Mechanical Seal Failure Evidence: What to Record Before Cleanup

Before cleaning or discarding a failed mechanical seal, preserve the as-found condition, leak location, operating timeline, support-system state, face wear, deposits, secondary-seal condition, and pump interfaces. A failure mark is evidence, not a complete diagnosis. The goal is to compare competing explanations and avoid ordering a replacement that reproduces the same conditions.

Make the equipment safe before collecting evidence

Failure inspection begins only after the equipment has been isolated, depressurized, drained, decontaminated, locked out, and released under the site’s safety procedure. Process residue, barrier fluid, cleaning chemicals, stored spring force, hot surfaces, and trapped pressure can remain in a seal or support system.

Do not dismantle a hazardous-fluid seal solely to improve the investigation. Follow the plant’s exposure controls, waste rules, and chain of custody. When evidence collection conflicts with safe handling, safety controls take priority and the limitation should be recorded.

Capture the as-found condition in a fixed sequence

Use the same sequence for every event so photographs and notes can be compared across equipment and time.

Before shutdown, when safe and available

Record the observed leak rate or pattern without placing personnel at risk. Note whether leakage is continuous, intermittent, vapor, droplets, a stream, or contained by a drain. Capture local pressure, temperature, level, flow, vibration, alarm, and support-system readings with their timestamps.

Before removing guards, piping, or the gland

Photograph the complete pump or vessel, equipment tag, seal area, gland connections, tubing, valves, reservoirs, coolers, filters, pressure devices, flow indicators, and any visible leak trail. Mark every tube before disconnection so a reversed or misplaced connection can be reconstructed.

During removal

Record coupling position, shaft movement, fastener condition, gasket surfaces, gland orientation, setting-device condition, sleeve location, fretting, corrosion, deposits, and evidence of pipe strain or mechanical contact. Keep the rotating unit and stationary seat paired to the same sample ID.

Before cleaning the seal

Photograph both faces under consistent light, the complete wear track, chipped edges, deposits, blisters, cracks, discoloration, elastomer shape, springs or bellows, drive features, anti-rotation details, and leakage paths. Include a scale and a sample label in the frame.

The Hydraulic Institute’s mechanical-seal guide includes common seal failures, the benefits of failure analysis, troubleshooting approaches, and a typical seal observation report. A repeatable observation form is often more valuable than a large set of unlabeled photographs.

Reconstruct the operating timeline

The latest alarm is not always the initiating event. Build a timeline long enough to include the last known stable period, recent maintenance, process changes, and support-system changes.

Time or event	Process evidence	Equipment evidence	Seal/support evidence
Last stable run	Fluid, concentration, solids, flow, pressure, temperature	Speed, vibration, bearing temperature	Flush/barrier conditions, normal leakage
Maintenance or change	Cleaning, recipe, concentration, utility change	Alignment, bearing, sleeve, coupling, piping work	New seal, repair kit, tubing, set points
Startup	Priming, venting, warm-up, pressure rise	Acceleration, rotation, vibration	Support system established before rotation?
Leak onset	Process state and alarm sequence	Movement, noise, vibration change	Leak location, rate, reservoir or pressure change
Shutdown	Pressure decay, drain-down, cooling	Coast-down or reverse rotation	Trapped pressure, loss of circulation, leakage after stop

Use time-stamped control-system trends where available, but preserve their context. A discharge pressure trend is not necessarily the pressure at the seal chamber. A barrier-system gauge reading taken after shutdown may not represent the value at leak onset.

Inspect four evidence groups separately

Seal faces

Record the width, continuity, and position of the wear track; chips at inside or outside diameter; radial cracks; heat-check patterns; glazing; grooving; transferred material; embedded particles; and deposits. Measure flatness or surface condition only with a method appropriate to the component and after documenting the uncleaned state.

Secondary seals and flexible elements

Record cuts, spiral damage, flattening, swelling, hardening, stick-slip marks, extrusion, loss of elasticity, deposits, and freedom of movement. For bellows, record visible distortion, deposit packing, corrosion, weld condition, and whether the assembly moves freely. Do not infer an elastomer compound from color alone.

Hardware and interfaces

Check sleeve fretting, set-screw marks, drive damage, gasket imprint, gland contact, stationary-seat fit, anti-rotation features, shaft runout, axial play, chamber condition, bearing condition, and alignment evidence. Compare measured values with the equipment drawing or approved installation limit rather than a generic internet value.

Support system

Inspect valve positions, filters, orifices, coolers, reservoirs, accumulators, pressure source, flow path, vents, drains, instruments, alarms, and contamination. Confirm that tags match the intended connection function. Retain fluid samples only under a controlled procedure that prevents mixing or mislabeling.

Flowserve’s official Seal Failure Analysis App description presents visual identification as a way to help troubleshoot and prevent failures. Visual references are useful for naming features, but a photograph match should generate a hypothesis, not close the investigation.

Convert observations into competing hypotheses

Avoid a one-mark/one-cause diagnosis. The same appearance can result from different sequences, and one event can create several secondary marks.

Observation	Possible explanations to test	Evidence that can separate them
Discontinuous or off-center wear track	Misalignment, runout, distortion, face damage, incorrect assembly	Shaft/chamber measurements, face geometry, gland and bearing condition
Heat marks or radial cracking	Loss of lubrication, flashing, insufficient circulation, excessive closing load, thermal shock	Pressure/temperature timeline, fluid vapor behavior, flush flow, face pair
Abrasive grooving	Process solids, dirty flush, installation contamination, damaged counterface	Particle/source analysis, filter condition, line cleanliness, deposit location
Secondary seal swelling or sticking	Chemical incompatibility, temperature exposure, deposits, wrong groove or compound	Compound record, process/cleaning media, temperature history, dimensions
Bellows or spring deposit packing	Crystallization, coking, solids, poor circulation, dead zone	Process behavior, temperature cycle, chamber/flush arrangement, deposit analysis
Sleeve fretting or drive damage	Loose fit, motion, incorrect setting, worn bearing, assembly error	Fit measurements, runout, axial movement, installation record

For every hypothesis, write what would have to be true and which evidence would contradict it. If the needed evidence was lost during cleaning or disposal, record the gap instead of converting an assumption into a finding.

Preserve samples and records for a technical review

Label every component with equipment tag, removal date, position, orientation, and sample ID. Store paired faces and secondary seals so they cannot be mixed with another failure. Keep original photo files, operating trends, drawings, purchase records, maintenance notes, and the inspection form under one event reference.

The review package should include:

• a one-page event timeline;
• as-found and disassembly photographs;
• pump, chamber, seal, and support-system drawings;
• operating and alarm trends with tag definitions;
• current fluid and cleaning-media information;
• material and prior repair records where available;
• measured equipment condition and instrument status;
• the proposed causes, contradicting evidence, and open questions.

Change the replacement scope only when the evidence supports it

An obsolete seal code or damaged sample may still need an OEM replacement review, but cross-reference alone does not address the failure environment. If the duty has changed, repeat the application review rather than copying the old materials and arrangement.

The chemical pump application page provides a route from failure evidence to the relevant seal families and process inputs. Any material, arrangement, support-plan, equipment, or operating change should be approved by the responsible plant and equipment authorities.

A defensible outcome may be “insufficient evidence to isolate one cause.” That is still useful when the record identifies which measurements, trends, samples, or commissioning checks must be added before the next run.