Supply-side airflow is straightforward: air leaves the equipment through defined supply ducts and exits at registers with relatively small cross-sections and moderate velocity. The supply side is designed to move air efficiently to specific destinations.
The return side has the opposite characteristics. Return plenums and ducts are typically larger in cross-section than supply ducts because they serve the same airflow volume through fewer, larger openings. Larger cross-sections mean lower velocity. Lower velocity means smaller pressure signals, which makes direct measurement less precise.
Three factors compound the difficulty:
None of these problems make return measurement impossible. They mean you need to pick the right method for the configuration in front of you and understand the accuracy limits.
When the return filter is located at the air handler (not at the return grille), the TrueFlow Grid replaces the filter and measures the total airflow entering the equipment. This is the easiest and most accurate return-side measurement when the setup allows it.
This method captures all return airflow, including any duct leakage from unconditioned spaces. It does not distinguish between conditioned return air and leaked air. For duct leakage analysis, combine it with individual grille measurements (see M3).
Setup:
airflow_source = "trueflow".When the return trunk is accessible and has a section of straight ductwork, a Pitot traverse (M4) can measure return airflow. This is practical for homes with exposed return ductwork in basements or utility rooms.
The traverse procedure is identical to a supply-side traverse. The challenge is finding enough straight duct. Return trunks often transition from duct to plenum within a few feet of the air handler. If you can get at least 3-4 duct diameters of straight run, a traverse is viable.
For low-velocity return ducts (below 500 FPM), use a hot-wire anemometer instead of a Pitot tube for better resolution.
Every air filter has a pressure-flow relationship: as airflow increases, the pressure drop across the filter increases. If you know the filter's pressure-drop curve and measure the actual pressure drop, you can estimate the airflow.
This is not as accurate as a TrueFlow reading, but it provides a useful cross-check or a reasonable estimate when direct measurement is not practical.
Steps:
Limitations: The curve is for a clean filter. A dirty filter has a higher pressure drop for the same airflow, so a dirty filter will underestimate the actual airflow. Filter curves also assume uniform air distribution across the filter face, which may not hold if ducts enter the plenum asymmetrically.
If the filter is at the return grille rather than at the air handler, use a capture hood (CPS EasyHood or Testo 420) placed over the grille to measure the airflow entering the return system from the conditioned space. For multiple return grilles, measure each one and sum the readings.
This method measures only the air from the conditioned space. It does not capture duct leakage. That is useful if your goal is to compare conditioned-space return airflow to air-handler airflow for duct leakage estimation.
Most residential systems have more than one return grille. Measuring total return airflow requires accounting for all return paths.
Newer construction often includes a return grille in each bedroom and main living area. To measure total return airflow:
On a system with 4-6 returns, this takes 20-30 minutes with a capture hood.
Older homes and many tract-built homes have one or two large return grilles, often in a central hallway. These grilles are large enough that a single capture hood may not cover the entire opening.
Options:
A platform return uses the space beneath the air handler as the return plenum. Air enters through openings in the platform sides or bottom. These returns are common in closet or utility-room installations.
Measuring airflow in a platform return is difficult because the "plenum" is not a defined duct with measurable dimensions. The best approach is to place the TrueFlow Grid at the filter slot at the air handler inlet. This captures all air entering the equipment regardless of how it got into the platform space.
Technician with TrueFlow Grid installed at the air handler filter slot, using a smartphone to operate the TEC app during airflow measurement
measureQuick uses airflow data in several diagnostic calculations:
The source of the airflow data matters. A TrueFlow measurement (airflow_source = "trueflow") is weighted as a direct measurement in diagnostics. A manual entry (airflow_source = "manual") or an estimated value (airflow_source = "estimated") carries wider uncertainty, which measureQuick accounts for in its confidence assessment.
Turbulence in the return plenum causes pressure fluctuations. Wait 5-10 seconds for the reading to settle before recording. If fluctuations persist, check for doors opening and closing in the conditioned space, wind effects on the building, or a blower motor with speed cycling (variable-speed motors adjusting to load).
The most common reason is a dirty filter. A filter loaded with debris has a higher pressure drop for the same airflow. If the filter is visibly dirty, replace it and re-measure, or note the discrepancy in your documentation. Damaged filters (bent frames, torn media) can also produce readings that deviate from published curves.
A difference of 5% or less is normal and can be attributed to measurement tolerance and density variation. A difference of 10% or more suggests return duct leakage between the grilles and the air handler. See M3 for duct leakage analysis techniques.
Some platform returns pull air through gaps in the platform framing rather than through defined grilles. In this case, you cannot measure the airflow entering the platform from the conditioned space. Measure at the air handler filter slot with TrueFlow for total return airflow and use the static pressure screening to assess duct leakage.
Prerequisites (complete these first):
Follow-up articles (next steps after this one):
Related in other domains:
If you get stuck or this article does not answer your question: