Sudden Break or Long Slow Decline: Decision Tree

Why this matters

Whether a part broke in a moment or died over years changes what you look for next. Sudden means an event to hunt down - a surge, a shock, a blockage. Slow means a condition to correct - wear, neglect, a bad environment. Miss the classification and you replace the part without touching what killed it, and the replacement fails on the same path. This tree gets you to the right mode fast, including the sneaky middle case that looks sudden but was not.

Start here: make it safe, then look for a run-up

If the failed equipment is still a hazard, deal with that first: de-energize and verify dead, relieve pressure, isolate water from live parts, let hot parts cool. Then inspect the failure surface.

One question drives everything below: does the damage show a run-up, or an instant? Get to the surface itself - the fracture, the burn, the worn face - because that is where the answer lives, not in the customer's account.

Step 1: Is there wear leading into the failure

• If the part was clean and sound right up to the break (bright fracture, no groove, no graded wear): lean sudden. Go to Step 2.
• If there is a run-up (a wear groove, polish, taper, layered deposit, or graded heat color building toward the failure): lean slow. Go to Step 3.
• If you see both - progression rings ending in a bright, rough final zone: stop. This is fatigue, slow then sudden. Treat it as a slow failure that ended fast; jump to Step 3 and note the sudden finish. Calling this "sudden" is the classic warranty mistake.

Step 2: Confirm the sudden path - find the trigger

A sudden failure is an event overwhelming a good part, so the real diagnosis is the trigger, not the part.

• Electrical event? Check for a recent surge, outage, or lightning; look at what else on the circuit failed at the same time. See related: Multiple Units Fail After the Same Power Event.
• Mechanical shock or overload? Water hammer, a seized load, a jam, a dropped or struck unit.
• Foreign object or slug of contamination? A single blockage, an object drawn in, a dose of the wrong chemical.
• A protective device that should have caught it? If a fuse, breaker, valve, or limit failed to protect, that is a second fault to address.

Caveat: a "sudden" event can be the final straw on a part already weakened, so glance for hidden prior wear before you close it as purely event-driven.

Step 3: Confirm the slow path - find the condition

A slow failure is a process, so the diagnosis is the condition that drove it.

• Plain wear or age? Compare the wear against expected service life for the hours and conditions.
• Neglect? Missing maintenance, clogged filtration, dry lubrication, packed debris.
• Bad environment? Corrosive air or water, heat, vibration, moisture - the setting is eating the part.
• Misapplication or undersizing? A part working beyond its rating wears fast and looks slow. See related: Undersized vs Failing.

Caveat: a fast-aging environment can make months of wear look like years, so bracket the duration by the conditions, not by the wear alone.

Step 4: Fix the mode, not just the part

• Sudden: replace the part and address the trigger - tame the surge, fix the water hammer, clear the source, restore the protection that failed. Skip this and it recurs on the next event.
• Slow: replace the part and change the driver - the maintenance interval, the environment, the sizing. Skip this and it recurs on the same clock.

Recap

1. Make it safe, then read the failure surface for a run-up versus an instant.
2. No run-up leans sudden; a run-up leans slow; both together is fatigue, slow-then-sudden.
3. Sudden: hunt the trigger and any protection that failed.
4. Slow: hunt the condition - wear, neglect, environment, or misapplication.
5. Repair the mode, not just the broken part, or you will see it again.

References

• Trade-standard failure-analysis practice (fracture and wear signatures)
• Manufacturer documentation on failure modes and protective-device function
• See related: Telling a Sudden Failure from a Slow One; Correlating a Fault with a Recent Power Event; Undersized vs Failing (decision tree)