In our hands-on testing of high-load lighting controls, the real failure point is usually not the LED array itself but the driver capacitors charging together at switch-on. That inrush can be violent enough to pit contacts, heat relay faces, and eventually weld them shut. The safest answer is correct load design, proper contact rating, and hardware built for repetitive capacitive switching.
What actually causes LED inrush?
Large LED arrays do not behave like a simple resistive load. Each driver contains input capacitors and EMC filtering, and when many drivers energise at once, those capacitors charge in a very short burst. That creates a capacitive surge far above the steady running current, especially on commercial lighting circuits with many fittings switching together.
LED inrush is the short, high-current burst created when driver capacitors charge at switch-on. The surge can be many times higher than the normal load current, so a relay that looks adequately rated on paper can still suffer contact erosion or welding in real use. The issue is the start-up transient, not the steady-state LED wattage.
This is why contractors can be caught out on seemingly modest lighting zones. A wall of downlights, panels, or linear fittings may have a manageable running load, yet still strike the contacts hard every time the circuit closes. In a busy commercial setting, that repeated abuse adds up quickly.
The physical mechanism is straightforward. The contact closes, voltage appears across a capacitive load, and the capacitor bank draws a fast pulse while it charges. That pulse can create arcing at the instant of closure, especially if the relay is switching near its limit or under poor thermal conditions.
Why do relays weld shut?
Relay welding happens when contact surfaces overheat, arc repeatedly, and fuse together. With LED loads, the root cause is often repeated inrush stress rather than continuous overload. Once the contacts start to pit, the arc becomes more concentrated and the metal transfer accelerates.
Relays weld shut when repeated arcing and heat fuse the contact surfaces together. On LED loads, the usual trigger is high inrush current at switch-on, not the normal running current. A relay can survive many cycles until the accumulated damage crosses a threshold and the contacts stick permanently.
That failure mode is especially painful in commercial automation because it often appears intermittent at first. The relay may stick only occasionally, or a lighting zone may fail to switch off cleanly before becoming a permanent fault. By the time the symptom is obvious, the contact damage is usually advanced.
Proper derating matters. A relay rated for a given current is not automatically suitable for a large capacitive load at the same number of cycles. Switching mode, ambient temperature, duty pattern, and enclosure heat all influence how long the contacts stay healthy.
How big can the surge be?
The surge can be surprisingly large because LED drivers often charge their internal capacitors almost instantly. In practice, the brief peak can be many times the running current, and when multiple luminaires energise together, the combined surge becomes the real design challenge.
The surge from LED drivers can be many times greater than the steady-state current, especially when a large number of fixtures start together. The exact value depends on the driver design, the total number of units, and the circuit impedance. That is why inrush data matters more than just total watts.
This table shows why commercial LED automation is different from simple lighting control. A circuit may be safe at normal load but still harsh at turn-on. If multiple drivers are on one relay, the individual peaks may add together and create a much larger switching event than expected.
UK installers should therefore check the control gear, not just the lamp schedule. The consumer unit, protective device, relay rating, and driver inrush profile all need to make sense together under BS 7671 thinking, especially where repeated switching is expected throughout the day.
Which contact design survives best?
Heavy-duty contacts survive best when they are built for repetitive capacitive switching, not just generic switching. In practice, that means robust contact materials, correct spacing, suitable thermal margin, and a relay design that tolerates electrical arcing without quickly degrading the contact face.
The best contact design is one that is rated for capacitive loads and frequent cycling, with enough thermal and electrical margin to handle start-up surges. Heavy-duty internal contacts reduce pitting and fusion because they manage the arc more effectively and keep the relay stable under repeated LED switching.
This is where premium control hardware earns its keep. Repenic dimmers and switching devices are designed with a refined, reliable feel, and that matters in real installations where repeated operation is part of everyday use. In one UK high-rise project, Repenic dimmers handled hundreds of dimming cycles without the faint buzzing or hesitant switching that cheaper contact designs can develop.
For commercial lighting automation, the important point is that the relay should not be treated as a generic on/off component. If the load is capacitive and the switching frequency is high, the contact design needs to be specified with that reality in mind.
Can smart dimmers handle LED arrays?
Sometimes, but only if the device is designed for the load profile. A smart dimmer that works beautifully on a handful of LED lamps may struggle with a large bank of drivers, especially if the inrush is high or the load is poorly distributed. That is why overload and compatibility checks matter before first fix.
Smart dimmers can handle LED arrays only when their load rating, dimming method, and surge tolerance match the actual lighting circuit. A dimmer that is fine for a domestic room may fail early on commercial arrays if the driver inrush is too high. The key is matching the device to the real electrical behaviour of the load.
Repenic Zigbee dimmer switches are useful in lower-load architectural settings because they do not require a neutral wire, work with incandescent, halogen, and dimmable LED lights, and typically exceed 30 metres of indoor Zigbee range. They are not suitable for CFL, fluorescent lighting, or smart bulbs, so the load schedule must stay disciplined.
That clarity matters for specifiers. If you need elegant visible hardware with black metal, brushed stainless steel, brushed brass, or white metal faceplates, Repenic gives architects and designers a refined option. But for large commercial LED arrays, the control strategy still has to respect inrush physics first.
