Why Does My Contactor Weld or Fail to Open Under Load? Causes and Fixes
Why does a contactor weld or fail to open under load? A welded contactor is a switching failure in which excessive arc energy — typically generated during interruption of currents exceeding the device's AC-3 or AC-4 duty rating — fuses the silver-alloy contact faces together, preventing the normally open contacts from separating even when the coil is de-energized. Failure to specify adequate breaking capacity, ignoring inrush multipliers on motor loads, or misapplying an AC-1-rated device on AC-4 duty accelerates contact erosion and raises weld probability with each switching cycle. This guide covers the physics of contact welding, the energy math behind arc formation, AC-1/AC-3/AC-4 duty classification differences, field diagnostic procedures, and contactor selection criteria to prevent recurrence.
Welded contactors are one of the top three causes of unplanned motor control center (MCC) downtime in industrial facilities. We've seen it on conveyor drives in cement plants, on pump starters in water treatment, and on HVAC compressors in data centers. The pattern repeats. The fix is rarely "buy the same part again."
What Actually Happens Inside a Welded Contactor?
Contact welding is a metallurgical event, not an electrical one in the simplistic sense. When the moving contact approaches the fixed contact during closing, or separates during opening, an arc forms across the gap. That arc — typically 4,000 K to 20,000 K at the plasma core — locally melts the silver-cadmium-oxide (AgCdO) or silver-tin-oxide (AgSnO₂) contact tips. If the contacts are still in motion when the molten pool solidifies, you get a microweld. Repeat that thousands of times and the microwelds accumulate into a macroweld that the return spring cannot break.
Two physical mechanisms dominate. First, making welds occur during inrush — typically 6× to 8× the full-load current of an induction motor for the first 50–100 ms. Second, breaking welds occur when the contactor opens against an overload or stalled rotor. Both are governed by the same I²t energy, but they fail differently in the field.
The Role of Bounce
Contact bounce is the silent killer. When a contactor closes, the moving armature impacts the magnet core, and the contacts bounce 1 to 5 times over a period of 0.5–3 ms. Each bounce restrikes an arc. In our experience, contactors fed by undersized control transformers — where the coil pull-in voltage sags below 85% of Uc — show 3× the bounce duration and fail at roughly one-third of their rated electrical life.
Why Does My Contactor Weld or Fail to Open Under Load?
Let's answer the headline question directly. There are seven causes we see in 95% of field failures. Ranked by frequency:
1. Wrong Utilization Category (AC-1 Selected for AC-3 Duty)
This is the most common procurement mistake. A 40 A AC-1 contactor (resistive load, cos φ ≥ 0.95) and a 40 A AC-3 contactor (squirrel-cage motor, breaking running current) look identical in a catalog sidebar but they are not interchangeable. Per IEC 60947-4-1 Annex A, AC-3 testing requires making at 6× Ie at 0.35 lagging power factor and breaking at 1× Ie. AC-4 (plugging, jogging) requires breaking at 6× Ie. If you put an AC-1 device on an AC-3 application, the contacts will weld within hundreds of operations, not the rated million.
2. Coil Voltage Out of Tolerance
IEC 60947-4-1 §7.2.1.2 specifies that AC coils must operate from 85% to 110% of rated control supply voltage Uc. Below 85%, the magnetic pull is insufficient to overcome the spring force quickly enough — the armature "hovers" and contact bounce becomes severe. Above 110%, the coil overheats and shortens life. We've troubleshot a packaging line in Mexico where contactors welded weekly. Root cause: a 230 V coil fed from a 208 V control transformer with a 6% sag during peak hours. The contactor saw 195 V — 85% of rated. Replacing the contactor changed nothing. Adding a 240–208 V buck-boost autotransformer eliminated the failures.
3. Cycling Frequency Above Rated
Mechanical and electrical operating frequencies are specified in IEC 60947-4-1 Table 7. A typical AC-3 contactor is rated for 600–1,200 operations per hour. Exceed that on a chiller short-cycling because of a faulty thermostat, and the contacts cannot dissipate heat between operations. Tip temperature climbs, oxide layers break down, and welding follows.
4. Phase Loss or Unbalanced Voltage
Single-phasing causes the motor to draw 1.7× to 2.0× rated current on the remaining phases. The contactor sees this as a sustained overload. If the upstream overload relay is set loose or has failed, the contactor will eventually weld trying to break that current. We saw this on an irrigation pump in Spain — a blown fuse on L2 caused the contactor on L1 and L3 to carry 240% of Ie for 18 minutes before the operator noticed.
