Seal Failure Decoded: Diagnosing the Root Cause

Why does my mechanical seal keep failing on the same pump?

It's one of the most common questions I'm asked. And in 36 years of investigating seal failures across process plants, the answer is almost never the seal itself.

If you're replacing the same seal on the same pump for the third time, you're not doing maintenance. You're living inside a pattern you haven't diagnosed yet.

The failure report says wear. Leakage. End of service life. That's a description. It isn't a diagnosis. And until you treat it as a diagnostic problem rather than a procurement problem, the pattern repeats.

Here's where I start.

Root Cause One: The Temperature the Historian Has Been Logging and Nobody Has Read

The most common cause of repeated seal failure isn't dramatic. It's a 15 to 20 degree gap between the design assumption and what the process is actually running.

That gap doesn't trigger an alarm. It doesn't show up in a weekly report. But it's devastating to an elastomer that was never specified for the real operating condition.

FKM has a serviceable temperature range. Run it 15 to 20 degrees above that ceiling consistently and you don't get years of service life. You get months. The material hardens, loses its ability to conform, and the seal faces run without the film they need.

The data is almost always in the historian. It's rarely been looked at in the context of the seal specification.

Before ordering a replacement, pull the actual logged temperature for that service point. Compare it to the elastomer's rated limit. If there's a gap, you've found your root cause, and a material upgrade alone won't fix it. The operating condition has to change, or the specification has to be written for the condition that actually exists.

Root Cause Two: The Shaft Movement Nobody Measured

Mechanical seal faces are finished to within two microns. The hydrodynamic film that keeps those faces apart, the controlled leakage that makes the seal work, exists in that same order of magnitude.

Half a millimetre of axial shaft movement doesn't sound like much. To a component engineered in microns, it's a wrecking ball. It exceeds the seal's compensation range, collapses the film, and puts the faces into direct contact. The faces don't last long after that.

This is measurable. A dial indicator at the stuffing box under operating load gives you the number in minutes. Most plants have never taken that measurement on a pump with a chronic seal failure history.

Take it before the next replacement goes in.

Root Cause Three: The Flush Plan That's Delivering the Problem

A Plan 11 flush, recirculating from pump discharge back to the seal chamber, is the right solution for clean, cool process fluid. It is the wrong solution for anything carrying suspended solids.

When you recirculate an abrasive slurry back across the seal faces, you're not protecting the seal. You're sandblasting it. The faces score. The elastomers degrade. The seal fails. You fit a higher-specification replacement and the cycle starts again, because the flush plan hasn't changed.

The fix isn't a better seal. It's a different flush arrangement, Plan 32 external flush with clean compatible fluid, or a Plan 41 with a cyclone separator to remove solids before recirculation.

If your plant is running Plan 11 on abrasive service, that's worth checking before the next seal goes in.

Root Cause Four: Chemical Attack the Specification Never Accounted For

This one is designed in before the seal ever runs.

Process fluids drift. Cleaning cycles introduce chemistry that wasn't in the original specification. Concentrations change. A fluid that was compatible with the specified elastomer five years ago may not be compatible with what's actually running through the line today.

The result is chemical attack, swelling, softening, blistering, or surface degradation on the elastomer that looks like thermal damage but isn't. Replacing like-for-like doesn't solve it. The elastomer family needs to change.

If the failed seal shows swelling or surface blistering rather than hardening, that's the direction to look. Cross-reference the current process fluid, not the original line designation, the actual fluid, against the elastomer's chemical resistance data. They may no longer match.

The 30-Second Inspection Most Engineers Skip

Before the failed seal goes in the bin, inspect the dynamic O-ring seat, the bore where the O-ring travels to track shaft movement.

Look for a witness mark. A band of fretting or polishing concentrated at one end of the travel range rather than distributed across it tells you the shaft was running at its axial limit on every rotation. The seal was maxing out its compensation range continuously. That's not a seal failure. That's a shaft movement problem presenting as a seal failure.

Thirty seconds. Most of the time, that evidence goes in the scrap bin unread.

Four Questions Before the Next Replacement

If a seal has failed more than twice on the same service point, answer these before specifying a replacement:

• What is the actual logged temperature in the seal chamber, not the design figure, the historian figure?
• What is the measured axial shaft movement at the stuffing box under operating load?
• What fluid is actually reaching the seal faces, and is the flush plan fit for that fluid?
• What does the failed seal show, hardening, swelling, blistering, fretting, face scoring? Each failure mode points somewhere different.

The replacement might still be a material upgrade. But the root cause is almost always one of these four. Fix the condition, not just the component.

If Your Seal Has Been Replaced More Than Twice

Send me the failure history and the current specification. I'll tell you within 24 hours whether there's a root cause worth addressing, and what I'd look at first.

No obligation. No sales process. Just a second set of eyes from someone who has been diagnosing this pattern for 36 years.

office@acumenseals.co.uk | 07767 842601 | acumenseals.co.uk

---

Andrew Sykes MCGI, 36 Years Solving Seal Problems

For independent seal failure diagnosis and specification review, contact Acumen Seals & Pumps Ltd.