Difference between Using Perfect E and Symmetry Perfect E

I am trying to reduce the simulation time for my model, so I tried to apply symmetry boundaries. When apply the "Symmetry perfect E" and "Symmetry perfect H" to appropriate boundaries. I get a similar result compared to simulating the full problem domain. I do not have a port in my simulation so I applied Perfect E and Perfect H to the boundaries I initially applied Symmetry Perfect E and Symmetry Perfect H respectively. But I got a different result.

I would like to know what the difference is between using Symmetry Perfect E/Symmetry Perfect H and Perfect E/Perfect H.

Perfect E and Perfect H B.C.s have the property of imaging. that means the electric current J and magnetic current M will have a virtual image inside the PEC and PMC at the opposite position. Balanies Text books "Advanced Engineering Electromagnetics" talks about this very clearly.

With this principle, there is no difference between PEC and Symmetric PEC or PMC and Symmetric PMC. It just depends on how you use it. But the key is the EM field will be mirrored by PEC and PMC that is why they call it "Symmetric BC". But its the exact same BC.

Wiki has a article for this
http://en.wikipedia.org/wiki/Method_of_image_charges

There is also a youtube video talk about this
<iframe class="restrain" title="YouTube video player" width="640" height="390" src="//www.youtube.com/embed/N-1F6o2tx80?wmode=opaque" allowfullscreen="true" frameborder="0"></iframe>

Per-your question, you did not describe your problem clearly, what is the model you are simulating? Show some pictures

Thank you for your response. I also got a reply from ANSYS. The field solutions are exactly the same but for post processing the symmetry BCs are used (the mirrored part of the problem are included in post processing) while for the Perfect E and Perfect H, post processing results are generated of just the visible problem domain. Here is my simulation domain below, the full domain and the quarter domain.

When I calculated the radiated power but the structure in the box, I get the same answer when I apply the symmetry E/H as BC, but when I use the Perfect E/H, I got exactly a quarter of the radiated power. That is the difference I observed and in agreement with the response I got from ANSYS. The major difference between is in the post processing.

Thanks