Views: 0 Author: Site Editor Publish Time: 2026-10-23 Origin: Site
A cartridge seal is supposed to be the easy one. Pre-set at the factory, pressure-tested, boxed up to drop on the shaft. Then the pump runs and the gland is wet before the shift ends. Most of those day-one leaks are not the seal. They are a missed step during the install.
The whole point of a cartridge is that the factory does the hard part, the manual face setting, before it ships. Spring load lands correct, faces sit where the drawing says. That same pre-set is where installs go wrong, because the clips that hold the setting have to come off at exactly the right moment and the length has to survive until the gland is torqued down. Leave one clip and the faces never move.
Check | What gets missed | What it does to the seal |
|---|---|---|
1 Setting clips | Clips left on, or wrong working length | Faces never seat, or springs over-compressed and faces wear out fast |
2 Secondary seals | O-ring twisted, wrong groove, wrong compound | A second leak path opens, independent of the faces |
3 Gland bolts | Uneven or over-tightened, no cross pattern | Stationary face distorts, seal leaks across the faces |
4 Sleeve and chamber | Scored sleeve, burrs, dirt on faces | Stable compression never forms, immediate weep |
5 Flush and prime | Pump not flooded, no flush flow, trapped air | Faces run dry for seconds and scuff |
1, the setting clips. Start with the clips. A new seal that weeps on its first shift almost always has a clip still in it. The factory pre-loads the spring to the right working length and holds it with clips or a spacer. Bolts down with those clips still fitted and the faces cannot reach their running position, so the seal weeps from the first turn. The error hides well, the clip sits flush and looks like part of the gland. Pull every clip only after the set screws are tight on a clean shaft, then turn the shaft by hand and feel for bind before the first start. Working length matters just as much. Too short and the springs sit over-compressed, faces gone in months. Too long and they cannot hold the faces together at startup or standby, so it leaks. The number comes off the seal drawing, measured from the chamber face. On a semi-open impeller pump, set the impeller clearance first, because moving the impeller moves the shaft and changes the compression. Shops that skip this one send seals back as defective when the real fault was a measurement.
2, the secondary seals. The O-rings do the quiet work, and they fail just as quietly. A ring in the wrong groove extrudes the moment pressure lands and leaks. Two rings, static and dynamic, different compounds, get swapped in a hurry and the wrong one ends up in the wrong seat. A ring nicked while sliding over an edge opens a leak path that has nothing to do with the faces. Material bites too. EPDM swells in petroleum oil and grease, so a dab of petroleum jelly or mineral oil on an EPDM ring starts a weep without anyone touching the faces. Only the lube that ships with the seal belongs on those rings, and the faces stay clear of it. A rubber bellows gets fitted with the fluid its maker calls for, nothing else.
3, the gland bolts. Bolts look like the last job. They are usually where a clean install turns into a leaking one. Pulled down uneven or too hard, the stationary face distorts or the mating ring tilts and the seal leaks straight across the faces. Nuts go on across the gland, alternating, in steps, to the torque the seal instructions give. No pipe slipped over the wrench for extra reach. Tightening harder is the reflex that makes it worse. A seal weeping on the first shift gets its setting and faces checked before the wrench comes back out, because more load only adds distortion. Most crews learn this one after they have already made it worse.
4, the sleeve and the chamber. The seal is only as good as the metal behind it. A sleeve grooved from old packing, a chamber face with burrs or corrosion pits, a polished face that picked up a fingerprint on the way in, any of those stops stable compression from forming. The leak reads as seal failure while the cause sits in the support geometry. A cartridge does not forgive poor hardware. Shaft, sleeve and stuffing box get cleaned, the chamber face checked flat and free of burrs, the faces covered until they seat. One Texas Gulf plant set a fresh seal on a scored sleeve, gland wept the first shift, seal took the blame. The sleeve was the problem, and it cost them a second teardown.
5, the flush and the prime. The pump has to be ready before the seal is. Casing filled with the right fluid or flush, flow confirmed at the flush port, chamber vented, because trapped air on a vertical pump dry-runs the faces in seconds and scuffs them before they ever warm up. A dry start burns faces faster than nearly any duty the pump will see. A dual seal takes barrier or buffer fluid first or the inboard faces run dry. Plenty of day-one leaks are really start-up leaks, the seal mounted fine and the support side not ready. A seal can go in perfectly and still fail in the first hour because nobody confirmed the flush. That one shows up in the reports as a seal problem.
Five checks cover most single-seal cartridge work. Order matters less than doing all five, every time. A seal that leaks on day one is rarely a bad seal. It is a step someone skipped. This list stops at single seals, dual arrangements add barrier fluid and a different failure set, and that is its own job. FBU builds pre-set, pressure-tested cartridge seals, the EDMA, GSMH, GDMH, EDMH and ESMH lines for ANSI B73 centrifugals, Sulzer APP compatible pumps, pulp and paper and general chemical duty. Each unit ships with the setting locked and the working length marked, so what happens at the bench decides whether it earns the rating or leaks inside a shift. The seal arrives ready. The rest is on the install.
Why does a cartridge seal leak immediately after installation?
The usual causes are a clip left in, gland bolts pulled uneven, faces contaminated or scarred, or a pump started dry with no confirmed flush. The seal itself is rarely at fault on the first shift. Shops that blame the seal first usually find an install step they skipped.
Is leaving the setting clips in really that common?
It is the most repeated mistake across pump rooms and skids, bar none. The clip sits flush against the gland and looks like it belongs, so it stays. Faces never reach running position and the pump weeps before the operator finishes the round. Anyone who has torn one down twice in a week has seen it.
Can a scored shaft sleeve cause a same-day leak?
Yes. Grooves, pits or burrs on the sleeve or chamber face stop stable compression from forming, so the seal weeps on the first shift even with the cartridge fitted right. The seal takes the blame while the hardware is the cause. A scored sleeve is a sleeve that should have been replaced, not a seal that should have been returned.
How tight should the gland bolts be?
Nuts go on across the gland, alternating, in steps, to the torque the seal instructions give. Uneven or over-tight load distorts the stationary face and the seal leaks across it. A new seal weeping at start gets its setting and faces checked before the wrench comes out. More torque is not the fix, it is usually the cause.
