Velocity Pressure and Pitot Tube Traverse
Anemometers guess. A pitot traverse measures. Learn how velocity pressure works, how to run a proper equal-area grid, and how to turn a column of manometer readings into a defensible CFM number.
Skip the arithmetic
Drop your average velocity and duct size into the airflow tool and get CFM instantly.
In This Guide
Total, Static, and Velocity Pressure
Every duct carries three related pressures, and understanding how they connect is what makes a pitot tube work. Static pressure (SP) is the pressure the air pushes outward against the duct walls in every direction. Velocity pressure (VP) is the pressure created purely by the air's motion, felt only when you face into the stream. Total pressure (TP) is the sum of the two.
The Relationship
Rearranged: VP = TP − SP. A pitot tube measures both TP and SP at the same point, and a manometer connected across the two ports does the subtraction for you.
A pitot-static tube is really two tubes in one. The impact (total pressure) port is the hole at the tip, pointed directly upstream, so it catches both the static push and the ram of moving air. The static ports are the small holes around the side of the barrel, which sense only static pressure. Connect the tip port to the positive (+) side of a digital manometer and the side ports to the negative (−) side, and the gauge reads velocity pressure directly in inches of water column (iwc). No math on the pressures themselves is needed.
The 4005 Velocity Pressure Formula
Velocity pressure is not velocity. To turn a manometer reading into an air speed in feet per minute (FPM), you use the standard-air velocity pressure equation:
Velocity from Velocity Pressure
- V = air velocity (FPM)
- VP = velocity pressure (inches water column)
- 4005 = constant for standard air (0.075 lb/ft³, 70°F, sea level)
The constant 4005 bakes in standard air density. At high altitude, in a hot furnace supply, or in a cold outdoor run, air density shifts and the true constant changes. For most residential and light commercial work at reasonable temperatures, 4005 is close enough. When you are more than a few thousand feet up or measuring air well above 100°F, apply a density correction to stay honest.
Once you have an average velocity, airflow is simple geometry:
Velocity to Airflow
Area is the duct cross-section in square feet. If you measured the duct in inches, divide the square-inch area by 144 first, or use CFM = (V × Area in²) / 144.
Why a Single Reading Lies
Air does not move at one uniform speed across a duct. Friction drags the air near the walls almost to a stop, while the core in the center races ahead. Poke a probe into the middle of a duct and you read the fastest air in the whole cross-section, which can overstate the true average by 20 to 40 percent. That error flows straight into your CFM number and, from there, into your tonnage-per-CFM and delta-T conclusions.
A traverse solves this by sampling many points spread across equal areas of the duct, then averaging them. Done right, it is the most accurate field method for measuring airflow and is the reference that flow hoods and inline sensors get checked against.
Average velocities, not pressures
Because VP grows with the square of velocity, averaging the raw VP readings and then taking one square root overstates the answer. Always convert each VP to FPM first, then average the FPM values. This is the single most common traverse mistake.
Laying Out the Traverse Grid
The goal is one reading per equal-area zone. Two grid styles are standard, and the log-Tchebycheff spacing is preferred because it weights points to better represent the wall region.
Rectangular Duct
Divide the face into a grid of equal-area rectangles and read at the center of each. Use a minimum of 16 points; large ducts warrant 25. Never fewer than 4 points per side.
Round Duct
Traverse on two perpendicular diameters, 6 to 10 points on each, placed at the centroids of equal-area concentric rings. That is 12 to 20 readings total.
Location matters as much as point count. Turbulence from fittings ruins a traverse, so measure in a straight section: at least 7.5 duct diameters downstream and 2.5 diameters upstream of any elbow, tee, transition, or damper. For rectangular duct, use the equivalent diameter. When you cannot find that much straight duct, more downstream distance is worth more than upstream distance, and you should note the compromise on your report.
