The Fault Changed After My PM: Disturbed vs Pre-Existing Decision Tree

Why this matters

This is the close cousin of the brand-new fault, but harder. Here the complaint that brought you out is the same complaint, except it now presents differently after your maintenance. An intermittent noise became constant. A slow leak became a steady drip. A circuit that tripped occasionally now trips every cycle. The customer reasonably asks: did your work make it worse, or did your work simply expose a problem that was always going to surface?

The stakes are the same as any disturbed-versus-coincidence question, but the diagnostic challenge is sharper because you cannot point to a clean before-and-after. The fault existed in both states. You are reasoning about a change in degree, not a change in existence. Disciplined techs separate two real mechanisms: your maintenance physically altered the fault's behavior, or your maintenance changed an operating condition that unmasked a pre-existing weakness. Both are common. Both have distinct fingerprints.

Symptom presentation

The hallmark is continuity of fault type with a change in character. The plumbing leak is at the same fitting, now faster. The HVAC noise is from the same bearing, now louder. The electrical trip is on the same breaker, now more frequent. The location and family of the fault did not move. Its intensity, frequency, or consistency did.

A second pattern is a restored-baseline effect. After a PM, the system often runs harder or differently. A cleaned coil moves more air and loads the blower more. A flushed water heater recovers faster and cycles the gas valve more. A retorqued panel carries fuller current without the resistance of a loose lug. When the system returns to design performance, a marginal component that was coasting under reduced load is suddenly stressed to its real limit.

Quick checks

• Did your PM restore the system toward design performance? If yes, suspect that the restored load unmasked a marginal part.
• Did your PM apply direct physical force to the fault zone (torque, vibration, thermal cycling, fluid pressure)? If yes, suspect direct disturbance.
• Is the change in degree consistent with more load, or with mechanical damage? Faster failure under heavier load points to unmasking. A new crack, nick, or backed-out fastener points to disturbance.
• Was the failing component already on a documented wear trajectory? Pre-existing weakness changes the honest framing even when your PM accelerated the timeline.

Isolation tree

Step 1: Confirm fault continuity. Verify the failing component is the same one implicated in the original complaint. If it moved to a different part, you are in new-fault territory, not changed-fault territory.

Step 2: Classify the change. Is it more frequent, more intense, or more constant? More constant often means a marginal threshold was crossed. More intense often means added load. A genuinely new failure mode (a different noise, a different trip signature) suggests direct disturbance.

Step 3: Test the restored-load pathway. Measure the operating condition you changed. Higher airflow, higher pressure, fuller current, faster recovery. If the system now runs at or near design and the weak part is failing under that legitimate load, the part was pre-existing and your PM unmasked it. This is not damage you caused. It is a hidden weakness your correct work revealed.

Step 4: Test the direct-disturbance pathway. Inspect for mechanical insult exactly as in any disturbed-fault case: fresh marks, partial seating, wrong torque, pinched gaskets. If the fault changed because you physically altered the part, the evidence is on the part.

Step 5: Adjudicate. If both pathways are plausible, weigh which mechanism better explains the specific change in character. Restored load explains intensity and frequency changes under design conditions. Direct disturbance explains new failure modes and insult evidence.

Confirming diagnosis

Pre-existing-unmasked is confirmed when you can show the system now runs at design performance, the failing part shows independent wear, and the heightened symptom is what you would expect from that part under full legitimate load. ISO 13379-1 condition-monitoring logic supports this: a part on a wear trajectory crosses its functional threshold sooner when stress rises, even though the wear was already there.

Direct-disturbance is confirmed by insult evidence plus a plausible physical pathway from your procedure to the fault.

Next steps

If unmasked pre-existing: document the restored performance reading, show the worn part, and frame the repair as the next stage of returning the system to health. This is an honest, defensible position when backed by measurement.

If disturbed: correct at no charge and revise the procedure.

Either way, capture the operating-condition measurement. A before-and-after performance number (airflow, pressure, current, recovery time) is the evidence that distinguishes these two outcomes and protects you in any dispute.

References

• ISO 13379-1, Condition monitoring and diagnostics of machines, data interpretation and diagnostics techniques.
• ISO 17359, Condition monitoring and diagnostics of machines, general guidelines.
• NFPA 70B, Standard for Electrical Equipment Maintenance, condition-based maintenance and performance baselining.
• ACCA Standard 4, Maintenance of Residential HVAC Systems, restoring equipment to design performance.