Journal

Can smoke show how much a leak affects the whole-building result?

Passivebase Editorial Desk

A smoke trace can show where air moves through the envelope, but it cannot tell you how much that path contributes to the whole-building leakage result. To estimate the effect of sealing it, measure the building with a calibrated fan before and after the repair under comparable conditions. The difference estimates the change in the tested result, subject to measurement uncertainty and any changes to the test boundary or setup. [2] [3]

Smoke locates a path; the fan measures total flow

During a blower-door test, a fan pressurizes or depressurizes the building. A smoke generator can help identify leakage sites while the fan is operating. The smoke observation identifies a location; the fan measures the airflow needed to maintain the test pressure. For the residential and light commercial buildings covered by its guidance, the US Department of Energy’s Building America Solution Center describes airflow at 50 Pa as CFM50 and says ACH50 is calculated using enclosure volume. [2]

Use a paired measurement to estimate the change

A practical diagnostic sequence is to record the test boundary and setup, measure the enclosure, locate and document the suspected path, seal it, then repeat the measurement with the same configuration. Compare airflow at the same pressure and use the same reporting method. The Building America Solution Center procedure, based on RESNET standards, calls for documenting as-found conditions, indoor and outdoor temperatures, site elevation and conditioned-space volume. These details help show whether the two measurements are comparable. [2]

In a single-zone test, if the enclosure is measured at the same 50 Pa before and after sealing, the difference between the two airflow results estimates the effect of the repair on total flow at that pressure. It approximates the repaired path’s flow only when that path experiences the test pressure and the seal is complete. If the path runs through a buffer space such as an attic, wall cavity or floor void, sealing it can change the pressure in that space. An incomplete seal or changes to temporary seals, openings or test conditions also weaken the comparison. Treat the subtraction as an estimate of the change in the tested result, not a direct measurement of isolated leak flow.

What the instrument limits mean

ANSI/RESNET/ICC 380-2022 specifies a maximum error of 5% of measured flow for the airflow meter used in an enclosure test. For the manometer, the maximum error is 1% of the reading or 0.25 Pa, whichever is greater. At a 50 Pa test pressure, 1% is 0.5 Pa, so the stated limit is 0.5 Pa. These are equipment limits in the 2022 standard, not a complete uncertainty estimate for a particular test or a guarantee that two tests will agree within those values. [3]

Consider a hypothetical single-zone test with readings of 1,000 CFM50 before sealing and 960 CFM50 after. The observed difference is 40 CFM. At the standard’s 5% maximum airflow-meter error, the respective limits are 50 CFM and 48 CFM. For a rough illustration only, assume each error is independent and uniformly distributed within its limit; converting each limit to a standard uncertainty gives 50/√3 and 48/√3 CFM. If the same fan, gauge and configuration are used for both tests, part of the meter error is systematic and largely cancels in the difference, so treating it as independent gives a conservative (larger) uncertainty; repeatability then dominates. If repeated readings also show an assumed standard deviation of 10 CFM for each test, combining these four independent contributions by root-sum-square gives about 42 CFM for the difference. Under those stated assumptions, the observed 40 CFM change is not clearly distinguishable from the combined uncertainty. The assumed repeatability and error distribution are illustrative, not measured facts; a tester should estimate them from the actual equipment and repeat readings. [1] [3]

The National Physical Laboratory’s Measurement Good Practice Guide No. 11 explains that a measurement result is incomplete without an uncertainty statement. It identifies the instrument, measured item, environment, operator and other factors as possible contributors, and describes statistical analysis of repeated measurements as one way to estimate uncertainty. Resolution is the smallest difference that can be meaningfully distinguished, but resolution alone does not establish whether a change is detectable. Ask for the uncertainty estimate and the repeat readings when a before-and-after difference is small. [1]

What the report should let you check

A useful diagnostic record links the local observation to the whole-building measurements: where smoke moved, what repair was made, the test pressure and airflow before and after, the test boundary, temporary seals and information on repeatability or uncertainty. The Building America Solution Center guidance calls for test results and documented test conditions, while the National Physical Laboratory guide explains why a measurement result needs an uncertainty statement. If the observed change is small, compare it with the uncertainty estimated for those tests. [1] [2]

Sources