What Back Pressure Does at the Base of a Stack

Why this matters

Every drainage fault a customer describes sounds like every other one: a smell, a noise, a slow fixture. Back pressure is the one that leaves a mark, because it is the only fault inside the drainage piping that pushes outward. A negative event pulls air in through a seal and makes a noise. A positive event lifts water out of a seal and puts it on the floor, in the room, on the finished side of the barrier. Ground-floor fixtures in tall buildings are where this lives, and the reason techs misdiagnose it is that they arrive with a mental model built on gurgling upstairs fixtures and try to apply it at the other end of the pressure gradient.

The sign convention, and why it is the fastest diagnostic you have

A drainage stack is not at one pressure. Water falling as a film on the wall drags air down with it, which pulls the upper stack below atmospheric. That same air has to escape at the bottom, where the water piles into the horizontal and fills the bore, which pushes the lower stack above atmospheric. Negative on top, positive at the bottom, in the same pipe, at the same instant.

That gives you a discriminator you can collect from the doorway before opening a tool bag:

Direction is observed, not inferred. If a report says gurgle and there is standing sewer water on the floor, one of the two is wrong, and the water is the honest witness.

Where the positive zone comes from

Follow the air. It came in at the terminal, rode down inside the stack with the film, and now has to get out. At the base, the water is no longer a film with a hole in the middle. It decelerates, piles up, and fills the bore for some distance into the building drain. The air arriving behind it meets a moving plug of water.

The air is not compressible enough to be politely absorbed, so its pressure rises until it can force its way through the plug or find an alternative opening. Every trap connected to the stack below the point where the sign flips, and every trap connected to the building drain inside the plugged zone, is an alternative opening.

The air-circuit article covers the supply half of that loop. What the base adds is that the water itself is the restriction, so this is the one place in the system where the water problem and the air problem stop being separable.

What decides its size

Three things, and only one of them is under the designer's control.

How much water is falling at once. The pressure rises with the flow rate arriving at the base, which is why this is a morning-peak fault and why a stack that is fine all day fails between six and eight.

How completely the base plugs. A gentle transition holds less water in the bore than an abrupt one, and an offset partway down the stack makes a second, higher plug that does the same thing in the middle of the building.

Whether there is a relief path. A vertical air path tied near the base gives the air somewhere to go that is not a trap. Whether the building has one is the single largest factor, and the vent takeoff article covers how to tell by walking the risers.

What the seals can take, and how to measure against it

The trade designs to a system pressure criterion: keep pressure inside the drainage piping within about plus or minus 1 inch of water column, stated against a nominal two-inch trap seal. That criterion and its conditions belong to the trap-seal article; the two numbers you need here are the nominal 2.0 inch depth and the plus or minus 1.0 inch criterion, and both get printed whenever a measured pressure is compared to a trap.

Three measurement rules decide whether your number means anything.

Capture the peak, not the average. A trap responds to a transient lasting a second or two. A damped or averaging gauge reports the mean of that transient, which is always lower. If a damped gauge is all you have, the reading is a lower bound on the peak and must be written with one inequality and no plus or minus, because you do not know how much of the peak the damping removed.

Carry the instrument's bound when comparing against an absolute limit. A manometer specified at plus or minus 1 percent of full scale on a 10 inch water column range carries a fixed bound of 0.10 inches of water column. The basis is full scale, so it does not shrink as the reading shrinks, and its character is a systematic offset, the same sign on every reading that instrument takes. Comparing one reading against the 1.0 inch criterion is an absolute comparison, so the offset does NOT cancel and both ends of the bound get checked. It only cancels when you subtract two readings from the same instrument.

Add contributions that push the same way, as bounds. Two independent peaks acting on the same seal in the same direction are worst-case bounds, and worst-case bounds add linearly and are reported as bounds.

The gate: which zone is this fixture connected to

One question decides everything about a fixture's symptom, its risk, and its fix.

Is this trap connected where the sign is negative, or where it is positive? In practice that means: is it on a branch high enough in the stack that the falling film is stealing air past it, or is it connected low enough that it sees the plug at the base, either on the stack itself or on the building drain inside the plugged run?

Run the gate on two fixtures in the same building and you get opposite answers, opposite symptoms, and opposite work.

Worked example: one stack, two fixtures, opposite answers

Five-storey residential building, four-inch soil stack, complaints from the ground floor only, worst on weekday mornings.

