The Breaker Tripped. The Generator Was Innocent.

The building test was going fine until the second chiller started.

400 kVA set, running sweet at maybe sixty percent. The chiller contactor pulled in, the lights took their usual half-second dip — and then the whole room went dark. Not the engine. The engine never missed a beat. The main breaker on the generator had opened, politely and completely, at about eighty-five percent load.

Everybody turned and stared at the generator like it had done something wrong. It hadn’t. I’d seen this movie before. The breaker did exactly what it was built to do. Somebody had just asked it to do the wrong job.

Here’s the thing most people get backwards: the breaker bolted to your generator is the least understood part of the whole installation. Engine, alternator, controller — everyone respects those. The breaker gets treated like a light switch with delusions of grandeur. Then one Tuesday afternoon it ruins your day, and the generator takes the blame for it.

After twelve years of getting called out to “generator faults” that turned out to be breaker problems, I’ve got a list. Five mistakes. I keep seeing the same five.

Mistake one: sizing the breaker at exactly 100% of the set’s rating

A thermal-magnetic breaker is calibrated at around 40°C ambient. That’s the datasheet number. Now put that breaker inside a canopy enclosure, in direct sun, on a 38-degree afternoon. The air inside that box is pushing 55, maybe 60 degrees. Your “800-amp” breaker is now, thermally speaking, roughly a 680-amp breaker. Nobody told it. It just trips earlier.

So the set is pulling a sustained 720 amps — a 500 kVA machine at 400 volts works out to about that (the kVA-to-amps math works like this) — and the breaker has been baking in its little oven all afternoon. An hour into the test: click. Darkness. The generator was at maybe ninety percent of its standby rating. The breaker thought it was overloaded, and by its own thermometer, it was.

The field rule electricians use is 125% of continuous current for the breaker frame and trip. It exists for exactly this reason. Size the breaker for the current the set will actually carry, on the hottest day, in the hottest box — not for the number printed on the alternator nameplate in an air-conditioned test cell.

Mistake two: leaving the trip unit on factory defaults

Breakers ship with settings. Long-time pickup at 1.0x, instantaneous somewhere around 5 to 8x. Those defaults are fine for a distribution board in an office. They are not fine for a generator feeding motors.

An across-the-line chiller pulls roughly six times its running current for a few hundred milliseconds. To the instantaneous element, that looks exactly like a fault. It does its job. It opens. I worked on a site that ran perfectly for three years — then a new tenant fitted out the floor above, added a second chiller, and the first hot Friday afternoon the main breaker started tripping like clockwork. Nothing about the generator had changed. The load had.

The fix is boring and it works: set the instantaneous pickup above the worst-case asymmetrical inrush, and give the short-time element enough delay to ride through a motor start. One more wrinkle — a generator’s short-circuit current is small and it decays fast (subtransient reactance doing its thing), so a trip curve that behaved perfectly on utility power can misbehave on generator power. Don’t leave defaults. Get the settings coordinated for the actual load, on the actual source. If motor inrush is new to you, I wrote about what starting current does to generator sizing here — the breaker is the other half of that story.

Mistake three: thinking the breaker protects the alternator

It doesn’t. The main breaker protects the cable. The alternator is protected by the controller — or it isn’t, and then nothing protects it.

I saw the aftermath of this one at a small factory. The set had been running at about 115% of rating for the better part of an hour during a utility outage — someone had “temporarily” added load, twice. A thermal-magnetic breaker at 1.15x takes an hour or more to trip; that’s what the time-current curve says, roughly. The alternator’s windings gave up at around forty minutes. When we opened it up, the breaker was fine. The cable was fine. The varnish on the windings had gone the color of toast.

The breaker did its job. Its job was the cable. Protecting the machine is the job of the controller’s overcurrent protection, set on an inverse-time curve that sits underneath the alternator’s thermal damage curve. If those settings are wrong — or if someone disabled them because of “nuisance trips” — the breaker will happily watch your alternator cook. An old-timer I work with disagrees with me on the margin here; he says a slightly oversized breaker with tight controller settings is the safest combination going. I’ve come around to his view. The point stands either way: know which device is protecting what.

Mistake four: not touching the breaker for ten years

Molded-case breakers are mechanical things. Grease hardens. Contacts oxidize. Springs take a set. A breaker that sat closed for a decade can absolutely refuse to open when it should — or refuse to close when you need it, which is how you discover the problem at 2 a.m. during a real outage instead of on a Tuesday when there’s time to fix it.

There’s a button on most of them marked push-to-trip. Almost nobody pushes it. Push it once a year. Exercise the mechanism. And get an infrared camera on the lugs during your annual maintenance — I found one lug running at about 95 degrees while its neighbors sat at 45, on a set that “ran fine.” Thermal cycling had walked the lug loose over the years. Twenty minutes with a torque wrench to the manufacturer’s spec and the hotspot was gone. That connection would have failed eventually, and it would have failed hot.

Mistake five: the undervoltage coil that “adds safety”

This one is my favorite, because it was installed with the best intentions. Somebody specified an undervoltage release on the generator’s main breaker — voltage disappears, breaker opens. Sounds safe, right?

Then a big motor starts. Voltage dips to seventy-five, eighty percent for a few hundred milliseconds. Completely normal. The undervoltage coil doesn’t know that. It sees voltage below its dropout threshold and does the only thing it knows: it trips the breaker. Right in the middle of the motor start. The one moment you most need the breaker to stay closed.

I watched a facilities manager go through three rounds of “the generator is faulty” before someone thought to look at the breaker accessories. If you want the controller to be able to open the breaker on a real fault, use a shunt trip wired to the controller’s fault output — then the controller decides, based on actual fault logic, not a dumb coil reacting to a normal voltage dip. An undervoltage release on a generator main breaker is a solution looking for a problem, and it usually finds one.

The short version for commissioning day

If you’re signing off on a new installation, this is the whole breaker checklist. It takes an afternoon:

• Push-to-trip test on the main breaker, witnessed and logged.
• Trip settings recorded — long-time, short-time, instantaneous — matched to the actual load, not the factory defaults. Ask the builder (Pavo or whoever supplied the set) for the settings sheet from the factory test.
• Lug torque checked to spec and marked, then an IR baseline photo of every connection under load.
• A real load-bank step that includes your biggest motor start, watching the breaker ride through it. (This belongs in the commissioning checklist, not in a footnote.)
• An annual exercise date in the maintenance log. Breakers are like muscles.

One more field note before I go: a tripped main breaker is also the most common reason a “generator won’t start” call turns out to be nothing. Check the breaker before you touch the engine.

And the next time the lights go out and someone says the generator failed — look at the breaker first. Nine times out of ten, in my experience, the “generator fault” is a breaker doing exactly what somebody told it to do. The generator was innocent. It usually is. If you want a second pair of eyes on a breaker size or a trip-settings sheet, talk to us — send the single-line diagram and the load list, and we’ll tell you straight whether the breaker is the problem.