What an Air Vent Is Doing in a Steam System

Why this matters

An air vent is a small brass fitting on the end of a main or the side of a coil, and it is routinely removed, plugged, or quietly deleted when a component is swapped for a cheaper one. The system then keeps running at the right pressure with the right gauge readings and delivers less heat every morning, which is the hardest kind of fault to sell because nothing is broken. The vent is not a trim item. It is the device that makes space available for steam, and in a closed pipe there is no other way for the gas already in there to leave.

Most of what is useful about a vent is best said by ruling out the three devices it gets confused with, because each of those confusions produces a specific field mistake.

It is not a safety relief valve

A relief valve protects a pressure boundary against overpressure. It is a code device, sized and set to a stamped pressure, and the sibling article in this library on what a safety relief valve on a steam system protects covers it properly. It operates rarely and its operation is an event.

A vent operates on every start-up, passes gas continuously during warm-up, and has nothing to do with pressure limits. Plugging a vent because the system "already has a relief valve" removes a process function and leaves the protective one untouched, and the system will not complain in any way an operator recognises.

It is not a vacuum breaker

This confusion runs in the opposite direction and it is the more expensive one, because the two devices are both about air and they work against each other by design.

A vent expels air, on warm-up, when the space is filling with steam. A vacuum breaker admits air, on cooldown, when the steam in a closed space condenses and the volume collapses. Both belong on the same piece of equipment and neither substitutes for the other.

The reason a vacuum breaker exists is the same volume ratio that makes hammer dangerous: steam condensing into water loses hundreds of times its volume, so a shell or coil that is shut off while full of steam pulls a vacuum as it cools. Under vacuum, condensate cannot drain out against atmospheric pressure at the trap discharge, so the equipment holds a pool of water until it is next called, which is the flooded condition the sibling article on finding water hammer works through. Fitting a vent where a vacuum breaker was needed does not drain the equipment. Fitting a vacuum breaker where a vent was needed leaves the equipment air-bound on warm-up and adds a path for outside air to enter.

It is not a trap, but a trap may contain one

A trap discriminates by phase: it passes liquid and holds back vapour. A vent discriminates by temperature: it passes anything cooler than saturation and closes on anything at saturation. Those are different tests and they catch different things, which is why a system needs both.

The overlap is that a float-and-thermostatic trap contains both in one body, a float for the condensate and a thermostatic element for the air, and that is the main reason it is the default choice on coils, shells and any equipment that has to warm up from cold. A trap without a thermostatic element does the liquid job and none of the gas job. Swapping one for the other is covered in the worked example below, because it is how most vents actually disappear.

What the vent actually discriminates on

The common thermostatic vent is a sealed capsule containing a liquid chosen so that its own boiling point tracks a little below the saturation curve of steam. As the surroundings approach saturation temperature for whatever pressure the system happens to be at, the fill boils, the capsule expands, and the valve shuts. As soon as the surroundings fall a few degrees below saturation, the fill condenses, the capsule contracts, and the valve opens.

The point worth holding is that this makes the device self-adjusting across pressure. It closes at a temperature a few degrees under saturation whether the system is at 2 psig or 100 psig, without anybody setting it, because the fill's curve and steam's curve run roughly parallel. Air is cooler than the steam around it and cool condensate is cooler than saturation, so both get passed, and steam does not.

The alternative is a bimetallic element, which opens and closes on absolute temperature rather than on a curve. It is more tolerant of hammer and of superheat and correspondingly less clever: because it does not track pressure, it needs setting for the operating pressure and it lags, so it blows through longer at start-up and closes later than a balanced-pressure capsule. Which of the two suits an application is the manufacturer's call for that pressure range.

Where the air actually ends up

Here is the direction that gets stated backwards constantly, because people carry the intuition over from hot water systems.

In a hot water system air rises, because air is far less dense than liquid water and it collects at high points. In a steam space, air is the heavier gas. Air's molecular weight is about 29 against water vapour's 18, so at the same temperature and pressure air is roughly 1.6 times as dense as steam and it does not float to the top of a steam-filled shell.

What actually places the air is the advancing steam front, which pushes the gas ahead of it toward the point most remote from the steam inlet, and the cold metal, which condenses steam locally and leaves the air behind at the surface. So a vent belongs at the end of a main, at the point of a coil or shell that is furthest from the steam connection, and at every dead end. On equipment, take the location from the equipment manufacturer rather than from the top of the casing.

Failure in each direction

A vent has exactly two failure modes and they present nothing like each other.

Failed closed is the silent one. Air cannot leave, so the far end of the run or the remote end of the coil never reaches saturation temperature. The gauge still reads full pressure, every trap still works, nothing leaks, and the only symptom is output. The field signature is a large gap between the saturation temperature for the gauge pressure and the actual metal temperature at the remote end. The sibling procedure for warming a line sets the working limit: a remote-end reading that plateaus more than about ten degrees below the saturation temperature for that pressure means air or water is still occupying that space.

