You have 2 VFD's. One is for the supply fan and the other is for the return fan if I'm reading this right. These VFD's also have a bypass on them. I've never seen both contactors try to pull in at the same time. (VFD and Bypass) These units have built in protection for that.
To me it seems it's in the programming of the building management system. They could have a bad line of code in their program which is pulsing their control point in their BMS cabinet. This would cause the control relay to be pulling in and out. The contactor is then following suit until it burns out. Because of the high starting current, the contacts would be arcing and thus pitting. I've seen what a messed up program can do. Just a thought.
I think this is the most likely scenario, or something like it. In a typical VFD-Bypass arrangement, the bypass is sometimes controlled by a manual switch for maintenance purposes only. However, another application is to have it controlled automatically so that if the VFD goes into fault, the control automatically switches over to Bypass mode. That doesn't / shouldn't mean the Bypass contactor COMES ON when the VFD faults, only that the Bypass is ENABLED should the VFD fault. The control circuit should still be determining when and where the FAN comes on and off.
I have seen a couple of related issues with BMS systems along these lines. One is that someone misinterpreted the concept of automatic bypass and it comes on every time the VFD faults, even if the fan was not needed. That's bad, because if the BMS is monitoring static duct pressure, the pressure goes up and it must do something about it, which could make it start to cycle fans on and off to try to maintain balance.
The other scenario is that someone did not think about the consequence of only one VFD faulting and going on Bypass, but the other is still on a VFD. So for instance if the Supply fan VFD faults and goes into Bypass, that means it will go full speed. But if the Return fan is still on a VFD and the accel rate is slow, the duct pressure builds up too quickly and the supply fan is shut down. But the return fan, being still on a VFD, then decels too slowly and the duct pressure drops quickly, so the supply fan is turned on again and the cycle repeats itself. This causes rapid cycling of the supply fan and on Bypass, can burn out the contactor. The solution is to make sure that if EITHER fan VFD faults and goes into Bypass, so does the other.
You need to get a full understanding of what the BMS is doing with these fans and why before you spin your wheels much more trying to find phantoms. VFDs don't usually affect an entire power system enough to cause something like this.