EMC filter selection and installation while there is power line filter leakage current

Question

I am working in telecom company which is willing to install EMC telephone line filters and EMC power line filters on single facility that has all the equipment. I have been reading about filters issue from last two to three weeks. I have seen that all the manufacturers that are building these type of filters are providing leakage current data for EMC power line filters for single phase and three phase power line filters but no such specs for telephone line and data line filters. My query is:

1. Can power line filter, telephone line filter and data line filter be install on same single facility even though power line filters have leakage current?
2. How I can separate the power line filters from other filters on single facility so that telephone line does not get noise from leakage current of power line filter due to the ground loop?
3. Will separate ground/earth for each filer will eliminate the noise problem or not?
4. Is there any other solution that I can use to keep telephone lines secure from noise of leakage current of power line filters.

Answer 1

The leakage current specification for power line filters is provided for applications which are sensitive to earth leakage, such as medical equipment. It can be safely ignored in a system where all inter-equipment and out-bound interconnects are balanced and run over twisted pair (noise incoming on a system earth reference is common mode noise to a properly balanced input, and is at a low enough frequency that the DC CMRR of the input should still mostly hold).

So, to answer your questions in order:

1. Yes -- there is no problem with this. The leakage current of a powerline filter must be <5mA anyway, otherwise you'll see GFCI/RCD nuisance tripping all over the place, and nobody wants to have the shame of building an ersatz GFCI tester now, do they?
2. TL;DR -- stop worrying about the "ground loop", it won't break anything in a reasonably well-designed system. For the gory details, I highly recommend Henry Ott's fantastic work, Electromagnetic Compatibility Engineering.
3. Earth itself is a poor conductor. It's the equipotential between the filters that counts; this is typically accomplished with a reference plane or reference grid in the equipment area/room. See the linked reference for more details!
4. Unless you're in some sort of alternate universe where the Bell System never existed, your telephone lines are balanced twisted pairs -- which, when applied correctly, have excellent rejection properties at 60Hz. So, I would not be concerned in the least about "back coupling" unless you hear hum; if you are hearing hum, I'd check for imbalances in the lines before I started diddling with random ground-related things.