5. Locked Rotor / Stalled Conditions
A jammed conveyor or seized pump bearing presents a locked-rotor current of 6–10× FLA indefinitely. AC-3 contactors are not designed to break this current — that's AC-4 territory. If the overload relay class is wrong (Class 10 on a high-inertia load that needs Class 30), the contactor opens against locked-rotor current and welds.
6. Short-Circuit Withstand Coordination Failure
Per IEC 60947-4-1 §8.2.5.1, contactors are tested under short-circuit conditions in coordination with a Short-Circuit Protective Device (SCPD) — typically a fuse or motor-protection circuit breaker. Type "1" coordination allows contactor damage but no hazard to persons. Type "2" coordination requires the contactor to be reusable after the fault, with only light contact welding permitted that can be separated by a screwdriver. If your fuse is oversized or you've replaced an aM fuse with a gG fuse, the contactor sees more let-through I²t than it was tested for and welds permanently.
7. Environmental Contamination
Conductive dust (carbon, metal swarf), salt fog, hydrogen sulfide (H₂S in wastewater plants), and silicone vapors degrade contact surfaces. Silicone is particularly insidious — it polymerizes in the arc and forms an insulating film that increases contact resistance, which heats the tips, which accelerates oxidation, which causes welding.
The Energy Math: Why Contacts Weld
The fundamental quantity is arc energy. If you understand this formula, you understand why every contactor selection decision matters.
Formula: Arc Energy at Contact Separation — Source: IEC 60947-4-1 Annex K, IEEE Std 1584-2018
Earc = ∫ Uarc(t) · Iarc(t) dt
| Symbol | Description | Unit |
|---|---|---|
| Earc | Total arc energy dissipated at contact tips per opening | J |
| Uarc | Instantaneous arc voltage (typically 15–30 V for AgSnO₂) | V |
| Iarc | Instantaneous arc current | A |
| t | Arc duration (typically 5–15 ms for AC at 50/60 Hz) | s |
For a practical estimate on a 50 Hz AC-3 break at 6× Ie of a 50 A contactor, Earc ≈ 25 V × 300 A × 0.010 s = 75 J per operation. The contact tip mass is about 1.5 g of AgSnO₂ with a specific heat of 234 J/(kg·K) and melting point of 962 °C. A single break does not melt the tip through, but the cumulative effect over 100,000 operations explains erosion.
Diagnosing a Welded Contactor in the Field
When you arrive at a panel with a contactor that won't drop out, here is the procedure we use.
Step 1: Verify the Coil Is De-Energized
Measure across the A1–A2 coil terminals with the control circuit commanded OFF. You should see 0 V. If you see voltage, the problem is upstream — a stuck PLC output, a faulty auxiliary contact, or a wiring error. Don't assume welding until you've confirmed the coil command is removed.
Step 2: Lockout, Then Inspect Contacts
Apply LOTO per NFPA 70E and OSHA 1910.147. Remove the arc chute. If contacts are visibly stuck, attempt to separate them with an insulated tool. If they separate with light force, that's a Type 2 coordination outcome — acceptable for re-test. If they're fused solid, the contactor is scrap.
Step 3: Measure Contact Resistance
With contacts manually held closed, measure pole-to-pole resistance with a 4-wire micro-ohmmeter. New contacts read 50–200 µΩ. Above 1 mΩ indicates erosion; above 5 mΩ means imminent failure regardless of whether the device still operates.
How to Specify a Contactor That Won't Weld
Selection is where most weld failures are prevented. Procurement teams who buy on price alone end up paying 10× in downtime. Here's the framework.
Match the Utilization Category Precisely
For installation contactors used in distribution boards switching resistive heating, lighting, or HVAC compressor loads, AC-7a (non-inductive household-like) and AC-7b (motor loads in household applications) are the relevant categories per IEC 60947-4-3. The ABB 1SBE111111R0611 ESB16-11N-06 is rated 16 A AC-7a / 8.5 A AC-7b at 230 V and is purpose-built for distribution board mounting. For DC-controlled applications where coil response time matters more, the ABB 1SBE111111R0602 ESB16-02N-06 with 2NC contacts and DC coil is the right pick.
Don't Undersize for Inrush
For 400 Hz aviation ground power, naval, or specialized installations, magnetic actuator dynamics differ significantly. The ABB 1SAE231111R0622 ESB25-22N-06 and ABB 1SAE351111R0640 ESB63-40N-06 are 400 Hz-rated installation contactors that handle the higher di/dt without the bounce instability that plagues 50/60 Hz devices misapplied to 400 Hz buses.
