How to Tell Flash Steam From a Leaking Trap

Why this matters

Both look like a white cloud coming out of a pipe. Both are hot, both are loud, and at the same discharge point and the same pressure they look close to identical, because appearance is set by velocity and by how fast room air condenses the vapor, not by where the vapor came from. Every shop that tries to tell them apart by looking replaces good traps. There is one test that separates them cleanly, it takes about two minutes, and it only works under three conditions that have to be established first.

The discharge is the hazard, not the diagnosis

Nothing in this procedure requires approaching a plume, and approaching one is the way people get hurt doing it.

Any discharge at velocity is transparent for the first stretch out of the opening and only becomes visible once it has entrained enough room air to condense. That transparent stretch is the hottest and fastest part of the jet, and it will cut skin before you feel heat. Observe from a body length off and to the side, never from underneath, never downwind, and never put a hand or a face into a plume to judge its temperature. If you must probe a discharge path at all, use a long-handled tool from the side, never a hand.

Where the test requires opening a station test valve, crack it slowly from the side of the stem, with the discharge piped to a drain or a secured hose directed away from every walkway, in heat-resistant gloves and face protection under your PPE hazard assessment per 29 CFR 1910.132 and 1910.133, and never open one that discharges into an occupied space. Sustained observation beside a blowing discharge falls under your hearing conservation program per 29 CFR 1910.95.

The gate, stated with the conditions it needs

At a discharge point served by exactly one trap, of an intermittent-discharge type, that currently has load: does the discharge stop between cycles?

Every clause in that sentence is load-bearing, and each one names a case where the gate cannot run.

• Exactly one trap. A shared receiver vent takes overlapping discharges from many traps, so it plumes continuously in a perfectly healthy system. The gate returns "no stop" on a good system and is worthless there.
• Intermittent-discharge type. A float and thermostatic trap discharges continuously under load by design, and so does a bucket trap on a heavy steady load. The gate returns "no stop" on a good trap. The trap-type table in the testing article names which bodies are which; read the tag before you run the gate.
• Currently has load. A trap on idle equipment does not discharge at all, so the gate returns a clean "stops" on a trap that is failed closed. Confirm the equipment is in service first.

If any of the three fails, do not run the gate. Fall back to the temperature-span method below, which needs no conditions beyond two gauges.

Step 1: Establish the three conditions, in that order

Check the trap tag for the body type. Check with operations that the served equipment is running. Then trace the discharge point on the isometric or by eye and confirm nothing else ties into it, including any bypass around the trap itself, verified by hand on the stem.

The bypass check is the one that gets skipped and the one that produces a false "failed open." A cracked bypass rejoining the trap outlet reads exactly like a blowing trap on every instrument, and that failure has its own article.

Step 2: Watch and listen for at least three consecutive cycles

Put an ultrasonic listener or a contact stethoscope on the trap body and watch the discharge point from a safe distance at the same time. Time the quiet intervals with a watch and write down the seconds.

Three cycles is the floor, not the target. One cycle tells you nothing, because a healthy trap that happens to be mid-discharge when you arrive looks continuous. Two tells you there is a rhythm. Three lets you see whether the rhythm is stable.

Step 3: Read the outcome

Outcome A, the discharge stops. Bursts of plume, matched one for one with the snaps you hear, and genuine silence between them at the sensor. That is a working trap, and the cloud you see is flash from the condensate it just released. The size of that cloud is not evidence of anything, because flash from high-pressure condensate is a small fraction of the mass occupying a very large fraction of the volume. The flash article owns that fraction and the volume ratio.

Outcome B, the discharge never stops. Continuous plume, continuous ultrasonic level, no rhythm at all, on a trap type that is supposed to be intermittent, with load confirmed. That is live steam through the seat.

Step 4: Confirm on temperature before you write it down

The gate is fast and it is not the whole evidence. Close the test valve, let the station settle, and take the confirming reading with a contact probe on brushed bare fittings, upstream and downstream, using the same instrument on the same surface finish so its systematic offset cancels in the difference between the two.

Read both against the two saturation temperatures. Inlet saturation comes from the trap inlet gauge and return saturation from the return header gauge, and steam tables own both conversions. The distance between those two temperatures is the span, and a downstream reading sitting low in the span is condensate discharge, while one sitting high in the span is live steam feeding that pipe.

