A duct traverse measures air velocity at multiple points across a duct cross-section, then averages those readings to determine the mean velocity through the duct. Multiplying the mean velocity by the duct's cross-sectional area gives the airflow volume in CFM.
This is a direct measurement method. Unlike the enthalpy method or fan-curve matching, a traverse does not depend on equipment capacity assumptions or refrigerant conditions. It measures the actual speed of the air moving through the duct at the point where you take the readings.
A duct traverse is the right choice when:
A traverse is not the fastest method. TrueFlow or a capture hood will give you a result in a few minutes. A traverse takes 30-60 minutes depending on duct size and the number of measurement points. Use it when the other methods are not practical or when you need the highest confidence in your reading.
A Pitot tube is an L-shaped metal tube that measures the difference between total pressure and static pressure at a single point in the duct. That difference is velocity pressure, which converts directly to velocity. The Pitot tube connects to a manometer (such as the DG-1000) that reads the velocity pressure.
Pitot tubes have no moving parts, no batteries, and no calibration drift. They are accurate across a wide velocity range (500-8,000 FPM). Below 500 FPM, the velocity pressure signal is very small and difficult to read reliably.
A hot-wire (thermal) anemometer measures air velocity directly by sensing how much heat the moving air strips from a heated wire element. It reads out in FPM on a digital display.
Hot-wire anemometers are better at low velocities (below 500 FPM) where a Pitot tube struggles. They are also easier to read since they display velocity directly rather than requiring a pressure-to-velocity conversion. The trade-off is that they require periodic calibration and can be affected by temperature extremes.
For most residential HVAC duct traverses, either instrument works. If you already own a DG-1000, a Pitot tube is an inexpensive addition.
The standard traverse pattern is the equal-area method. You divide the duct cross-section into equal-area sections and take one velocity reading at the center of each section.
For round ducts, the standard approach is to measure along two perpendicular diameters (a cross pattern). ASHRAE recommends a minimum of 5 points per diameter for ducts up to 24 inches, and more for larger ducts.
For a typical residential round duct (6-14 inches):
The measurement points are not evenly spaced. They follow a log-linear or log-Tchebycheff distribution that places more points near the duct wall where velocity changes rapidly due to friction. ASHRAE Standard 111 provides the exact insertion depths as a fraction of the duct radius.
For a 5-point traverse on a 12-inch round duct, the insertion depths from the near wall are approximately: 0.6", 2.0", 3.0", 4.0", 5.4" (measured from the duct wall to the center of the Pitot tube tip).
For rectangular ducts, divide the cross-section into a grid of equal-area rectangles. The minimum number of points depends on the duct dimensions:
| Duct Size | Minimum Grid |
|---|---|
| Up to 12" x 12" | 3 x 3 (9 points) |
| 12" x 24" | 3 x 4 (12 points) |
| 24" x 24" or larger | 4 x 4 (16 points) |
Mark the measurement positions on the outside of the duct. Drill access holes along one or two sides. Insert the instrument to each grid center point and record the velocity.
For a 12" x 18" rectangular duct with a 3 x 3 grid, the nine measurement points are at the centers of nine 4" x 6" rectangles.
[Diagram] 3x3 traverse grid on a rectangular duct cross-section. Each cell represents an equal-area rectangle, and the measurement point is at the center of each cell.
| | Column 1 | Column 2 | Column 3 | |---|---|---|---| | Row 1 | Point 1 (center) | Point 2 (center) | Point 3 (center) | | Row 2 | Point 4 (center) | Point 5 (center) | Point 6 (center) | | Row 3 | Point 7 (center) | Point 8 (center) | Point 9 (center) |
For a 12" x 18" duct, each cell is 4" x 6". Insert the pitot tube or hot-wire anemometer to the center depth of each cell and record the velocity reading. Average all 9 readings to calculate mean duct velocity.