Setting up the taps. Two pressure taps, one at the fifth-floor branch cleanout and one at the ground-floor branch cleanout, both above the flood level of the fixtures they serve. Fit a threaded, gasketed test adapter rather than cracking a plug and holding a tube in it: this test deliberately drives the line positive, and a loose plug at the ground-floor tap will vent sewer water at whoever is holding it. Splash goggles and gloves on for anyone at a tap, everybody standing to the side of the opening rather than in front of it, and the tap fitted and sealed before the first discharge is called for. Do not tap below the flood rim of a fixture on that branch, and do not enter a manhole or wet well to reach a better tap point: those are permit-required confined spaces under 29 CFR 1910.146 in general industry and 29 CFR 1926 Subpart AA in construction, requiring a permit, atmospheric testing, ventilation, an attendant and a rescue plan, and nobody follows anybody in.

Instrument. Peak-capturing digital manometer, 10 inch water column range, plus or minus 1 percent of full scale, so a fixed bound of 0.10 inches of water column, systematic, on every reading today.

The event. Four upper-floor water closets discharged simultaneously.

Ground-floor tap. Peak reads plus 1.4 inches of water column.

• Carry the fixed bound, because this is an absolute comparison: the true peak lies between 1.3 and 1.5 inches of water column.
• Against the criterion of 1.0: both ends exceed it. The conclusion survives the instrument bound, which is the only reason it is worth writing down.
• Had this come from a damped gauge instead, the honest statement would be peak >= 1.4 inches of water column, with no interval.

Fifth-floor tap, same event. Peak reads minus 0.6 inches of water column.

• With the same fixed bound carried, the true value lies between 0.5 and 0.7 inches of water column negative.
• Against the criterion of 1.0: both ends are inside it. This branch is in the negative zone and is not currently at risk, though it has less margin than the number suggests once anything else is running.

Adding the second contribution as bounds. The building's real morning peak is not four closets. The ground-floor laundry discharges into the same stack, measured separately at the same tap as a plus 0.6 inch water column peak. Both push the seal the same direction, toward the room, so as worst-case bounds they add linearly: combined peak <= 2.0 inches of water column. One inequality, no plus or minus.

What that lands on. The ground-floor trap's nominal depth is 2.0 inches. A combined event bounded at 2.0 inches of water column is not a burp, it is the whole seal, which is why the complaint on Monday mornings is a wet floor and an open pipe rather than a noise. The four-closet test alone, at 1.4, explains the ordinary weekday burp. Two different events, two different severities, one mechanism.

Which fixture gets the work. The ground floor. Note what that means for the usual instinct: the fifth-floor branch has the noisier symptom history and is the one people ask about, and it measured inside the criterion. Chasing the gurgle upstairs would have consumed the visit.

The failure mode. The single most common wrong move on this fault is to add an air admittance valve at the ground-floor fixture. An AAV is a one-way device that admits air and is designed not to pass it back out. On a negative event it works. On this event it is closed, by design, at exactly the moment relief is needed, and the customer has paid for a part that cannot act. The base-of-stack article covers the rest of the non-fixes.

How to verify the reading rather than trusting it

Repeat the event three times and take the highest peak, not the average of three. The seal fails on the worst one. If your three peaks spread more than the instrument's own bound, the spread is real variation in the discharge, and the top of that spread is the number that matters.

Prove the sign with water, not with the display. Refill the ground-floor trap, mark the level, run the event, and look. Positive means water displaced outward: splash outside the trap, a wet ring, a level that came back low after being thrown rather than drawn. If the level was drawn down instead, your tap is telling you positive and your fixture is telling you negative, and one of them is on a different branch than you think.

Check the tap is on the branch you believe. Buildings routinely have a ground-floor drain tied somewhere the drawings do not show. Run water into the fixture and confirm you see it at the tap before you trust any pressure from that tap.

Re-test after the repair at the same taps with the same instrument. Because the instrument's offset is systematic and identical on both visits, the before-and-after difference cancels it, and that difference is a far stronger number than either absolute reading. It is the one comparison in this whole procedure where the instrument bound stops mattering.

References

• The plumbing code as adopted and amended by your local jurisdiction (the IPC or UPC family), for stack venting, relief venting and the connection rules at the base of a stack; it binds through the adopting ordinance and the permit rather than on its own
• 29 CFR 1910.146 (general industry) and 29 CFR 1926 Subpart AA (construction), permit-required confined spaces, for any sewer manhole, wet well or interceptor
• Manufacturer documentation for the manometer in use, for the accuracy basis (percent of full scale against percent of reading) and whether it captures peaks
• See related: What a Drainage System Is Actually Doing With Air; Why the Bottom of a Stack Is the Worst Place in the System; Why a Trap Seal Is the Only Barrier and What It Is Made Of