What Flash Steam Is and Why It Looks Like a Failure

Why this matters

Every healthy steam system produces a cloud at its condensate vents, and shops replace good traps every year trying to stop it. Flash is not a leak, not a fault, and not evidence of anything wrong. It is the arithmetic of what saturated water does when its pressure drops, and it is unavoidable in any system where condensate leaves a high pressure and arrives at a lower one. What makes it read as a failure is that a small fraction of the mass turns into an enormous fraction of the volume, and the eye is measuring volume.

Handling flashing condensate

Flash happens the instant the pressure drops, which means it happens at the outlet of whatever you just opened, in your face rather than downstream.

Open a drain, a test valve or a blowdown by cracking it slowly from the side of the stem with the discharge piped to a drain or directed by a secured hose away from every walkway, in heat-resistant gloves and face protection under your PPE hazard assessment per 29 CFR 1910.132 and 1910.133. Never open one that discharges into an occupied space. Never stand in front of the outlet, and never put a hand into a plume to judge whether it is "only flash," because any discharge at velocity is invisible for the first stretch out of the opening and that invisible stretch is the hottest part of it. Sustained work beside a venting receiver falls under your hearing conservation program per 29 CFR 1910.95.

What is actually happening at the pressure drop

The physics of phase change itself belongs to the latent heat articles and is not re-derived here. What matters for steam systems is a bookkeeping fact about saturated liquid, meaning condensate sitting at the boiling point for its own pressure, which is the state condensate leaves a trap in.

Saturated liquid at a high pressure holds more sensible heat per pound than saturated liquid at a low pressure can hold, because the boiling point itself is higher. Drop the pressure and that excess sensible heat has nowhere to go. It cannot leave as heat, because there is nothing colder in contact with it, and it cannot stay as sensible heat, because the water can no longer be that hot and remain liquid. So it converts: a fraction of the pound boils instantly, and boiling it consumes exactly the excess.

That is why the flash is not new energy and not lost energy at the moment it forms. It is the same energy re-expressed. It becomes a loss only if you then vent it.

The fraction, and which way it moves

The fraction that flashes is a straight enthalpy balance:

Flash fraction = (liquid enthalpy at the high pressure - liquid enthalpy at the low pressure) / latent heat at the low pressure

All three of those come from steam tables, and the steam tables own the answer for your actual pressures. Representative table values for the range most field service systems run:

Saturated liquid runs about 0.017 ft3/lb across that whole range, which is the number that makes the volume argument below work.

Direction. A larger pressure drop makes a larger fraction. Flashing 100 psig condensate down to 15 psig gives (309 minus 218) divided by 946, which is 9.6 percent. Flashing the same condensate all the way to 0 psig gives (309 minus 180) divided by 970, which is 13.3 percent. Dropping further gives more, because the numerator grows faster than the denominator does.

What it does not depend on. Not the trap, not the trap's condition, not the pipe size, not the load. Fix the two pressures and the fraction is fixed. That single sentence is the most useful diagnostic property flash has, and it is why a plume that changes at unchanged pressures and unchanged load cannot be flash.

Why 13 percent of the mass looks like all of it

Take one pound of 100 psig condensate arriving at an atmospheric receiver.

• The vapor: 0.133 lb at about 26.8 ft3/lb is about 3.56 ft3.
• The liquid left: 0.867 lb at about 0.017 ft3/lb is about 0.0147 ft3.

The vapor occupies roughly 240 times the volume of the water it left behind. That ratio is the entire reason flash reads as a catastrophic leak: 13 percent of the mass, filling a space 240 times larger, moving fast enough to carry itself out of the vent as a visible cloud, while 87 percent of the mass slides quietly into the receiver where nobody looks.

Check the energy closes, because it is a good sanity test on any flash number you compute. Boiling 0.133 lb at about 970 Btu/lb consumes about 129 Btu. The sensible heat given up by the pound in going from 309 to 180 Btu/lb is 129 Btu. Those match by construction, which is what tells you the balance was set up correctly.

What flash forces on the piping

Three consequences follow directly, and all three are places shops get surprised.