Do not brush or cut pipe insulation to find bare metal. Thermal system insulation on steam and condensate piping is presumed asbestos-containing material under 29 CFR 1926.1101 in construction, with general industry duties at 29 CFR 1910.1001, until sampled, and that is an inhalation route no glove addresses. Use a fitting that is already bare.

Worked example: one gate, two stations on the same header

Two drip stations on the same 100 psig main, both fitted with the same thermodynamic disc trap, both with a test valve to atmosphere. Return header gauge reads 2 psig. From steam tables, inlet saturation is about 338 F and return saturation is about 219 F, so the span is 119 F.

Conditions first. Both tags read disc trap, which is an intermittent type. The main is in service, so both have load. Both bypasses were checked by hand and found fully closed. All three conditions hold at both stations, so the gate can run at both.

Station 1, 90 seconds of observation. Seven discharge bursts, quiet intervals timed at about 11 seconds each, with a clear silence at the sensor between them. Seven cycles of roughly 11 seconds quiet plus a short discharge fills the 90 seconds, which is a useful arithmetic check that you counted what you thought you counted. Outcome A.

Station 2, 90 seconds of observation. Continuous plume with no visible interruption, continuous ultrasonic level, no snaps at all. Outcome B.

The confirming temperatures, test valves closed. Contact probe on the outlet stub at each station, on metal that was already bare, brushed across the pipe with a long-handled brush in heat-resistant gloves and face protection.

• Station 1 reads 226 F. That is 7 F above return saturation, which is about 6 percent of the 119 F span.
• Station 2 reads 318 F. That is 99 F above return saturation, which is about 83 percent of the span.

Both methods agree at both stations, which is the point of taking the second one. Station 1's plume was flash off correctly discharged condensate. Station 2 is passing live steam.

Check the direction claim rather than assuming it. A trap discharging condensate delivers water at return conditions to that stub, so the stub sits at the low end of the span. A trap passing live steam delivers steam at inlet conditions, which condenses in the stub and drives it toward the high end. Station 1 sits 6 percent up the span and Station 2 sits 83 percent up it. Both ends of the range are named and the readings land where the mechanism puts them.

What would have inverted this. If Station 2's trap had been a float and thermostatic body instead of a disc trap, the continuous discharge would be normal and the gate would have called a working trap failed. The temperature reading would still have been the deciding evidence, which is why step 4 is not optional. And if the return header had been running high, both stubs would read further up the span with neither trap passing, so the span itself has to be rebuilt from current gauges rather than from last year's numbers.

The failure mode of skipping the conditions. A survey run without the trap-type check produces a list where every float and thermostatic trap on a loaded service is marked failed open. Those become replacements, outages, and entries in the plant's failure history, and the replacement bodies produce the identical reading at the next survey, which then gets logged as an early failure of a new trap.

How to verify your call held up

• Re-run the gate at the same station a week later on anything you called Outcome A that still bothers you. A trap in early seat wear shortens its quiet interval before it goes continuous, so the recorded seconds are the early warning.
• Check that every Outcome B call has a temperature reading behind it. Sound-only calls are where the false positives live, and they are the calls that survive into a work order.
• Check that both temperature readings came off the same surface finish with the same instrument. An upstream reading on bare casting against a downstream reading on a jacketed line is not a comparison, because the offsets do not cancel.
• Where you could not run the gate, say so on the sheet rather than recording a state. "Shared vent, gate not applicable, called on temperature span" is a usable record. A bare "OK" on a station the method could not resolve is how a fault survives three surveys.
• If several stations on one header come back Outcome B at once, read the return pressure at a gauge before ordering traps. A loaded return stalls traps, and that is a system finding, not a population of trap failures.

References

• 29 CFR 1910.132 and 29 CFR 1910.133, PPE hazard assessment and eye and face protection, for opening station test valves on saturated condensate
• 29 CFR 1910.95, occupational noise exposure, for sustained observation beside a blowing discharge
• 29 CFR 1926.1101 and 29 CFR 1910.1001, asbestos in construction and general industry, covering presumed asbestos-containing thermal system insulation
• Steam tables for the two saturation temperatures that define the span, and trap manufacturer data for the discharge pattern of the specific body
• See related: What Flash Steam Is and Why It Looks Like a Failure; The Trap That Tested Good and Was Passing Steam