I'm fairly new at this so here goes: We have a Star connected gen. bus with a floating neutral for one of our power plants which is rated at 13.2k. It's grounded by a zig-zag transformer rated at 50A continuous current and 400A 10sec rating capacity. So I was task to figure out how to calculate the temperature for the zig-zag. First I thought I2R but there is no resistor.
Any help will be greatly appreciated.
Well, there certainly is 'R' unless you've got a superconducting zig-zag so that will be some of the heat accounted for. The rest will be iron losses in the core. There's no way for us to calculate or even make a good estimate of the temperature rise: best bet is probably to call the manufacturer with the serial number and ask for a copy of the works test data.
ScottyUK is right; the maker's data is just about the only place to start for the losses.
How warm the bank actually becomes when placed in service will of course be determined by the sum of the heat generated due to losses plus the value of the neutral return current as determined by how well balanced the loads are across the phases of your system; the less neutral return current there is the less heat will develop from that source.
IEC 600076-1 ch. 4 Demonstration of ability to withstand short circuit indicates how to calculate the temperature rise- up to 10 sec short-circuit duration. No difference between connection type-except number of windings-two or more.
However, one has to know the winding conductor cross section area since current density [J] has to be calculated priorly.
The temperature rise during a 10 second overload will be basically due to thermal mass if the thermal time constant is much greater than 10 seconds. But,, back to basics: specifically stating that it is a zig-zag winding does not seem to have any bearing on and method which calculates temperature rise. Since you stated that there is "no resistor" we get the feeling that calculation of thermal rise of transformers is a very new issue to you.