Return lines are sized for vapor, not for water. A return carrying that pound is carrying 3.56 ft3 of vapor and 0.0147 ft3 of liquid. A line sized on the condensate flow rate is undersized by orders of magnitude in volumetric terms, and the result is velocity, noise, erosion, and a return pressure that climbs. That climbing return pressure then stalls traps upstream, which is how one sizing decision produces a population of apparent trap failures.

Trap capacity tables assume a stated differential. A published trap capacity is given at a specific pressure differential, because flash forming in the orifice occupies part of it. Read the capacity at the differential the trap will actually see, which is inlet pressure minus return pressure, not inlet pressure minus atmospheric.

The first fittings downstream of a trap take the erosion. The two-phase mixture leaving a trap is accelerated by the flash forming in it, and the outside of the first bend is where that momentum lands. That produces a distinctive smooth, bright thinning with no corrosion product, and the article on how return lines fail from the inside covers reading it.

Worked example: a receiver vent that a shop tried to fix twice

An atmospheric receiver takes condensate from a 100 psig service. It vents a heavy continuous plume. Two traps on that service have already been replaced on the theory that one of them was blowing.

Run the fraction. Both pressures are known: 100 psig in, receiver at 0 psig. From the table, (309 minus 180) divided by 970 is 13.3 percent. So 13.3 percent of every pound of condensate reaching that receiver is going to appear as vapor, in a healthy system, with every trap perfect.

Run the volume. At 26.8 ft3/lb, that 13.3 percent of the mass is about 240 times the volume of the liquid it came from, using the two specific volumes stated above. The plume is not a small signal. It is the largest thing happening at that receiver by volume, and it is supposed to be there.

Now test whether anything is wrong. The fraction is fixed by the two pressures, so ask whether either has moved. The gauge log shows the header steady at 100 psig and the receiver at atmospheric throughout. Condensate load is set by the equipment, and production hours are unchanged. Under those three conditions the flash rate cannot have changed, so any change in the plume would have to be added mass, and no change was recorded. The plume was always like that; nobody had looked at it before.

What the two trap replacements bought. Nothing, and worse than nothing, because both replacements were logged as failures and now sit in the plant's failure-rate history, inflating the rate used to set survey intervals. A fabricated failure rate produces a survey interval tuned to a population that does not exist.

What would have made this a real finding. If the receiver had been vented to a lower pressure than before, the fraction rises and the plume genuinely grows with nothing wrong. If the header pressure had been raised, same result in the same direction. If production hours rose, more condensate arrives and the plume grows in proportion, still with nothing wrong. And if all three held steady and the plume grew anyway, that is added mass, which is a real fault and points at a path around a trap rather than at the traps themselves.

How to verify you are reading flash and not a fault

• Get both pressures at gauges before forming any opinion, because the fraction is meaningless without them and every argument above is built on the pair.
• Compute the fraction from steam tables for your actual pressures, not from a remembered figure. Thirteen percent is the 100 psig to atmospheric case specifically, and a 15 psig to atmospheric service flashes about 3.9 percent by the same balance, which is under a third of it.
• Establish whether the plume changed, and against what. A plume that is merely large is normal. A plume that grew while both pressures and the load held steady is added mass and needs a path traced.
• Check whether the discharge point serves one trap or many. A vent serving many traps plumes continuously even when every trap is cycling correctly, because the discharges overlap. That case cannot be resolved by watching the vent at all.
• Before condemning a trap on a plume, check the return header pressure. A return running high both raises the plume at the vent and stalls traps, and replacing traps into that condition puts new traps into the same stall.

References

• Steam tables for saturation temperature, liquid enthalpy, latent heat and vapor specific volume at the pressures actually measured, which are the authority for every value in the balance
• 29 CFR 1910.132 and 29 CFR 1910.133, PPE hazard assessment and eye and face protection, for opening drains, test valves and blowdowns on saturated condensate
• 29 CFR 1910.95, occupational noise exposure, for sustained work beside a venting receiver
• Trap manufacturer capacity data, read at the differential the trap will see rather than at inlet pressure minus atmospheric
• See related: How to Tell Flash Steam From a Leaking Trap; What a Flash Tank Is Recovering