PAGE