Dear All, Mr.Hamid’s post on 24th April 2008 is correct.But it is not the whole truth.I would like add as as follows: 1. The neutral current, as already pointed out, will be the difference between the primary and secondary currents. 2. From the diagram given in the riddle, it is noted that there is an unloaded tertiary winding. This will tend to make the zero sequence currents flowing in the primary and secondary for a Ph-G fault more than than the case without the tertiary winding. 3. Further, the tertiary winding will make the distribution of zero sequence currents between the primary and secondary different.It will be no more inversely proportional to the primary and secondary voltages. The primary contribution will be less, depending upon the tertiary leakage reactance. 4.Since the current in the neutral will be 3 times the the difference between the primary and secondary zero sequence currents,the presence of tertiary winding will make the actual fault current in the neutral more than that expected without it. By properly designing the leakge reactance of the tertiary the neutral current can be made higher and the relay can be made sufficiently sensitive without increasing the setting. Your comments please.