Why Condensate Return Lines Fail From the Inside

Why this matters

Condensate return piping fails from the bore outward, so by the time a wet spot appears on the lagging the wall under it is already mostly gone, and the section next to it is usually in the same condition. Shops replace the leaking joint, and the line leaks again a few feet away within the year. The section you cut out is a written record of which mechanism did it, and it can be read in a few minutes: the location of the damage, its shape, and whether there is corrosion product sitting in it name the cause and the owner of the fix before any water sample is drawn.

Before you cut anything out

A condensate line that feels cool can still be pressurized. Temperature is not a pressure test, and a return that has been idle since the weekend is at whatever the receiver is at. Do not judge by hand and do not cut on the assumption a line is dead.

Isolate at both ends, lock and tag under your written energy control procedure per 29 CFR 1910.147, drain to a piped discharge directed away from every walkway, confirm zero at a gauge on the section itself rather than assuming the drain finished, and let it cool before any cut.

Three separate hazards attach to the cutting itself. The first two are inhalation routes, which means gloves and glasses do nothing for either of them.

• Insulation. Thermal system insulation on condensate piping is presumed asbestos-containing material under 29 CFR 1926.1101 in construction, with the general industry duties at 29 CFR 1910.1001, until sampled. Removal is done by a trained crew under the applicable standard, not by the crew doing the pipe work.
• Coatings. Cutting or grinding painted steel of unknown vintage can release lead, with duties at 29 CFR 1926.62 in construction and 29 CFR 1910.1025 in general industry, and respiratory protection supplied under a written program per 29 CFR 1910.134. Strip or test the coating in the cut zone before a torch or a grinder touches it.
• Hot work. Torch cutting requires a hot work permit and fire watch under NFPA 51B in the edition your authority having jurisdiction has adopted, which reaches you through that permit, alongside 29 CFR 1910.252 for general industry or 29 CFR 1926.352 for construction.

Where the line runs through a below-grade pit, vault or tunnel, treat that space as a permit-required confined space until evaluated under 29 CFR 1910.146.

Why the bore is the aggressive side

The outside of a return line sees room air. The inside sees hot water that has just been steam, carrying whatever came over with it, moving in a partly-filled channel at velocity, and re-boiling every time the pressure steps down.

The chemistry of why returning condensate turns aggressive belongs to the water treatment articles, which own dissolved gas carryover and its control, and is not re-derived here. What this article owns is the mechanical consequence: each of the three interior mechanisms attacks a different part of the pipe's geometry, and that is what makes them separable by eye.

Three signatures, and what each one names

The groove. A gravity return runs part full, so the condensate stream occupies the bottom of the bore and the top of the bore sees vapor. Attack that depends on contact with the liquid therefore concentrates in a channel along the invert, and the top half stays near nominal wall. The signature is the contrast between the two, not the depth by itself.

The pits. Air entering the system produces localized attack rather than a channel, so the damage is discrete and deep instead of long and shallow. The useful part is the location: air gets in where pressure goes below atmospheric, which means vent connections, pump seals and packing, and any section that pulls a vacuum as it cools after shutdown. Follow the pits back to a place that can go subatmospheric and you have found the ingress.

The erosion. Condensate leaving a trap flashes, and the vapor formed in the outlet accelerates the whole mixture. That momentum lands on the outside of the first bend it meets. Erosion removes metal mechanically and takes the corrosion product with it, so an eroded surface is bright and smooth where a corroded one is rough and loaded with scale. Absence of product is the diagnostic, not the shininess.

What an ultrasonic thickness survey will and will not show you

A thickness gauge is the right tool and it lies in one specific, predictable direction on a grooved pipe.

A grid survey reports where you landed, not the minimum. A groove may be a fraction of an inch wide. A one-inch grid over it mostly lands beside it, so the survey reports near-nominal wall on a pipe that is two thirds gone along a narrow line. That is a systematic bias in one direction, toward reporting more wall than exists, so taking more readings on the same coarse grid does not improve it. The fix is a fine traverse along the groove, found first by eye or by feel on a cut section.

The instrument's own error does not cancel here. A calibration or couplant offset common to every reading cancels when you subtract two readings from each other, which is the case where a fixed offset is harmless. Remaining-wall is an absolute number, not a difference, so the offset lands on it at full strength. Calibrate on a known block of the same material and confirm the reading on a section of the same pipe you can measure with a caliper.

