Nothing. This isn't addressed by the NEC.
Wow, that's surprising it's not addressed in any way. I actually came across the question when having to familiarize with cable limiters. It seems some feel these would somehow protect the remaining cables but I don't see how if each one is only protecting the cable it's attached to.
Cable limiters do not protect from overloads. They are more like an instantaneous trip breaker. Their intent is to clear the faulted cable in a parallel set and let the remaining cables carry the load.
In your original example parallel #1/0's, what would cause a single cable to fail and not cause a fault where the OCPD would open?
Not sure I follow. If one of the parallel cables became disconnected somehow (faulty lug connection) and did not touch ground, why would that trip the OCPD? The single remaining cable is now carrying the full load with OCP at twice the limit of it's ampacity.
Why would the cable become disconnected? The likelihood of a proper installed cable suddenly disconnecting itself is so infinitesimal that the NEC doesn't consider it. A cable's insulation failing or faulting to ground is more likely but that's already protected by the OCPD.
Where would you draw the line on things like this in the NEC? What about someone changing the trip plug in a breaker to double its trip setting, or someone replacing 225 amp fuses with 400 amps fuses? Both of those are much more likely than one conductor of a parallel set becoming disconnected without causing a fault.
