A termination is where a conductor stops being a conductor and becomes a joint, and joints are where electrical systems fail. When one fails the question is almost always attribution: did an installer make it badly, did a product fail to perform as designed, or was the failure set in motion by a materials decision made long before either of them arrived.
Push-in versus screw terminations
Push-in or back-wired terminations rely on a spring member maintaining contact force on a conductor. Screw terminations rely on a clamped joint held by applied torque. The two age differently and fail differently, and both have well-documented histories.
A spring member that relaxes, or that was engaging a conductor of the wrong diameter, loses contact force gradually. A screw termination that was under-torqued starts with insufficient force; one that was over-torqued may have deformed or nicked the conductor, reducing cross-section and creating a stress concentration. Each of those leaves distinguishable physical evidence at the joint.
Torque is a recoverable fact
Terminations are specified with a torque value, printed on the device or in its instructions, and that value is a requirement rather than guidance. Whether it was achieved is partly recoverable after the fact from thread and seat deformation, witness marks, and the conductor's own impression.
Re-torquing practice matters too. Some terminations are specified for a check after initial thermal cycling, and in a facility with a maintenance programme the absence of any record is a finding. Marking a terminal after torquing is cheap, common practice and makes the question answerable later.
Dissimilar metals
Aluminium-to-copper joints are the classic case and the reason a great deal of prescriptive practice exists. The two metals differ in thermal expansion and hardness, aluminium creeps under sustained clamping load, and its oxide is insulating rather than conductive. A joint that was sound on installation can loosen over years of thermal cycling with nobody touching it.
The mitigations are specific: connectors listed for the combination, appropriate compounds, correct preparation of the conductor surface. Their presence or absence is observable at the joint, and their absence is a straightforward finding where an aluminium conductor was landed on a device listed only for copper.
Reading the joint
The physical record at a failed termination is usually informative. Oxide and heat discolouration graded by distance from the contact point, annealing of the conductor, arc pitting on the mating surfaces, and displacement or embrittlement of insulation all record a history.
The distinction worth drawing early is between damage that developed over time and damage that occurred in the terminal event. A joint that degraded over months carries a gradient; one that was destroyed by an external fire tends to carry uniform damage consistent with its surroundings rather than centred on the contact.
Why protection never operates
A degrading termination generates heat at normal current, which is the reason these failures are dangerous. Overcurrent devices protect against current beyond a threshold, and a bad joint does not create one. It converts a normal load into localised heating at a small area, and nothing in a conventional protection scheme is watching for that.
Anyone reasoning that the breaker did not trip so the connection was fine has the logic inverted. That is the expected behaviour, not evidence of a sound joint.
Product history is public
Terminations and the devices carrying them have been the subject of recalls, listing withdrawals and standards revisions. The published record is a legitimate investigative resource: a known issue with a device family is not proof about an individual unit, but it identifies what to examine and whether the condition was known to the manufacturer.
Listing status is a related question with a documentary answer. Whether the device was listed for the conductor material, the conductor size and the environment it was installed in is checkable, and a mismatch shifts the analysis.
Thermography would have found it
A degrading termination announces itself thermally long before it fails, which is why infrared survey is standard practice in industrial and commercial maintenance programmes. A joint running tens of degrees above its neighbours under comparable load is a finding, and it is findable with a handheld camera in minutes.
That makes the maintenance record relevant in both directions. Where surveys were performed and the joint was flagged and not repaired, the record establishes knowledge. Where no programme existed on equipment where one is customary, the omission is itself the subject. Either way the documents usually exist, because thermography produces images with timestamps.
Sequencing the examination
Terminations are fragile evidence and easily destroyed by the act of removal. The productive order is to document the joint in place, record conductor routing and any witness marks, radiograph enclosed assemblies before opening them, and only then disassemble under a protocol — measuring what force was required to loosen a screw, if that is still meaningful.
Where several parties have an interest, agreeing that protocol in advance is far easier than arguing about a teardown one party performed alone. A termination can be examined properly once.
This article is general technical orientation, not a failure analysis, an engineering opinion, or advice on any specific matter. Determining the cause of a particular incident requires hands-on examination by a credentialed expert.