Worked example: three sections off one return main

A 2 inch schedule 40 return main, nominal wall 0.154 inch, has leaked three times in fourteen months. Three sections come out under the isolation, drain-and-verify and hot work controls named above, and each is split lengthwise on the bench after it has cooled, then read with a calibrated thickness gauge and a light. Splitting a removed, cooled, depressurized section is bench work; splitting anything still tied into the main is not.

Section 1, a horizontal run mid-main. A groove along the invert, minimum 0.052 inch. That is 0.052 divided by 0.154, or 34 percent of nominal remaining, so 66 percent gone. The rest of the circumference reads 0.150, essentially full. Corrosion product sits in the groove.

Contrast is the finding. Sixty-six percent loss in a narrow channel against near-nominal wall two inches away is not general thinning and it is not external. It is attack concentrated where the liquid runs, and the owner of that fix is water treatment. Replacing this section with identical pipe returns you here.

Section 2, a vertical riser near the receiver. Scattered round pits, one of them through-wall, and wall between the pits reading 0.148, which is 96 percent of nominal. Product over the pits.

Ninety-six percent of the wall intact and a hole in it is the whole signature. This pipe has no general loss at all. The fix is not the pipe and not the water treatment program by itself; it is finding where air is getting in, and this riser's proximity to the receiver puts the vent and the pump seals at the top of that list.

Section 3, an elbow about 18 inches downstream of a trap outlet. Outside of the bend reads 0.061 inch, which is 40 percent of nominal remaining, 60 percent gone. The surface is smooth and bright with no scale anywhere on it. The inside of the bend reads near nominal.

No corrosion product on a 60 percent loss means the metal was removed mechanically. The location, on the outside of the first bend after a trap, names flash-accelerated flow. The owner of this fix is sizing and geometry: return line volume sized for the vapor rather than the liquid, more straight length between the trap outlet and the first change of direction, or a heavier-wall fitting where a change of direction is unavoidable.

Check the readings against the grid problem before believing any of them. All three minima came from a traverse along the visible damage on a split section, not from a grid on the outside of an intact pipe. Had this been an in-service grid survey, section 1's groove would very likely have been missed entirely and section 3's bend would have been read at the wrong point on the circumference, because the loss there is on one side.

What the three findings together say. Three sections, three mechanisms, three different owners. A plant that replaces all three with identical pipe has bought one outcome: the same three failures on the same schedule. A plant that reads them has a water treatment action, an air ingress hunt, and a piping sizing correction, and only the third of those is a pipe job.

What changes the reading

• A pumped or flooded return runs full, so the liquid contacts the whole circumference and the invert groove signature disappears into more general circumferential loss. The absence of a groove on a flooded line is not evidence the water is fine.
• Intermittent service leaves the line wet and idle. A batch process return sits full of standing condensate between runs, which changes where air ingress matters and can put pitting in places a continuously flowing line would not have it.
• Non-ferrous or stainless returns do not produce the same product. The location logic still holds, the colours and the tubercles do not, so read location and geometry first and material appearance second.
• A line that was replaced in sections over the years is a mixed population. Date the sections before treating a failure interval as a property of the whole main.

How to verify a return-line finding before you spend on it

• Split and look at the bore, on a removed and cooled section only. Everything above depends on seeing the inside surface, and no external survey substitutes for it on the first failure.
• Record whether product is present or absent on every measured minimum. That single field separates erosion from corrosion and it is the field most inspection sheets do not have.
• Map the failures on the isometric before ordering pipe. Three failures clustered downstream of traps is a sizing finding. Three failures spread along horizontal runs is a chemistry finding. The map makes that visible and a list of dates does not.
• Traverse along the damage rather than gridding across it on any follow-up thickness survey, and calibrate on a known block of the same material before the first reading.
• Where the finding is air ingress, look for the subatmospheric point rather than more pipe. Vents, pump packing and cooling sections are where to start, and the pits themselves point back along the line toward it.

References

• 29 CFR 1910.147, control of hazardous energy, for isolating, draining and verifying zero pressure before cutting condensate piping
• 29 CFR 1926.1101 and 29 CFR 1910.1001, asbestos in construction and general industry, covering presumed asbestos-containing thermal system insulation
• 29 CFR 1926.62 and 29 CFR 1910.1025, lead in construction and general industry, with respiratory protection under a program per 29 CFR 1910.134, for cutting or grinding coated steel
• NFPA 51B, in the edition adopted by your authority having jurisdiction and binding through the hot work permit, together with 29 CFR 1910.252 or 29 CFR 1926.352 for the corresponding welding and cutting duties
• See related: What Flash Steam Is and Why It Looks Like a Failure; Why a Failed-Closed Trap Is Found First and Fixed First