What a Vacuum Breaker Is For

Why this matters

A vacuum breaker is a small check valve that admits air into a steam space, and it is the cheapest fitting on the equipment it protects. It gets left off drawings, painted shut, plugged during a repair because it was "leaking air," or deleted when somebody decides the coil already has a vent. The equipment then runs correctly every single day and holds a pool of condensate every single night, and the bill arrives years later as a shell that has been sitting in oxygenated water since the last inspection, or as a split tube on the first hard freeze.

The reason nobody catches it is that the fault only exists when the equipment is off. Every reading you can take while it is running is correct.

Before any of this is checked on a live system

Steam at any usable pressure burns severely and the first stretch of a leak out of a fitting is invisible, so a vacuum breaker is never located, tested or confirmed by hand, by feel, or by reaching toward a hiss. Work from beside any opening and keep a body length off the line of discharge.

Fitting a breaker, replacing one, or fitting the compound gauge described below means breaking a joint. Isolate the steam supply and the condensate return, lock and tag the isolation as control of hazardous energy under 29 CFR 1910.147 in general industry, or under the construction energy control requirements of 29 CFR Part 1926 on a construction site, drain to a point piped away from any walkway, and confirm zero on a gauge you have just watched read something else. Then let the metal fall below 120 F, verified with a non-contact reading taken at arm's length, before a wrench goes on it. A return line opened warm flashes at the opening and carries its water out with it.

A trap runs on a differential and nothing else

A steam trap is a valve that distinguishes condensate from steam. It has no pump in it. When it opens, the only thing that moves water through it is the pressure on the inlet side being higher than the pressure on the outlet side, and the outlet side is the sum of three things: the pressure in the return main, the static head of any vertical rise between the trap and that main, and the friction in the return line.

Two conversions carry all of the arithmetic in this article. Atmospheric pressure is 14.7 psia, so a gauge reading of 2 psig is 16.7 psia. And a column of water exerts 1 psi for every 2.31 feet of height at 60 F; hot condensate near 212 F is roughly 4 percent less dense, so one psi supports nearer 2.4 feet of it. The 2.31 figure is the conservative one for a lift requirement, because it makes the pressure you need come out larger, and it is the one this library uses.

Nothing about that changes when the equipment shuts off. The return main does not drop, the riser does not empty, and the trap still needs a differential. What changes is the inlet side.

Where the vacuum comes from

Saturated steam at 15 psig occupies roughly 800 times the volume of the water it condenses into. That ratio is read off a steam table at the pressure you are actually at, and it is why a closed steam space is not a stable container.

Shut the supply valve on a coil or shell that is full of steam. The metal keeps giving up heat to the room or to the air stream, the steam keeps condensing, and 800 volumes become one. There is no source of gas to replace it, because the supply valve is shut and the return side sits at higher pressure than the collapsing space. Do not lean on the trap to check that return: most trap families have no reverse-flow function at all, which is why a check valve after the trap is standard on a lifted or common return. Without one the riser can drain back into the cooling coil on top of the vacuum. The pressure inside falls below atmospheric within seconds, and it keeps falling as the metal cools.

       vacuum breaker
     (admits air on cooldown)
              |
  steam in -> [ coil steam space ] <- metal cools,
              |                        steam collapses
            trap
              |
        vertical rise
              |
  ---- return main (its own pressure) ----

The question that decides whether a breaker belongs

Does this steam space ever sit closed while it is still cooling, with condensate that still has to leave against a downstream pressure?

Where the answer is yes, the space will go sub-atmospheric and the trap will lose its differential exactly when there is water to move, so a breaker belongs on it. That covers anything with an on-off or modulating supply valve: coils, shell-and-tube exchangers, jacketed vessels, storage calorifiers, unit heaters, sterilisers, anything that cycles.

Where the answer is no, a breaker adds an air ingress path and buys nothing. A drip leg on a live header carries header pressure continuously, and header pressure does not collapse because the header is not a closed volume. Fitting a breaker there gives oxygen-bearing air a route into a wet system for no drainage benefit, and it is one more seat to leak steam out of when it wears.

The distinction is not equipment size and it is not pressure. It is whether the space is ever closed and cooling.

What it is not, and what it does not fix

An air vent and a vacuum breaker both handle air and they work in opposite directions: the vent expels air on warm-up so steam has somewhere to go, the breaker admits air on cooldown so condensate has somewhere to go. Both belong on the same coil and neither substitutes for the other. That comparison is worked properly in the sibling article on what an air vent is doing, and it is not re-derived here.

The part that gets missed is sharper: a vacuum breaker restores the inlet side to about atmospheric pressure, and nothing more. It cannot push condensate up a riser into a pressurized return, because atmospheric pressure is the ceiling on what it can give you. If the outlet side of that trap sits above 14.7 psia when the equipment is off, the breaker does not drain the equipment. It is necessary and it is not sufficient, and the worked example below is that case.

The related condition where a device fails at light load with the supply valve throttled, rather than at shutdown, is stall, and it is diagnosed by a different comparison. The sibling procedure on recognising stall owns it.

Worked example: the preheat coil that is soft on the first call and fine by nine

A steam preheat coil in an air handler, fed at 15 psig, so about 250 F saturation. The complaint is that the first call of the morning does not hold the leaving-air setpoint. Air-side rise on that first call came in at 22 F against a logged 48 F once the unit had been running twenty minutes. Nobody had ever caught it after nine in the morning, and two visits had reported no fault found.

The three numbers that resolve it, taken before anything is opened. Return main gauge: 2 psig, which is 16.7 psia. Tape from the trap outlet up to the return main: 8 feet, measured rather than taken off a drawing, because return routing is the thing drawings are most often wrong about. At 2.31 feet per psi that lift is 3.46 psi. So the trap outlet sits at 16.7 plus 3.46, which is 20.2 psia, whenever that riser is full.

Running, everything is fine, and that is the trap. The coil at 15 psig is 29.7 psia against 20.2 psia at the outlet. That is 9.5 psi of drive, the trap discharges normally, and any test run on it at ten in the morning reports a healthy trap. Both earlier visits were correct about the trap.

Shut down, the inlet side falls out from under it. A compound gauge fitted at the coil, read the following morning before the first call, showed 8 inches of mercury vacuum. At 0.491 psi per inch of mercury that is 3.9 psi below atmospheric, so the coil sat at 10.8 psia. Against 20.2 psia at the trap outlet, the differential is 9.4 psi backwards. The trap is holding a pressure difference in the wrong direction all night, which is the job it is built for, and the coil holds every drop of water in it.

What the morning symptom actually is. The first call has to push that standing water out of the tubes before the coil can fill with steam, so the first call is a coil operating at a fraction of its surface. By nine it is clear and the rise is back at 48 F. The symptom is a start-up condition, not a load condition, which is what separates it from stall: stall appears at light load with the unit running, this appears on the first call after a shutdown regardless of load.

Why a breaker on its own does not fix it. Fit a vacuum breaker and the coil comes back to about 14.7 psia on cooldown. The trap outlet is still at 20.2 psia. The coil is 5.5 psi short of moving one drop, so it still stands full overnight. A shop that fits the breaker, invoices it, and leaves would get a call back with the same complaint, and would have been right about the mechanism and wrong about the fix.

What actually clears it. The lift has to come out of the trap's problem. Drop the trap to a vented receiver set below the coil and pump from there, or fit a device that supplies its own motive pressure; the manufacturer owns which of those suits the load. Take the gravity option with a 3 foot fall from the coil to a vented receiver: with the breaker admitting air the coil sits at 14.7 psia, the bottom of that 3 foot leg is 14.7 plus about 1.24, which is 15.9 psia. Note the constant just flipped direction: 2.31 ft per psi was the conservative choice when the column opposed the trap, and it is the optimistic one when the column is doing the pushing, so a fall uses the hot-condensate 2.41, and the receiver is at 14.7 psia. That is about 1.2 psi of gross drive, and the friction in the drop leg and the trap's own minimum operating differential come out of it before the coil drains. On a short fall to a receiver directly below the coil that usually leaves a working margin, but it is a margin to compute rather than assume: take the minimum operating differential from the trap's published data, and lengthen the fall if the leg is long or full of fittings.

And the breaker is still required in that arrangement. Take the same gravity drop with no breaker: the coil at 10.8 psia plus the 1.30 psi of leg gives 12.1 psia against a receiver at 14.7 psia, which is 2.6 psi backwards. The coil holds water again. Both parts are load-bearing, and either one alone leaves the equipment standing full.

The failure mode if it is left. A coil that spends every night full of water and every morning aerated by the air that eventually leaks in somewhere is a corrosion exposure on the tube side, and on a preheat coil facing outdoor air it is a freeze waiting for the right night. The freeze mechanism itself belongs to the sibling article on what a steam coil does and why it freezes.

Checking one that is already fitted

Fit or read a compound gauge on the steam space, and read it cold. A pressure gauge that only reads positive pressure cannot show you this fault at all; it sits at zero for both a healthy coil and a coil pulling 8 inches of mercury. That single instrument choice is why the condition survives on so many sites.

Take the reading during a real shutdown, not by closing the valve at eleven in the morning on hot metal. The vacuum builds as the metal cools, so a reading taken five minutes after shutdown understates it.

Confirm the breaker is admitting air and not steam. A breaker that has failed the other way lets steam out on every cycle rather than air in, which is a continuous loss and a burn hazard at the discharge. Do not put a hand or a face near it to check; look for the plume and for staining below the fitting, from beside it.

Check that the inlet is not plugged or piped somewhere useless. A breaker whose air inlet has been screened, painted, taped, or piped into a duct that is itself under negative pressure is a fitting with no air supply, and it will read as installed on any checklist.

References

• ASME Steam Tables, or an equivalent saturated-steam table, for the specific volume ratio and saturation pairs at the pressure you actually read
• 29 CFR 1910.147 for control of hazardous energy in general industry, the construction energy control requirements of 29 CFR Part 1926 on a construction site, and 29 CFR 1910.146 for below-grade condensate spaces
• Vacuum breaker, trap and coil manufacturer literature, which owns the cracking pressure, the size and the mounting location for a given piece of equipment
• See related: What an Air Vent Is Doing in a Steam System; How to Recognise Stall Before You Replace the Trap; What a Steam Coil Does and Why It Freezes