The Breaker That Never Tripped

Why this case is worth reading

A melted branch-circuit cable behind a breaker that is still sitting in the ON position is one of two things: a heat source the breaker was never designed to see, or a breaker that did not do its job. The first is common and the breaker is innocent. The second is rare, and when it happens the cause is often a part that came through a channel nobody could answer for. This case is told backwards from the shop's own job record, because by the time anyone was thinking about it, the damage was done and the paper was all there was. The gaps in that paper turned out to matter as much as the entries.

First, what the shop made safe

The call came in on a January evening: a burning smell in an upstairs bedroom, no smoke visible, the heaters "acting funny". The dispatcher told the homeowner to unplug everything in that room, leave the bedroom, and call 911 if she saw smoke, felt a warm wall or the smell got stronger. None of that happened, and a tech was there in 40 minutes.

On arrival the tech did three things before any diagnosis. He opened the panel cover and switched off the bedroom circuit's breaker, then proved the circuit dead at the bedroom receptacles with a tester checked on a known live source before and after, per NFPA 70E Article 120 (120.5 in the 2021 edition, renumbered in 2024), with work practices at 29 CFR 1910.333(b)(2). He scanned the bedroom walls and the ceiling below with a thermal camera for any spot warmer than the surrounding wall; there was none, so the fire department was not called. And he told the homeowner the circuit stays off and nobody resets that breaker until the shop says otherwise, then taped the handle in the OFF position and wrote "do not reset, see tech" on the tape.

Everything after that was diagnosis on a dead circuit.

The record, as it was written

The shop's job record for the two visits, condensed:

JOB 24-0119  bedroom 2 circuit, burning smell
01-14 19:52  CALL  smell, 2 space heaters on 1 strip
01-14 20:35  ARR   brkr 15A pos 12 found ON, not tripped
01-14 20:41  SAFE  brkr off, proved dead, IR scan clear
01-14 21:20  FIND  rcpt 3: cord end melted at strip
01-14 21:45  FIND  NM 14/2 browned full length,
                   worst at top plate bundle
01-14 22:05  RMVD  brkr pos 12, bagged, tag E-112
             (no photo in panel before removal)
01-16 09:10  BENCH E-112 weight 3.1 oz,
                   panel-line sample 4.5 oz
01-16 10:30  CUST  order confirmation, 2 x 15A,
                   bought by prior handyman, 3 yr ago
01-17 14:00  FIND  pos 9 brkr matches E-112 look

Each line is a step in the reasoning. What the record does not say comes up at the end.

What the cable said about current

The first question on a melted cable is whether too much current flowed in the whole circuit or too much heat was made at one point. They leave different marks, and they point at different culprits.

• A loose or failed termination makes heat at one spot: a backstabbed receptacle, a loose screw, a corroded splice. The cable within a few inches of that spot cooks and the rest of the run is fine. A breaker, even a perfect one, does not trip on this, because the current is normal; the heat comes from resistance at the joint. If that is what you find, the breaker is not your suspect.
• An overload heats the conductor along its whole length. The insulation discolors evenly over the run and fails first where heat cannot escape: cable bundled with other cables, packed in insulation, or pulled through a tight bored hole.

The 21:45 entry is the one that pointed at the breaker. The 14/2 nonmetallic cable was browned along its full 22 ft run, and worst where it passed through a top plate in a bundle of three cables with insulation packed around them. The receptacle terminations were tight and clean. The damage at the power strip's cord end was the strip being overloaded too, not the cause of the cable damage. So this was current, all the way back to the panel, and current is exactly what a breaker is there to stop.

How much current: two space heaters, each labeled 1,500 W at 120 V, draw 1,500 divided by 120, 12.5 A each, so 25 A together on a 15 A circuit. That is 25 divided by 15, about 167 percent of the breaker's rating, for hours each night. UL 489, the standard molded-case breakers are listed to, sets maximum trip times at 135 and 200 percent of rating, and 167 percent sits between the two, so a genuine 15 A breaker would have opened that circuit on the first cold night, typically within minutes and long before the insulation browned. This one had stayed closed through what the homeowner said was most of three winters.

