Static, Velocity and Total Pressure

Why this matters

A duct system gives you three pressures and most techs only ever read one of them. That is usually fine, right up to the visit where static reads normal and the rooms are still starved, or where static goes up across a fitting and someone writes "no loss here" on the report. Static pressure can legitimately increase in the direction of flow. Total pressure never can. Knowing which of the three you are holding is the difference between a reading that constrains the diagnosis and a reading that quietly rules out the actual fault.

Before you drill a test port

• Lock the blower's disconnect open and confirm the wheel has stopped before any part of you goes past the cabinet opening. A wheel coasting down looks stopped from three feet away. The mechanical isolation duty here is 29 CFR 1910.147.
• A drill bit does not know what is behind the sheet metal. Refrigerant lines, low-voltage bundles, line-voltage whips and gas piping all run along duct exteriors. Look on both sides where you can, feel for tubing, and drill with a stop collar set to just past the metal thickness.
• Cut sheet metal leaves an edge that will open a forearm. Deburr the hole, plug it when you are done, gloves on for the whole operation.
• If the cabinet interior is lined with fibrous insulation, drilling and probing releases airborne fibers. Work upstream of your own breathing zone with the blower locked off, and wear a fitted particulate respirator selected under 29 CFR 1910.134 rather than relying on a dust mask you did not fit test.

Three numbers, one equation

• Static pressure (SP) is the pressure the air exerts on the duct wall in every direction. It is what pushes air through everything downstream. Read it through a hole in the duct wall, perpendicular to the flow.
• Velocity pressure (VP) is the additional pressure you feel only when you face into the moving stream. It is a measure of the air's motion and it is always positive in the direction of flow.
• Total pressure (TP) is the sum: TP = SP + VP. It is the whole energy content of the air at that point.

For standard air (0.075 lb per cubic foot, roughly sea level at 70 F, dry), velocity comes out of velocity pressure as:

V (fpm) = 4005 x square root of VP (in. w.c.)

The 4005 is not a law of nature. It is the density-general form V = 1096 x square root of (VP / density) with 0.075 substituted in.

What static pressure is blind to

How much air is moving. Static is a pressure, not a flow. A system reading a healthy total external static can be moving badly short of design air, and a system reading low static can be worse off than one reading high.

The trap that catches people is the low reading. High static gets investigated because everyone was taught that high static means restriction. Low static gets a shrug. But a return that has come apart in an attic, a disconnected branch, or a filter rack drawing half its air from the mechanical room all reduce system static while destroying delivery to the rooms. Static went down because resistance went down, and resistance went down because the air found a shorter path that does not go through the house.

Where the resistance is. A single static reading is a sum of everything between that point and the reference. It cannot apportion.

What velocity pressure is blind to

Every part of the duct that is not directly at the tip of your probe. Velocity across a duct cross section is never uniform, so one VP reading is one filament of air. That is the entire reason for a traverse: a grid of readings averaged as velocities, not as pressures, because the square root is not linear.

Air that is moving slowly. Because VP goes as the square of velocity, it collapses at low speed and takes your resolution with it. Work it through on a manometer that resolves 0.001 in. w.c.:

• At 864 fpm, VP is 0.047 in. w.c. A 0.001 uncertainty is 2.1 percent of the reading, and because velocity goes as the square root, that is about 1 percent on velocity.
• At 200 fpm, VP is 0.0025 in. w.c. The same 0.001 uncertainty is 40 percent of the reading, which is about 20 percent on velocity.

Same instrument, same technician, same afternoon. Below roughly 600 fpm a pitot traverse stops being a measurement and starts being an opinion, which is why low-velocity returns and large plenums are flow-hood or tracer work rather than pitot work.

What total pressure is blind to

Total pressure is the honest one, and it is worth being precise about why. Total pressure always falls in the direction of flow. Every fitting, every foot of duct, every filter takes some and gives none back. If your downstream total is higher than your upstream total, there is a fan between them or you have made a measurement error.

What it will not do:

• It will not tell you where the loss occurred between two taps, only that it did. Locating it takes a series of taps, the same way a hydronic restriction hunt does.
• It will not tell you the air is going where you want it. A perfectly efficient duct that dumps into the wrong room has excellent total pressure behaviour and a failed job.
• It is fussy to read directly. The impact port has to face into the stream. A probe held a few degrees off in a swirling section after an elbow reads low, and you will not see that in the number.

