Cleared One, Now Another Surfaces: Stop or Continue Decision Tree

Why this matters

You cleared the fault you came for, and another surfaced. You cleared that one, and a third appeared. Now you face a different kind of decision than pure diagnosis: a scope and judgment call. Do you keep going until the system is fully sound, or do you stop, report, and let the customer decide? Push too far and you run up a bill the customer never authorized and may refuse. Stop too soon and you hand back a system that fails again tomorrow, generating a callback and the impression you did not finish the job.

This is not the same as anticipating a single masked fault. This is the situation where multiple faults keep surfacing in sequence, and you have to decide whether you are looking at a finite cascade with an end in sight or an open-ended pit of deferred maintenance. The right answer depends on safety, authorization, and whether you can see the bottom.

The pattern is universal: a neglected furnace where each fix reveals the next, a plumbing system where one repair pressurizes the next leak, an old panel where each corrected circuit exposes another, an appliance where one part swap surfaces the next worn component.

Symptom presentation

The signature is serial revelation: each repair, done correctly, exposes a new fault rather than completing the system. Two profiles matter. A finite cascade has a small number of series faults along one operating path, and you can reasonably see the end. An open-ended cascade is a system in general deferred-maintenance failure, where faults will keep surfacing because the whole equipment set is past its service life.

Distinguishing the two early is the core skill. A finite cascade is worth finishing on the trip. An open-ended one calls for a stop, a written assessment, and a decision the customer owns.

Quick checks

• Are the surfacing faults along a single operating path, or scattered across unrelated subsystems? One path suggests finite; scattered suggests systemic decay.
• Can you see the end? Estimate how many more faults likely remain. If you cannot bound it, treat it as open-ended.
• Does the next fault involve a safety condition? Safety changes the calculus toward must-address-or-make-safe-and-stop.
• Has the customer authorized this scope? Each new fault may exceed the original authorization.

Isolation tree

Step 1: Classify the cascade. Decide finite versus open-ended using the path test and the bounding test. Faults in series on one path that you can count point to finite. Faults across multiple aging subsystems with no clear end point to open-ended.

Step 2: Apply the safety gate. If a surfaced fault is a safety hazard, you must either correct it or make the system safe before leaving, regardless of scope or authorization. Safety overrides the stop-or-continue economics. A system left in an unsafe state is never an acceptable stopping point.

Step 3: Check authorization against accumulating scope. The original call authorized the original fault. Each new fault that adds material cost or time generally needs the customer's agreement before you proceed. Continuing past authorization on a non-safety repair is a billing and consent problem waiting to happen.

Step 4: For finite cascades, continue with consent. If the end is in sight, the faults are related, and the customer agrees to the running scope, finishing on one trip is efficient and serves the customer. Confirm authorization, then complete the path.

Step 5: For open-ended cascades, stop and report. When you cannot bound the work, stop at a safe, documented point. Produce a written assessment of what is failing and the likely trajectory, and let the customer decide between continued repair and replacement. This is the honest call on equipment in systemic decline.

Confirming diagnosis

Your scope decision is sound when it follows the gates in order: safety first, authorization second, finite-versus-open-ended third. ISO 17359 condition-monitoring guidance frames an item in general degradation as a candidate for end-of-life assessment rather than indefinite component-by-component repair, which is the formal basis for stopping and recommending replacement on an open-ended cascade.

A defensible stop is one where the system is safe, the customer is informed in writing, and the remaining work is theirs to authorize.

Next steps

For a finite cascade you finished: verify the full operating path runs clean and document every fault corrected.

For an open-ended cascade you stopped: leave the system safe, hand the customer a written assessment of remaining faults and the repair-versus-replace picture, and get authorization before any further work.

In both cases, keep a running log of faults found and cleared. It is the record that justifies your time, your stopping point, and your recommendation.

References

• ISO 17359, Condition monitoring and diagnostics of machines, end-of-life and degradation assessment.
• OSHA 29 CFR 1910, General Industry standards, not returning hazardous equipment to service.
• ISO 14224, reliability and maintenance data, multiple coexisting failure modes.
• PHCC and ACCA repair-versus-replace assessment guidance for systems in deferred-maintenance decline.