The breaker on the bench

Two days later the bagged breaker, tag E-112, went on the bench beside a breaker of the same panel line from the shop's approved supplier.

• Weight. 3.1 oz against 4.5 oz on the shop's postal scale. A thermal-magnetic breaker carries a bimetal strip, a magnetic trip assembly and copper and silver-alloy contacts, and those are the parts a counterfeiter has a reason to skimp on, so a breaker about a third lighter than its genuine counterpart is a strong flag. (4.5 minus 3.1 is 1.4 oz, which is 31 percent of 4.5.)
• Markings. The rating and catalog number were in a slightly different typeface, the plastic was a shade off, and the date code position a genuine breaker of that line carries was blank.
• The listing mark. Present, and its file number checked out in the lab's online directory. This is the trap: counterfeiters copy real file numbers from real products. A directory hit tells you the number exists, not that this object earned it. See related: How to Spot a Counterfeit Electrical Component, which owns the full inspection.

Nobody opened it. Cracking the case would have destroyed the evidence the brand owner and any insurer would want, and the shop had what it needed from the outside.

The purchase nobody at the shop made

The 10:30 entry came from the homeowner's own files. Three years earlier a handyman had been called because the bedroom breaker "kept tripping" in winter. He replaced it. The homeowner still had his order confirmation from an online marketplace: two 15 A breakers of the panel's line, at about a third of what the shop's supplier charges for one.

Read that against the physics and the story closes. The original breaker was tripping because two heaters on one circuit overloaded it, which is the breaker working. The handyman treated the trip as the fault, replaced the protective device, and the replacement did not protect. The second breaker from the same order was still unaccounted for, which is why the 14:00 entry exists: the tech went back through the panel looking for a match and found it at position 9, the bathroom circuit, by the same typeface and the blank date code. It was replaced the same visit, with the panel de-energized and proved dead before and after, and on restoring the tech stood to the side of the panel as the main closed. See related: Counterfeit Breakers and Why They Look Right.

What the shop told the customer

The conversation was plain and in this order. The bedroom circuit had been overloaded for years by two heaters on one circuit. The breaker that should have stopped it was not a genuine part and had been installed by someone else. The damaged 22 ft cable run would be replaced, the two suspect breakers already had been, and the heaters needed either separate circuits or one heater per room. The shop could vouch for the work it did and the parts it installed, not for the rest of the house's wiring, and recommended an inspection if she wanted that confidence.

The shop also told her to keep the bagged breaker, the bagged second breaker and a section of the damaged cable, labeled, because her insurer might want them, and that whether to file a claim was her decision. See related: How to Keep a Failed Part as Evidence.

What the shop reported

The owner filed a report on the U.S. Consumer Product Safety Commission's public product-safety site describing a counterfeit breaker that failed to trip under overload in a home, with the photos and weights. He sent the brand owner named on the label the same evidence and an offer to ship the parts, and the brand owner later confirmed in writing that E-112 was not its product. He reported the listing to the marketplace through its seller-report process. All three confirmations went into the job file.

Where the record was thin

The job record carried the case. Its gaps were instructive:

• No photo of the breaker in the panel before removal. The 22:05 line says so. A photo showing E-112 at position 12, ON, beside its neighbors, is the one image an insurer or a manufacturer asks for first, and it can never be taken again. The shop's evidence procedure now makes it the step before any removal.
• No record of what the handyman took out. The original breaker, the one that had been working, was gone. Its weight and markings would have been the cleanest comparison of all.
• The handyman's job left no record in the house at all, except the order confirmation the homeowner happened to keep. A swap of a protective device with no permit, no invoice naming the part and no source is exactly the gap that let a counterfeit sit in a panel for three winters. Every breaker the shop installs now goes into an installed-part record with its source and lot; see related: The Installed-Part Record SOP.

References

• UL 489, Molded-Case Circuit Breakers (the trip-time performance genuine breakers are listed to)
• NFPA 70E Article 120 (120.5 in the 2021 edition, renumbered in 2024), and OSHA 29 CFR 1910.333(b)(2), for proving the circuit dead
• U.S. Consumer Product Safety Commission public product-safety reporting site
• See related: Counterfeit Breakers and Why They Look Right
• See related: How to Keep a Failed Part as Evidence