The sign convention that gets people

On the suction side of a fan, static pressure is below atmospheric, so it reads negative. Velocity pressure is still positive, because the air is still moving. Total pressure on the suction side is therefore usually negative as well.

That is why total external static is taken as the sum of the magnitudes of the supply-side and return-side static readings rather than their difference: you are adding the push on one side to the pull on the other. The full procedure and the range comparison against equipment maximums belong to the static-pressure measurement card in this library; this article is about which of the three numbers you are holding.

One convention worth knowing because it appears on every fan curve: fan static pressure is fan total pressure minus the velocity pressure at the fan outlet. A fan curve plotted against static pressure and one plotted against total pressure are not interchangeable, and confusing them puts your operating point in the wrong place.

Worked: the fitting where static went up and the fan still paid

A 10 in. by 20 in. supply trunk, 200 square inches, which is 1.389 square feet. It carries 1200 CFM, then transitions to 20 in. by 20 in., 400 square inches, 2.778 square feet.

Velocities:

• Upstream: 1200 / 1.389 = 864 fpm
• Downstream: 1200 / 2.778 = 432 fpm

Velocity pressures, from VP = (V / 4005) squared:

• Upstream: (864 / 4005) squared = 0.047 in. w.c.
• Downstream: (432 / 4005) squared = 0.012 in. w.c.

Measured static upstream is 0.420 in. w.c., so upstream total is 0.420 + 0.047 = 0.467 in. w.c.

Say the transition itself costs 0.020 in. w.c. of total pressure, which is a loss the fan has to pay for. Downstream total is therefore 0.467 - 0.020 = 0.447 in. w.c. Downstream static is total minus the downstream velocity pressure: 0.447 - 0.012 = 0.435 in. w.c.

Read that back. Static rose from 0.420 to 0.435, an increase of 0.015 in. w.c., across a fitting that destroyed 0.020 in. w.c. of total pressure. The air slowed down, and 0.035 in. w.c. of velocity pressure converted back into static. The fitting handed back more static than it consumed, and kept the difference. That conversion is real and it is called static regain; system designers use it deliberately in trunk sizing.

A tech reading static only, on both sides of that transition, records a gain and moves on. A tech who also read velocity pressure at both taps has the loss, and can compare 0.020 in. w.c. against what the fitting should cost.

The constant assumes 70 F air at sea level, and your duct may be neither

The 4005 is standard-density air. Two field conditions move it enough to matter, and they move it in opposite directions:

• Hot supply air. Density scales inversely with absolute temperature at a given pressure. Air at 150 F is 530 / 610 = 0.869 times as dense as 70 F air, so about 0.065 lb per cubic foot. For a given velocity pressure the true velocity is 1 / square root of 0.869 = 1.073, about 7 percent higher than the 4005 constant reports.
• Altitude. Thinner air, same effect, larger at elevation. Use the density-general form and get your density from the equipment manufacturer's altitude correction table rather than guessing it.

Both corrections work the same way: put the actual density into V = 1096 x square root of (VP / density). Note the direction carefully. Low density means more velocity for the same velocity pressure, but less mass of air, and it is mass that carries heat. This is exactly why a furnace at elevation is derated even though the duct velocities look fine.

Confirming a traverse before you trust it

1. Average velocities, not velocity pressures. Convert each VP to a velocity, then average the velocities. Averaging the pressures first and taking one square root understates the answer, and understates it more the more uneven the profile is.
2. Check your traverse plane. You want a straight run of at least several duct diameters upstream of the plane. If you cannot get it, say so on the report rather than quoting the number to three digits.
3. Cross-check against the fan. Take the fan's static and its speed, read the flow off the published fan curve, and compare. Two independent estimates that agree within about 10 percent are a result. Two that disagree by half mean one of them is measuring something you did not intend, and the traverse plane is the usual culprit.
4. Repeat one grid point at the end. If the same hole gives a materially different VP than it did at the start, the system changed state mid-traverse and the whole grid is a mixture of two operating points.

References

• ASHRAE Handbook, Fundamentals, duct design chapter, for the static, velocity and total pressure relationship and static regain
• AMCA standards for fan performance definitions, including fan static pressure as fan total pressure minus outlet velocity pressure
• OSHA 29 CFR 1910.147, hazardous energy control for locking out a blower before reaching into the cabinet
• OSHA 29 CFR 1910.134, respiratory protection program requirements where fibrous duct liner is disturbed
• See related: HVAC Static Pressure Measurement Reference (total external static procedure and ranges); Airflow Measurement and Balancing Reference