9.1 Contact Resistance (Ductor) & Torque Verification

Why Contact Resistance Matters

A bolted joint or a breaker main contact that looks fine to the eye can still carry a high-resistance film, pitting, or insufficient contact pressure. Under load that joint heats (I-squared-R), accelerates oxidation, and can cascade into a failure or a fire during a fault. The contact resistance test — universally called the Ductor or DLRO (Digital Low-Resistance Ohmmeter) test — is the standard NETA MTS method for catching those joints before they fail.

How the Ductor Test Works

A micro-ohmmeter injects a known, regulated DC current (typically 100 A on power circuit breakers and bus joints, lower on control contacts) through the closed current path and measures the voltage drop across the joint. Ohm's law gives the resistance directly in micro-ohms (µΩ). Using DC (not AC) and a four-wire (Kelvin) connection cancels lead and contact resistance, so the reading reflects only the joint under test.

On a three-pole breaker you take a reading on each pole — line-to-load through the closed contacts — and compare phase-to-phase. You also compare to the manufacturer's published value or, on maintenance work, to the baseline from prior tests.

Acceptance Criteria

NETA MTS does not publish a single pass/fail micro-ohm number because resistance scales with the breaker's continuous-current rating (a 100 A frame legitimately reads higher than a 4000 A frame). The acceptance rule is comparative:

A useful rule of thumb on low-voltage power breakers is that <= 100 µΩ is good and > 300 µΩ indicates trouble. On medium-voltage air-magnetic or vacuum breakers the structures are larger and the same comparative rule applies — look at the spread and the trend, not an absolute number.

Torque Specifications

Loose joints are the #1 preventable cause of switchgear failure, and "loose" almost always means under-torqued. NETA's rule is explicit:

1. Torque per the manufacturer's published instructions first. If the OEM gives a value for that bolt size, material, and lubricant, use it.
2. Where no manufacturer data exists, use the torque table in the ANSI/NETA ATS/MTS Annex, indexed by bolt diameter and material (for example, 1/2 in. silicon-bronze vs. 3/8 in. stainless).
3. Use a calibrated torque wrench — and verify calibration is in date. Torque is a controlled, measured preload, not "good and tight."

Over-torquing is just as bad as under-torquing: it yields the fastener, crushes bus plating, and leaves the joint with less spring force than before.

Torque Verification Method

After a bolted connection is made (or on an existing joint during maintenance), NETA accepts two verification paths:

• Re-torque to spec with a calibrated torque wrench and confirm the bolt does not move when the wrench reaches the specified value. Movement means the joint was loose — disassemble, inspect the faces, and reassemble.
• Mark-and-recheck: apply a torque-seal / witness mark across the bolt and bus after torquing. On the next maintenance cycle, a broken mark means the joint has loosened (thermal cycling, vibration) and needs re-torque.

A common field trap is reaching for an impact gun. An impact gun cannot produce a known torque — it depends on bolt friction, joint stiffness, and run time — and the hammering can damage plating and threads. NETA considers impact tightening unacceptable for electrical bolted joints.

Why Both Tests Are Required Together

A joint can be torqued correctly and still read high on the Ductor if the contact faces are oxidized, pitted, or contaminated — torque only provides force, not a clean conductive interface. Conversely, a joint can have clean faces and still run hot if the clamping force is low. Running both tests gives you joint integrity: the torque wrench confirms preload, and the Ductor confirms conductive area. Skipping either leaves a gap.

Common Ductor Field Errors

• Using a 2-wire ohmmeter — lead and clip resistance (often tens of milliohms) swamps the joint reading. Always use a 4-wire Kelvin DLRO.
• Measuring across the wrong span — for a breaker pole the leads go line-to-load through the closed contacts; for a bus joint they go directly across the joint, not the whole bus.
• Ignoring polarity/current direction — on a DC test, thermoelectric EMF at dissimilar-metal junctions can offset the reading slightly; many DLROs auto-reverse and average.
• Not letting the reading settle — inductive joints (large reactor or transformer leads) need a few seconds for the test current to ramp; record the steady-state value.
• Forgetting to record the baseline — without a baseline, a rising trend across years is invisible; always log pole readings and temperature.

Worked Example

A 1600 A LV air circuit breaker is tested after maintenance. Pole A reads 42 µΩ, Pole B reads 38 µΩ, Pole C reads 91 µΩ. The spread (91 vs. 38) is well past the 50% comparative threshold. Before condemning the breaker, re-torque the line/load stabs per the OEM value (say, 45 lb-ft on 1/2 in. hardware). Pole C drops to 44 µΩ — the stab was loose. Re-test confirms the joint is now balanced and the breaker is returned to service.