I did a demonstration for some students of the electrical trade. One of them posed the question as to the ultimate failure voltage of a 600-volt rated cable. I thin the specimen we used was #4 AWG. I put the non-shield cable on a steel table for the test and connected a DC hipot. I let the class place bets as to the breakdown voltage. Nobody suggested that it might be more than 10 kV, but I cranked the voltage in a slow ramp to almost 50 kV. Realize, of course, that this was not a laboratory-grade test, just a demonstration of a point.
On CT's driven to saturation, the common failure mode is for the overvoltage to find the weakest insulation in the system. I have seen a set of terminal blocks catch fire. I have seen the connection board on a GE relay drawout case arc over. I would not want to bet on what happens in any given system unless a specific installation of overvoltage protection such as arc-tubes or varistors are installed.
old field guy