Well, it'd be a dull world if everyone always agreed with me
I don't think we've eliminated the heater, simply because the problem persists.
This appears to be an intermittant problem; the breaker doesn't seem to trip while the OP is actually there watching things.
What else can be done before you tear apart the service equipment? We can swap breakers with a known good one (done), we can megger the circuit itself, we can replace the device, and we can replace the affected load. If all those prove to be OK, then we're left with the impossible: there's a problem with the service equipment.
In troubleshooting, you can proceed one of two ways; you can start at one end and work your way along, or you can start in the middle and start isolating by halves. In either case, a necessary first step is to cast aside any previous thoughts you had, and start over.
In the examples I mentioned above, the fault occurred at irregular intervals. With the fridge the compressor had to be running; with the power strip, the bad connection seemed related to moving the various wires attached to it around, and with the cracked buss ordinary traffic vibrations influenced the amount of contact vetween the two sections.
Oddly enough, as soon as I remembered the cracked buss a theory explaining the OP's problem presented itself: poor contact on the neutral side would lead to a voltage fluctuation, which would lead to the heater (a pure resistance) drawing more current. I wonder if anything else has suffered from high voltage transients?