Coordinate the SCPD Properly
Get the manufacturer's coordination tables. ABB, Schneider, and Siemens publish them. They list, for each contactor and overload relay pair, the maximum prospective short-circuit current Iq and the required upstream fuse or breaker for Type 1 or Type 2 coordination. Don't guess. A 25 A installation contactor like the ABB 1SAE231111R0631 ESB25-31N-06 has a specific Type 2 fuse maximum — exceed it and welding under fault becomes likely.
Add Residual Current Protection Where Applicable
For installations requiring earth-fault protection, pair the contactor with an upstream RCCB such as the ABB 2CSF202001R1900 F202 AC-100/0.03. This won't prevent welding directly but it limits the consequences of a downstream insulation failure that could otherwise cascade into a phase-to-earth arc that welds the contactor closed.
Comparing AC-1, AC-3, and AC-4 Behavior
| Criteria | AC-1 (Resistive) | AC-3 (Squirrel-cage motor) | AC-4 (Plugging, jogging) |
|---|---|---|---|
| Making current (× Ie) | 1.5× | 6× | 6× |
| Breaking current (× Ie) | 1.5× | 1× | 6× |
| Power factor (test) | 0.95 | 0.35 | 0.35 |
| Typical electrical life | 1–3 million ops | 0.5–1 million ops | 100,000–200,000 ops |
| Application | Heaters, lighting | Pumps, fans, conveyors | Cranes, lifts, reversing drives |
| Derating from AC-1 | Reference | ~60% of Ie(AC-1) | ~25% of Ie(AC-1) |
The derating column is what catches procurement teams off guard. A device labeled "100 A AC-1" may only deliver 60 A of usable AC-3 current. We've audited spec sheets where the buyer compared one vendor's AC-1 rating to another's AC-3 rating and concluded the cheaper unit was equivalent. It wasn't.
Estimating Contactor Life for Your Duty Cycle
Manufacturers publish life curves in operations-versus-current at given voltages. To estimate life for your specific application, use the calculator below. It applies a standard inverse-power-law derating per IEC 60947-4-1 Annex K.
Field Cases: Three Real Failures and Their Fixes
Case 1: Cement Plant Conveyor Drive
A 75 kW conveyor motor in a cement plant in Turkey welded its contactor every 4–6 weeks. The contactor was rated AC-3, 110 A Ie, switching a motor with FLA 145 A. The buyer had matched the contactor to the motor's nameplate FLA without accounting for the conveyor's 18-second start time and high inertia. Solution: upsize to a 185 A AC-3 contactor with Class 30 overload relay. Failures stopped.
Case 2: Data Center Chilled Water Pump
A 30 kW pump cycled 8–12 times per hour because of a hunting pressure controller. The contactor's rated operating frequency was 600 ops/h but actual cycling was within spec — yet contacts welded after 4 months. Investigation revealed the issue wasn't cycling rate but that each start drew 7.2× FLA against a closed discharge valve (water hammer prevention). The system was effectively AC-4 duty. Solution: replace with AC-4 rated contactor and reprogram VFD soft-start.
Case 3: 400 Hz Aviation Ground Power Unit
An airport ground power unit serving 400 Hz aircraft buses experienced contactor weld failures within months of commissioning. The original specification used 50/60 Hz contactors. At 400 Hz, the AC magnetic circuit losses and arc dynamics differ substantially. Solution: replace with 400 Hz-specific units like the ABB 1SAE231111R0640 ESB25-40N-06 designed for 400 Hz operation.
Preventive Maintenance That Actually Works
Most maintenance programs over-specify visual inspection and under-specify measurement. Here's what we recommend based on IEEE Std 3007.2 and field data:
Annually, on critical drives: measure contact resistance under load, check coil pull-in and drop-out voltages against the 85%–110% Uc band, verify auxiliary contact integrity, inspect arc chutes for cracks, and torque check the power terminals to manufacturer spec (typically 2.5–6 Nm for installation contactors, up to 25 Nm for larger frames).
Every five years on duty cycles approaching 50% of rated life: replace proactively. The cost of a planned replacement is 5–10% of an unplanned failure including downtime.
Ready to Source Contactor?
- Browse in-stock contactor units
- Request a custom quote — response within 4 hours
- Talk to an engineer
Frequently Asked Questions
Can I free a welded contactor by tapping it with a hammer?