Add all velocity readings and divide by the number of readings. If one reading is dramatically different from the others (more than 50% above or below the average), check for obstructions, holes, or damaged ductwork at that point.
Example for a 9-point rectangular traverse:
| Point | Velocity (FPM) |
|---|---|
| 1 | 620 |
| 2 | 710 |
| 3 | 580 |
| 4 | 680 |
| 5 | 740 |
| 6 | 650 |
| 7 | 590 |
| 8 | 720 |
| 9 | 610 |
Average velocity = (620 + 710 + 580 + 680 + 740 + 650 + 590 + 720 + 610) / 9 = 656 FPM
For a 12" x 18" rectangular duct: Area = 12 x 18 / 144 = 1.50 sq ft
CFM = Average Velocity x Area
CFM = 656 x 1.50 = 984 CFM
If you are using a Pitot tube with a manometer, the manometer reads velocity pressure in inches of water column (IWC). Convert to velocity using:
Velocity (FPM) = 4,005 x sqrt(velocity pressure in IWC)
For a velocity pressure of 0.03 IWC: Velocity = 4,005 x sqrt(0.03) = 4,005 x 0.173 = 693 FPM
This formula assumes standard air density (0.075 lb/ft3, approximately 70F at sea level). At higher altitudes or temperatures, apply a density correction.
Turbulence from elbows, transitions, dampers, and takeoffs disrupts the velocity profile. ASHRAE Standard 111 recommends placing the traverse plane at least 7.5 duct diameters downstream of a disturbance and 3 duct diameters upstream of the next one. In residential ductwork, this ideal spacing is often impossible. Do the best you can; even 3-4 diameters downstream of an elbow significantly improves accuracy.
If you cannot find a location with adequate straight duct, increase the number of traverse points to compensate. More points average out the turbulence better. For a rectangular duct with a nearby elbow, move from a 3x3 grid to a 4x4 grid.
measureQuick accepts manual airflow entry. After completing the traverse calculation:
airflow_source field records this as "manual" since the value was not transferred automatically from TrueFlow.A manual entry does not carry the same automated validation as a TrueFlow reading, but it gives measureQuick the airflow data it needs for superheat, subcooling, and capacity calculations.
| Method | Accuracy | Speed | Best Use |
|---|---|---|---|
| Duct traverse (Pitot) | +/-3-5% when done correctly | 30-60 min | Ducts without filter slots; branch duct measurement |
| TrueFlow Grid | +/-5% | 5-10 min | Systems with accessible filter slot |
| Capture hood | +/-5-10% | 5-15 min per register | Individual register measurement |
| Temperature-split estimate | +/-15-25% | Automatic | Fallback when no direct measurement is available |
A properly executed traverse is among the most accurate field methods. The key qualifier is "properly executed." Sloppy point placement, too few points, or measuring too close to a disturbance can introduce 10-20% error.
These standards are written for commercial balancing work, but the measurement principles apply identically to residential ductwork. The primary difference is scale: residential ducts are smaller, which means fewer traverse points are needed but achieving adequate straight-duct distance is harder.
Some variation is normal. Velocity near duct walls is lower than at the center due to friction. If a reading is more than double or less than half the average, check that location for obstructions (dampers, turning vanes, takeoffs) or duct damage. Do not discard outliers without investigating; they may indicate a real duct problem.
At velocities below 500 FPM, Pitot tube velocity pressure is less than 0.016 IWC, which is at or below the resolution of most field manometers. Switch to a hot-wire anemometer for low-velocity ducts, or use a more sensitive manometer.
For residential work, 5 points per diameter on round duct and 9-16 points on rectangular duct provide adequate accuracy. Adding more points improves accuracy but with diminishing returns. Going from 9 to 16 points on a rectangular duct typically changes the average by less than 3%.
Move as far downstream as possible. If you cannot achieve 7.5 diameters, take extra points (increase from 9 to 16 on rectangular, or from 5 to 8 per diameter on round). Note the proximity to the disturbance in your documentation.
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: