Puzzling 3-phase system with aligned phases

[zigunov]

This one is really puzzling me! I usually don't post about electrician problem solving but I've never seen anything like this:

(Backstory)
We have this 500kW resistor bank for an air heater in our facility and the resistors are not really heating. We open the panel and find this 480V/280V 3-phase heater system with some seriously beefy contactors and an SCR for regulating the power through a PID controller. We put the controller in manual mode and request 100% of power out of it. The heater is drawing 0.5A from each phase (clamp meter measurement), which is not nearly enough for a 500kW system.

We go through the continuity tests and everything seems to be properly connected. The heater resistors seem to exist and have a reasonable resistance. We then notice something strange. There's 280V between the phases of the resistors and the ground (after the contactor that turns them on), but 0V across phase 1 and 2, 2 and 3, etc. When the resistors are turned off (contactor opens all three phases), the voltage across the phases is 480V.
(End of backstory)

So we decide to dig deeper and plug in a scope with 1:10 probes wired to each phase. Each probe is plugged to one of the phases, before the contactor terminals, whereas the grounds are all connected to the ground chassis of the panel. See below a picture of the scope trace right at the moment where the contactor is actuated (before = contactor open, resistors "off"; after = contactor closed, resistor "on").

I've never seen this and it makes no sense to me. As soon as the contactor hits, the three phases align. We see then 0V across phases, and the resistors don't heat. But WHY?? These resistors must be connected either in delta or wye configuration, and either way the phases should never align like that. I have no idea of how that could be. The resistors measure a real resistance across them, of about 5 ohms (though I'm not sure that means much because we don't know whether they are in a wye or delta configuration).

Anyone has any thoughts? What really stumped me is seeing the trace in the picture attached. I've never seen a 3-phase system have its phases align when a contactor is toggled like that!

[Ian.M]

Presumably the resistor bank is either Delta wired or Wye with no Neutral.  Its most likely that the feeds for two of the phases have a high resistance so their voltage collapses (to the remaining phase) under load.  How the contactor and control circuit manage to get power during this is another matter, maybe the fault is downstream of where they get power, or maybe they are running off the good phase  and Neutral, or maybe they are on a different circuit so independently powered.

Caution: Something that shouldn't be is probably getting *VERY* hot while the contactor is on, and may even catch fire!

[geggi1]

Test it with the voltage on and start measuring the voltages from the main contactor and follow the wiring towards the resistors. Sooner or later you will most likely find some broken wiring.
From my experience heaters are usually a nightmare to troubleshoot when they are large because because of the complexity of the wiring.

[Someone]

Caution

CAT I scope (and probes?)
CAT III energy
.... additional potential boom

[f4eru]

probably 2 phases of some element failed before your contactor, and, and, when all is off, capacitve coupling make it appear normal in open loop voltage.
Are there main fuses before arrival ?

[Gregg]

My advice is to measure the voltage drop across every conductor and connection with a quality meter.
My bet is that contactor contacts are pitted, high resistance and need replacement.  The resistance somewhere is so high that your scope or even a thermal image device probably won’t be of much use.