Mark P. Modera is a staff scientist in the Indoor Environment Group at

In essence, the improvements are based on our findings which point to better locations to measure pressure differences and which encourage testing while the HVAC system is running. In the first case, technicians thought that it was sufficient to measure pressure differentials between the inside and the outside of the envelope. We found better places to measure pressures--principally in the supply plenum. In the second instance, the ASTM protocols add testing for the system in operation. The blower door subtraction and flow hood methods have traditionally applied to systems that are not operating, an unnatural state.

New and Improved

Blower Door
Subtraction Method

Because the groups doing field tests in the past were accustomed to using fan pressurization--in other words, blower doors--to measure envelope leakage, the first commonly used duct measurement technique involved blower doors as well. The tests find the leakage areas for the total house (envelope plus all ducts), for the supply ducts plus envelope, and for the house envelope alone. The desired leakage rates can be found by subtraction:

Supply ducts plus envelope leakage - envelope leakage = Supply ducts leakage.

Total leakage - House envelope leakage = All ducts leakage.

All ducts leakage - Supply ducts and envelope leakage = Return ducts leakage.

The basic ASTM protocol specifies that the house be pressurized to 25 pascals (25 Pa). The standard specifies 25 Pa--not the 50 Pa commonly used in weatherization protocols--because it is close to the normal duct pressures that occur when the system fan is running.

Flow Hood Method

The second technique to come into common use involved the use of a direct measurement of duct leakage flow using a flow hood in combination with the blower door. Researchers devised this technique because the changes in blower door flows between the sealed and unsealed configurations are generally small compared to the total blower door flow, and therefore can be dramatically affected by the uncertainties introduced into blower door measurements. Measuring the duct leakage flow directly with a flow hood can significantly reduce uncertainty.

The ASTM flow hood measurement protocol consists of two blower door pressurization tests, one that measures the leakage of the supply ducts, and another that measures the leakage of the return ducts. Auditors can perform these simultaneously with two flow hoods.

Again, our improvement was to monitor the pressure differential across the duct system (utilizing some kind of static pressure probe, like a pitot tube) at two locations on the return side and two locations on the supply side. Rather than formulating a correction factor, as with the blower door subtraction method, we used the pressure reference in the leakage calculations.

Leakage Measurements
By the Book--
and In the Field

We studied leakage measurements in 30 California houses following the standard ASTM measurement methods, comparing them with the old methods and with each other. We studied both techniques in pressurization and depressurization modes, to check for accuracy.

Blower Door
Subtraction Results

In the study, we made the three sets of measurements while pressurizing with the blower door (and then depressurizing). We found that, without measuring duct pressures during sealed duct conditions and without employing the correction factor to offset biases that arise in the subtraction method, inspectors would underestimate return duct leakage by almost 50% on average, and total leakage by 30% on average. Underpredictions would range as high as 76% for the return leakage, and 64% for the supply leakage. The bias in subtractions of sealed duct test results from unsealed duct test results stems from leakage between the house and the theoretically sealed duct. The correction is based on treating the house plus the duct system as a series of leaks.

Flow Hood Results

Again, for the flow hood measurements, direct measurements of duct pressures--not just the pressure of the indoor space--for determining duct leakage characteristics are just as important as with the blower door method. Here, though, the actual measurements are used, not a correction factor.

Taking actual duct pressure measurements is more crucial for supply ducts than for return ducts, largely because supply registers and ducts resist air flow more than return registers and ducts. We found that the calculated supply duct leakage would be 33% too low on average, and as low as 76%, if the house pressure were substituted for the measured duct pressure.

Blower Door or Flow Hood?

The first and foremost conclusion is that, whichever method is employed, only by making internal duct pressure measurements and using appropriate analysis procedures based upon those measurements can reliable and unbiased estimates of duct leakage to unconditioned spaces be made.

That said, the flow hood method seemed to have an edge over the blower door subtraction method. Although we observed no significant bias between the two methods, the flow hood method seemed to be superior because of better internal consistency between pressurization and depressurization results, and it does not render impossible results such as negative leakage figures (see Table 1). These results are not that surprising, considering blower door test results can vary by as much as 6% from test to test without changing anything. This translates to an average uncertainty of 27% for total duct leakage, and 53% for either supply or return leakage.

We mentioned that we also used auxiliary ASTM techniques to measure duct leakage, specifically with the system fan running. The auxiliary tests confirm that the pressures driving duct leakage are much larger than envelope pressures. They can also help estimate the leakage rate of a given single duct system during normal operation, essential for analyzing the energy efficiency of the whole duct system. n

Data for 28 houses obtained using two methods specified by the American Society for Testing and Materials--the blower door method and the flow hood method. Note: Negative pressures are impossible and indicate computational errors.

Related Articles