Revision history [back]

The keyword here is "credible fault". If a fault is likely to occur in your network then it has to be studied. "Conventional" criteria in power system studies have been pass down based on past realities (lack of computational power, lack of data). Modern analysis tools are able to calculate 3ph, LG, LLG, LL faults on a single run. In a substation, the breaker duty capability should be above the max of those 4 fault currents. Now the issue is data quality (network data, sequence data).

I recommend to study all four types of faults in short circuit studies.