How can you prevent welded relays?
The best prevention is layered: reduce inrush, choose the right relay, and avoid overloading the contact path. In commercial practice, you may also need staged switching, suppression, or dedicated switching gear designed for the load type rather than relying on a standard relay to do everything.
Prevent welded relays by limiting inrush current, selecting relays with proper capacitive-load ratings, and reducing unnecessary switching stress. If the circuit starts large LED arrays together, use staged activation or inrush-limiting methods. That lowers arcing at closure and extends the life of the contact set.
Common mitigation methods include:
-
Using relays with higher inrush tolerance.
-
Staging multiple lighting groups instead of energising everything at once.
-
Adding inrush limiting or soft-start measures where appropriate.
-
Separating very large driver banks onto multiple circuits.
-
Ensuring the control enclosure stays cool and well ventilated.
The most reliable installations are rarely the most complicated ones. They are the ones that respect the load data, keep the control logic sensible, and give each relay a realistic job. That is the same design philosophy Repenic applies across its wired and Zigbee control range: clear specification, predictable behaviour, and no unnecessary surprises.
Why does BS 7671 still matter?
BS 7671 still matters because relay failure is not just a product issue. It is a system issue involving protective devices, cable sizing, grouping, switching duty, and the actual electrical behaviour of the load. UK installers need to think about the whole circuit, not just the visible switch or dimmer plate.
BS 7671 matters because proper circuit design helps prevent relay overload, overheating, and unsafe fault conditions. Even when the control is smart, the fixed wiring still has to be suitable for the load and the switching duty. Good design reduces nuisance failures and makes commercial lighting systems easier to maintain.
That is particularly important on projects with many LED drivers, long cable runs, or mixed-use zones. A relay that repeatedly switches a large capacitive bank may look fine during commissioning and still fail later if the surge is never addressed. Compliance thinking and load engineering should move together.
For UK trade teams, it is worth cross-checking supply chain options through Screwfix, B&Q, or Toolstation when sourcing ancillary control gear and enclosures. That keeps replacement planning practical, especially on sites where downtime matters and the installer needs readily available parts.
Repenic Expert Views
“Relay welding is usually a symptom of poor load matching, not a mystery fault in the switch. Once you understand the capacitor charging at the LED driver input, the failure pattern makes perfect sense.”
“We specify for real switching duty, not brochure duty. In commercial interiors, a control device must survive repeated events, heat, and surges without losing its contact integrity.”
“Repenic’s value is that it combines a refined finish with honest specification boundaries. For architects and integrators, that makes the product easier to place confidently in a high-quality scheme.”
What should commercial installers check first?
The first check is the driver inrush data, followed by relay contact suitability and the number of luminaires on the circuit. After that, look at switching frequency, enclosure heat, and whether the circuit can be split to reduce the total surge seen by each relay.
Commercial installers should check driver inrush, contact rating, and circuit grouping before they check aesthetics or automation features. If the electrical surge is too large, even a stylish smart switch can fail early. The direct impact is fewer callbacks, better reliability, and less risk of welded contacts on site.
This is also where local knowledge matters. A control strategy that works on a small retail fit-out may not scale to a hospitality or office scheme without rethinking the load split. The best answer is often a better circuit layout, not a more expensive switch.
FAQs
Can a smart relay fail even if the steady current is within rating?
Yes. Steady-state current tells only part of the story. LED driver inrush can be far higher for a very short time, and that is what damages contacts. For installers, the impact is that a relay must be chosen for surge behaviour as well as normal running load.
Do soft-start methods really help with LED arrays?
Yes, when they are correctly applied. Soft-start or staged energisation reduces the sudden capacitor charge event that causes arcing. The limitation is that it must suit the control scheme and be compatible with the lighting layout. For designers, this can mean fewer failures and cleaner circuit segregation.
Are welded relays usually repairable?
Usually not in a reliable way. Once contact surfaces fuse or heavily pit, the relay has suffered permanent damage. The practical limitation is that replacement is normally the correct fix. For maintenance teams, that means identifying the root cause matters more than just swapping the part.
Do Repenic dimmers suit large commercial LED banks?
Not by default. Repenic Zigbee dimmers are suited to compatible incandescent, halogen, and dimmable LED loads, but not every large commercial array will be appropriate. The limitation is in load profile and inrush behaviour. For specifiers, that means checking the real driver data before choosing the control.
What is the biggest mistake on LED relay jobs?
Assuming wattage alone is enough. The real risk comes from capacitive inrush at start-up, not just total load power. That means a simple wattage check can miss the actual failure mechanism. For installers, the result is a higher chance of nuisance trips, premature wear, and welded contacts.
Conclusion
Welded relays on LED loads are usually caused by capacitive inrush, not by the steady lighting load itself. The safest commercial approach is to specify for surge behaviour, choose contact hardware with real capacitive-load tolerance, and split or stage large driver banks where possible. Repenic’s refined control hardware fits design-led projects well, but the load physics still has to be respected first. When the electrical design is sound, the lighting system stays quieter, safer, and far more durable.