# A Busbar Connection Can Pass Its Torque Check and Still Be Quietly Cooking Itself Loose Author: Nitin Verma Author URL: https://insights.spans.co.in/author/nitin-verma Published: 2026-11-08 Meta Title: Electrical Connection Torque Check vs Thermal Creep | Spans Meta Description: A torque check on a busbar or cable lug connection confirms the clamping force at the moment it was measured. It does not confirm the joint stays tight under load, because thermal cycling and material creep loosen connections by a mechanism no torque wrench can detect. URL: https://insights.spans.co.in/busbar-torque-thermal-creep-wastewater Every MCC, switchgear busbar, and major cable lug at an [ETP](https://spans.co.in/effluent-treatment-plant-etp/) or [STP](https://spans.co.in/sewage-treatment-plant-stp/) gets its bolted connections checked against a specified torque value, confirming the fastener has been tightened to the clamping force the manufacturer calls for. That check is real and necessary: it catches a connection that was never tightened properly in the first place. It does not confirm the connection stays at that clamping force over months of service, because the mechanism that loosens a bolted electrical joint over time is not someone forgetting to tighten it, it is thermal cycling and material creep, a slow, load-driven relaxation that a torque wrench reading on day one has no way to anticipate. ![A torque check log shows every joint passing, connected to a four-step cycle diagram: creep relaxes clamp force, contact resistance rises, the joint runs hotter under load, and heat accelerates further creep](https://spans.co.in/images/blog/busbar-torque-thermal-creep-inline.png) ## Clamping force today and contact integrity over time are different claims A torque check answers: is this fastener tightened to the specified value right now. It is a force measurement, confirming the mechanical clamping that holds two conductors in contact. It says nothing about contact resistance, the electrical property that actually determines whether the joint runs hot under load, because a joint can be correctly torqued and still carry higher resistance than design intends, or can start at acceptable resistance and drift upward over time even though nobody has touched the bolt since installation. [Creep](https://en.wikipedia.org/wiki/Creep_%28deformation%29), the slow deformation of a material under sustained stress, acts on the conductors themselves, aluminum more than copper but both to some degree, gradually relaxing the clamping force a correctly torqued bolt originally provided. The bolt has not backed off. The material it is clamping has slowly deformed underneath it, and a torque check performed after that relaxation would likely still read as acceptable, because the check measures the bolt's resistance to turning, not the actual contact pressure between the conductors it is holding together. ## Why a passing torque record and a cooking connection can coexist Industry guidance on bolted electrical connections generally combines several checks precisely because no single one covers the whole failure progression: visual inspection, resistance measurement, torque checks, and infrared thermal imaging under load. A torque check alone misses the resistance and heat side of the picture entirely. [Thermographic inspection](https://en.wikipedia.org/wiki/Thermographic_inspection) catches a hot joint, but only once resistance has already risen enough to generate detectable heat, and only if the connection happens to be carrying load at the moment of the survey. In between a correctly torqued connection and a thermally obvious hot spot sits a gap where creep has already begun relaxing the joint, resistance has begun climbing, and neither the maintenance log's last torque check nor a thermal survey scheduled for a different quarter has any way to show it. The physics make this worse once it starts, not better. As a joint relaxes and resistance rises, the connection runs hotter under the same load. That added heat accelerates the very creep that caused the resistance to rise in the first place, a self-reinforcing cycle that, left alone, moves a connection from slightly relaxed to genuinely dangerous over a timescale that has nothing to do with the maintenance calendar. A connection sitting at the early stage of that cycle shows a clean torque check, no visible damage, and no thermal signature yet, because the cycle has not run long enough to produce one. This is the same shape of mistake we described in [the piece on anchor bolt vibration loosening](https://insights.spans.co.in/anchor-bolt-vibration-loosening-pump-foundation-wastewater): a torque check confirms clamping force at the moment it is performed, and a completely different physical mechanism, vibration there, thermal creep here, can undo that clamping force afterward without ever registering on a repeat of the same check unless someone goes looking for the specific symptom that mechanism actually produces. It is also close kin to [the piece on earth pit resistance versus touch and step voltage](https://insights.spans.co.in/earth-pit-resistance-touch-step-voltage-wastewater): both involve an electrical verification that is correct and necessary for its own narrow purpose while leaving a separate, more consequential condition, actual fault safety there, actual contact integrity under load here, to be caught only by a different kind of check that most maintenance programs never schedule with the same discipline. Picture an MCC feeding a bank of blower motors, commissioned with every busbar and cable lug torqued to specification and logged as compliant on the commissioning checklist. Eighteen months of continuous load cycling later, nobody has re-torqued or re-inspected those connections, because the torque check passed and nothing on the maintenance calendar calls for revisiting a connection that already passed. One aluminum-lugged joint, carrying a consistently heavy load, has been creeping under sustained clamping stress the entire time, its contact resistance climbing slowly enough that no single week looks alarming. The joint is now measurably hotter than its neighbours under full load, a difference a thermal camera would catch in minutes, but nobody has run one, because the panel's electrical connections are considered a solved problem once the torque sheet comes back clean. Left on its current trajectory, the self-reinforcing cycle of resistance and heat means this connection does not improve on its own; it only gets worse, and the first hard evidence anyone sees may be a nuisance trip, a discoloured terminal, or in the worst case a connection failure at exactly the moment the load it is carrying matters most. ## What actually confirming connection integrity requires The fix is not abandoning the torque check, which remains the right way to catch an improperly installed connection. It is treating thermal imaging under load, and ideally a periodic contact resistance or millivolt-drop measurement, as a separate and equally necessary check, scheduled on its own cycle rather than assumed to be covered because the torque record looks clean. For dissimilar-metal joints, aluminum lugs on copper busbar being the classic case at many plants, a Belleville spring washer at installation maintains preload through thermal cycling and creep in a way a solid washer cannot, closing off the mechanism before it starts rather than relying on a later inspection to catch it once it has. None of this needs to be an elaborate programme to be effective. An annual thermal survey of MCC busbars and major cable terminations under normal operating load, timed to actually catch the plant at or near its typical duty point rather than during a shutdown, finds the early-stage hot spots this failure mode produces well before they become urgent. Pairing that survey with a periodic resistance or millivolt-drop spot check on the handful of highest-load connections, the ones most exposed to the resistance-and-heat cycle in the first place, costs far less than the [downtime](https://spans.co.in/wastewater-treatment-cost/) a connection failure at a critical MCC would cause, and it catches exactly the condition a torque check was never built to see. If your plant's electrical connections are verified by torque checks alone, with no periodic thermal survey under actual load, that gap is worth closing with [Spans' electrical engineering team](https://spans.co.in/contact/) before a connection that passed every torque check on file quietly cooks itself into a failure nobody saw developing. --- This blog is powered by Superblog. Visit https://superblog.ai to know more. ---