Sometimes you can mechanically separate lightly welded contacts, but this is a temporary measure only. The contact surfaces are now severely degraded and contact resistance has risen, which means the next operation generates more heat and welding will recur quickly. Replace the contactor at the earliest opportunity.
Why does my contactor hum but not pull in?
This usually indicates low coil voltage (below 85% Uc), a damaged shading ring on the magnet pole face, or mechanical binding. The armature is partially attracted but cannot fully seat. Operating in this state for more than a few seconds will burn the coil and damage the magnetic circuit. Measure coil voltage at A1–A2 with the contactor commanded ON and confirm it sits between 0.85·Uc and 1.10·Uc.
Is it safe to replace a welded contactor with a higher-amperage model?
Only if the upstream short-circuit protective device coordination still holds and the auxiliary contact arrangement matches the control circuit. Going from a 40 A to a 65 A frame may invalidate the Type 2 coordination tested with the original SCPD, and the larger coil may have different inrush characteristics that overload the control transformer. Always check the manufacturer's coordination tables before substituting.
What's the difference between a contactor welding and an overload relay failing to trip?
They're related failure modes. The overload relay is supposed to detect sustained overcurrent and signal the contactor to open before damage occurs. If the relay fails to trip — wrong class, wrong setting, or thermal element fatigue — the contactor sees the overload longer than it was designed for and may weld. In any weld investigation, test the overload relay's trip function and verify class (10, 20, or 30) matches the load's starting time.
Do solid-state contactors eliminate welding?
Yes, because there are no mechanical contacts to weld. Solid-state contactors using SCRs or IGBTs have different failure modes — primarily thermal runaway and shorted junctions — and require heat-sinking and snubber design discipline. They cost 3–5× more than electromechanical equivalents and have higher conduction losses, so they're typically reserved for very high cycling rates (above 3,600 ops/h) or vibration-critical applications. For standard motor starting at 60–600 ops/h, a properly specified electromechanical contactor remains more cost-effective.
How do I know if my application is AC-3 or AC-4?
If the motor is allowed to coast to a stop before the contactor opens, it's AC-3. If the contactor must break the motor while it is still rotating at full speed against a load — for example reversing without first stopping, jogging for positioning, or plugging — it's AC-4. AC-4 duty requires breaking 6× Ie at 0.35 power factor and demands either an oversized AC-3 contactor (typically 2× the AC-3 rating) or a purpose-built AC-4 device.
Can vibration cause contactor welding?
Indirectly, yes. Severe vibration can cause contact chatter — momentary openings during normal carry — which generates micro-arcs and progressive erosion. Marine, rail, and mining applications often require contactors with mechanical latching or vibration-rated designs (per IEC 60068-2-6). If you're seeing welding on an installation with known vibration sources, check whether the contactor is rated for the vibration spectrum present.
Conclusion: Specification Discipline Prevents Welding
Contactor welding is not a random failure. It is the predictable consequence of a mismatch between the device's tested capability and the actual application duty. Every welded contactor we've investigated traced back to one of seven causes: wrong utilization category, coil voltage out of tolerance, excessive cycling, phase loss, locked-rotor breaking, SCPD miscoordination, or environmental contamination. None of these are mysterious. All of them are preventable with disciplined specification and basic preventive maintenance.
For procurement managers, the lesson is to stop comparing contactors on price-per-amp. The headline ampere rating means little without the utilization category, the rated operational voltage, the coordination class with upstream protection, and the expected mechanical and electrical operating life under your actual duty cycle. Two devices with identical AC-1 ratings can have AC-3 ratings that differ by 40%, and life expectancies that differ by an order of magnitude.
For engineers in the field, the lesson is to treat rising contact resistance as the leading indicator. Annual micro-ohm measurements on critical drives cost very little and predict welding months before it happens. Combine that with coil voltage verification and SCPD coordination audits and you eliminate the vast majority of contactor failures before they take a production line down.
When you do specify a replacement, match the duty: AC-7a/AC-7b installation contactors like the ESB16, ESB25, and ESB63 series for distribution board applications, AC-3 motor contactors for pumps and fans, AC-4 devices for plugging and jogging, and 400 Hz-rated devices for aviation and naval buses. Coordinate the SCPD per IEC 60947-4-1 Type 2 wherever the application allows reuse after fault. Verify coil supply stability. Document the duty cycle. Then the contactor will deliver the million operations the manufacturer promised — not weld at ten thousand.