Failed open is the loud one. Live steam blows continuously from the vent, which is a burn hazard wherever the discharge lands and a continuous loss the rest of the year. Do not put a hand or a face near a vent that is discharging to find out what it is doing, and do not attempt to plug or cap one on a live system; steam at usable pressure is invisible for the first stretch out of an opening and it cuts before it feels hot. Isolate the section, lock out the isolation under 29 CFR 1910.147, confirm zero pressure on a gauge known to be good, drain, and let the metal fall below 120 F verified by a non-contact reading taken at arm's length before the vent is unscrewed.

There is a third condition that is not a failure of the vent at all: a vent that has been piped away to a safe discharge point through a line that dips, so the dip fills with condensate and the water column blocks the gas. Vent discharge piping is run with a continuous fall away from the vent, and it never contains a low point, because a low point in a vent line is a liquid seal on a gas device.

Worked example: the trap swap that removed a vent nobody knew was there

A steam coil in an air handler, fed at 15 psig so about 250 F saturation, had been served by a float-and-thermostatic trap for years. The trap failed, and it was replaced with a thermodynamic disc trap on the reasoning that the disc trap has fewer moving parts and costs less to stock.

The coil worked. It was just worse. Return-air to supply-air temperature rise had been logged at 55 F through the previous winter and came in at 41 F afterwards, a loss of 14 F on a 55 F rise, or about a quarter of the coil's output. The complaint was written up as a coil fouling problem and a cleaning was quoted.

What the readings said. The steam gauge at the coil read 15 psig, so the saturation temperature in that coil should have been about 250 F, read from a steam table rather than remembered. Metal at the coil's inlet header read close to it. Metal at the return header, the point most remote from the steam connection, read 180 F. That is a 70 F gap, seven times the ten-degree limit above, at a location where the gauge insists there is 250 F steam.

Why the gap is the diagnosis rather than a symptom. Fouling on the air side reduces heat transfer everywhere, more or less evenly, and it does not produce cold steam-side metal; if anything it leaves the steam side hotter because less heat is being removed. A gradient this steep along the steam path, with correct pressure at the gauge, is a space occupied by something other than steam. The two candidates are water and gas, and the trap was discharging normally, which left gas.

Why the new trap caused it. The disc trap does the liquid job well. It has no thermostatic element, so it does the gas job not at all. The old float-and-thermostatic trap had been venting the coil on every start-up for years, through a part nobody thought of as a vent, and swapping it deleted the coil's only venting path. The coil then air-bound itself a little more on each cycle, with the air settling at the remote end because it is the heavier gas and because that is where the steam front pushed it.

How much of the loss is which mechanism. Both mechanisms in the sibling article on why air in steam is worse than air in water are running here. Do not try to split them in the field; the measurement that matters is the 70 F gradient, and it resolves to a single fix either way.

The fix and the confirmation. Either return to a float-and-thermostatic trap or keep the disc trap and add a thermostatic air vent at the coil's remote point, piped with a continuous fall to a discharge nobody stands under. The confirmation is the same two numbers taken again: return header metal within about ten degrees of the saturation temperature for the coil gauge pressure, and the air-side rise back at its logged 55 F.

The failure mode if it had been cleaned instead. A coil cleaning performed competently, changing nothing, followed by a quote for a new coil. Two visits and a part, on a fault that is a fitting.

Verifying a vent without opening anything

Compare remote-end metal against saturation for the gauge pressure. This is the whole test and it needs a gauge reading, a steam table and a temperature reading. Take the temperature with a contact probe or on high-emissivity tape, because bare shiny steel reads low on an infrared thermometer, sometimes by tens of degrees, which fakes exactly the fault you are looking for.

Watch the vent during a start-up rather than during running. A vent's entire duty is in the first minutes, so a vent that does nothing on a hot system is behaving correctly and testing it at eleven in the morning proves nothing.

Follow the discharge line for a dip, if it has been piped away.

Count the dead ends. Every closed end of every main, and every piece of equipment with a steam space, needs a venting path, whether that is a vent or a thermostatic element inside a trap. A dead end without either is a permanent air pocket, and it reads cold every morning without ever setting off an alarm.

References

• ASME Steam Tables, or an equivalent saturated-steam table, for the saturation temperature at the gauge pressure you read
• Trap and vent manufacturer selection literature, which owns the choice between a balanced-pressure capsule and a bimetallic element and the correct location on a given piece of equipment
• 29 CFR 1910.147, control of hazardous energy, for isolating and depressurising a section before a vent is removed
• See related: articles in this library on why air in steam is worse than air in water, on what a safety relief valve on a steam system protects, and on how to find where water hammer is being generated