What measures are employed to provide protection against ground faults on the load side of an UPS? I am in the UK and I have found that with mains healthy, earth loop impedance tests taken on the load side of the UPS often results in relatively high impedance values thus often rendering reliance for shock protection on mcbs alone less than adequate. However, I am also aware that there is ground fault protection built into the UPS which rapidly reduces voltage to the load in the event of a fault to earth. What I would like to know is just how this operates and can it be relied upon to provide appropriate shock protection or is it just designed to protect the UPS itself?
The output side of the UPS is effectively a separately derived system, so I would expect the neutral to be grounded in the unit.
So a phase-earth short on the output is effectively a phase-neutral short as far as the UPS is concerned.
Many UPS systems (and small inverters in general) just won't supply enough current to trip anything but the lowest rated MCBs, so it would be a case of how to determine the touch potential at the point where the fault occurs.
As I see it, it will depend upon:
1. The potential divider effect of the phase and earth conductors.
2. What the output voltage of the UPS drops to under the fault condition.
#1 is easily calculated from the cable sizes, but point #2 will depend upon the UPS design.
Unless the output side has cables so long as to limit the current severely (in which case the voltage drop under normal use would almost certainly be far too great anyway), or the UPS is a really large type, then I don't think the average UPS is going to deliver anything like the 40 or 50V touch potential at the far end that would be considered a hazard.
There's nothing to stop you running the UPS output through an RCD if you wish.
Paul, Many thanks for your reply. I am aware of the points you have made and I concur. I always take r1+r2 values and ensure neutral is earth referenced ( amazing the number of even large ups units that are not installed so! On an earth fault the collapse of voltage to the load caused by the ups current limiting protection may be the reason that earth fault loops are apparently high. I know that most people would not take earth loop readings as a rule on such circuits but I would have liked to compare notes or discuss the actual cause.
The source impedance of the UPS will be much higher than that of a normal supply xfmr, but the built-in current limiting could certainly result in apparently higher readings you'd get using a normal loop tester.
It would be informative to put a voltmeter on the UPS and observe it while the loop tester is operated at the far end.