The Fault That a Recent Repair Created
Why this matters
Some faults are not aging, not wear, and not the customer's doing. They were created by the last repair, sometimes yours. The act of fixing one thing disturbs, pinches, over-tightens, or leaves defeated another thing that was fine, and a brand-new symptom shows up within hours or days of the visit. Medicine has a word for a problem caused by the treatment, iatrogenic, and the trades have the same class of fault. Catching it before you leave saves a callback and your name. Recognizing it when someone else caused it saves you from chasing wear that is not there.
What a repair-induced fault is
It is a new defect introduced by the work itself, not the original complaint and not an unrelated coincidence. The tell is timing plus proximity: a symptom that did not exist before the visit, on or next to something the repair touched. That is different from a second unrelated part failing near the same time, which is its own attribution problem covered in the related article. Here the repair is the mechanism, not a coincidence of dates.
The common mechanisms
| Mechanism | What it looks like later |
|---|---|
| Over-torque or over-tighten | Crushed gasket, stripped thread, cracked fitting, slow weep |
| Pinch or chafe on reassembly | A wire, line, or hose caught against an edge, failing over days |
| Disturbed adjacent connection | A neighbor loosened or reseated while you worked, now intermittent |
| Interlock or guard defeated and not restored | A safety left bypassed after testing |
| Contamination introduced | Debris, sealant, flux, or air left in a line or contact |
| Wrong setting left behind | A valve part-closed, a switch flipped, a dial off its mark |
| A marginal neighbor bumped | A part one nudge from failing, pushed over by handling |
None of these are exotic. They are the ordinary byproducts of getting your hands into a system, which is exactly why disciplined techs check for them every time.
The timing and location tell
A repair-induced fault reads differently from wear. Wear builds slowly and announces itself; an induced fault appears abruptly and sits right where the work was. When a customer says the new problem started the day after service and it is on the same circuit, line, or assembly you opened, treat your own visit as the prime suspect until you clear it. Ego wants it to be a coincidence. The location usually says otherwise.
The pre-departure audit
The cheapest place to catch this is before you pack up. Run a short, deliberate audit:
- Safety first: restore every interlock, guard, cover, and safety you defeated to test, and verify it is back in service before anything else. A bypass left in place is the most dangerous induced fault there is.
- Re-check the torque and seat of every fitting and connection you made, and every one you leaned on to reach the work.
- Trace every wire, line, and hose you moved for pinch, chafe, or a route against a hot or sharp edge.
- Confirm every valve, switch, damper, and setting is back where normal operation needs it.
- Run the system through a full cycle and read it, do not just watch it start. Many induced faults only show once load, heat, or pressure returns.
When the callback comes anyway
If a new symptom lands right after your visit, check your own work the way you would check a stranger's, honestly and first. Verify what you touched is right, then look one ring outward at what you disturbed to get there. If you find your own pinch, crush, or bump, own it and make it right on your dime; that is the deal. If your work and everything you disturbed check clean, then you are into genuine second-fault attribution, and you hand it to the concurrent-fault method rather than tearing back into good work.
The discipline that prevents it
The techs who rarely cause these are not luckier, they are more deliberate: they note what they defeat so they restore it, torque to feel and spec instead of gorilla-tight, dress wires clear of edges, and cycle-test before they leave. Slowing the last five minutes down is far cheaper than a return trip and the credibility hit that rides with it.
References
- Trade-standard practice for post-repair verification and lockout restoration
- Manufacturer torque, routing, and reassembly specifications
- See related: Separating a Failed Repair From an Unrelated New Problem; The Signature of a Mismatched Replacement Part