Ventora — Your AI HVAC Companion
Velocity pressure conversions, duct math, and instant airflow answers in your pocket. Free on iOS.
Step-by-Step Traverse Procedure
- Find a clean section. Confirm the 7.5/2.5 diameter rule, or get as close as the jobsite allows, and pick a spot you can safely reach with the blower running.
- Mark and drill the access. Lay out insertion points on tape along the duct. Drill a 5/16" test port on each traverse line and plan to seal them afterward.
- Zero and connect the manometer. Total-pressure port to (+), static ports to (−). Zero the gauge with the ports open to the room so drift does not corrupt low readings.
- Face the flow. Insert the pitot with the tip pointed squarely upstream, parallel to the duct axis. Even a 10° yaw drops the reading; keep the tip aligned.
- Read every point. Let the value settle, record VP at each depth, and move to the next equal-area point. Do not skip the wall points just because they read low.
- Convert, then average. Apply V = 4005 × √VP to each reading, average the velocities, and multiply by duct area for CFM.
- Seal the ports. Plug the test holes so you do not leave new leaks in the system.
Worked Example: Rectangular Duct
Problem
A 20" × 12" supply trunk feeds a 3-ton system that should deliver about 1,200 CFM. You run a 16-point traverse. To keep the math readable, we'll average the 16 velocities into four representative VP readings below.
| Reading | VP (iwc) | √VP | Velocity = 4005 × √VP (FPM) |
|---|---|---|---|
| Core center | 0.090 | 0.300 | 1202 |
| Mid zone | 0.078 | 0.279 | 1119 |
| Near wall | 0.052 | 0.228 | 913 |
| Corner | 0.034 | 0.184 | 738 |
Step 1: Convert each VP to velocity (done in the table).
Note how the corner air moves at barely 60% of the core speed.
Step 2: Average the velocities
(1202 + 1119 + 913 + 738) / 4 = 993 FPM
Step 3: Find the duct area in square feet
(20 × 12) / 144 = 240 / 144 = 1.667 ft²
Step 4: Calculate CFM
993 FPM × 1.667 ft² = 1,655 CFM
Watch the shortcut error
If you had averaged the four VP values first (0.0635 iwc) and taken one square root, you would get 4005 × √0.0635 = 1,009 FPM and 1,682 CFM — inflated versus the correct 993 FPM / 1,655 CFM. The gap widens the more uneven the profile is.
Interpreting the result
1,655 CFM against a 1,200 CFM target (400 CFM/ton) means this trunk is moving well over design flow — worth checking the fan tap, damper positions, and whether this trunk is starving other runs. Airflow that is too high can hurt dehumidification just as low airflow hurts capacity.
Common Mistakes and Field Tips
- Averaging VP instead of velocity. Convert every reading first, then average the FPM.
- Measuring near a fitting. Elbows and dampers spin the air; honor the 7.5/2.5 diameter rule or note the exception.
- Skipping wall points. The slow air near the walls is exactly what pulls the true average down. Read it.
- Yawed pitot tip. Keep the tip parallel to the duct axis and pointed dead upstream; a cocked probe reads low.
- Reversed manometer hoses. Total to (+), static to (−). Swapped hoses read negative or garbage.
- Ignoring density. High altitude or hot/cold air shifts the 4005 constant; correct for it when conditions are extreme.
- Low VP, high noise. Below about 0.02 iwc the signal is small and jittery; a hot-wire anemometer or flow grid may be more repeatable at very low velocities.
Field tip: sanity-check against tonnage
Before you trust a traverse, compare it to the 400 CFM-per-ton rule of thumb. A 3-ton system should land near 1,200 CFM. If your number is wildly off, recheck your area math, your grid, and your probe orientation before blaming the equipment.
The AI HVAC Assistant in Your Pocket
Ventora runs your velocity pressure, duct, and load math, answers code questions, and troubleshoots in the field so you can focus on the work. 7-day free trial, then $19.99/month.