I remember reading some comments on the IEE forum about very low Ze values some months ago, and have just come across such a situation myself.
Clipped the loop tester onto the main terminals (installation & bonding isolated, of course), selected the 20-ohm range, hit the button, and presto: "0.00" on the display. Several repeated tests came up with the same result every time.
Now, the specified accuracy for the tester (Robin 4116) is 2% plus 2 digits, so all I can really say for sure is that the Ze on this service is probably less than 0.02 ohm.
Given this level of accuracy on the typical tester, it has to make me wonder what use Ze measurements are when we start getting down to such low values. That Ze I measured could be 0.02 ohm, giving a prospective fault current of 12kA. Or it could be 0.015 ohm with a PFC of 16kA, the maximum that the DNOs will usually specify by default on a single-phase supply to cover themselves, even though for many services the true value will be considerably lower. It might even be worse than that for all I know.
Even with a measured Ze of 0.04 or 0.05 ohm, the accuracy of the tester leaves sufficient scope for error that one has to wonder about the value of such measurements at these extremes with the typical loop tester.
Personally, I tend to think that the widespread adoption of digital readouts leads people to infer a degree of accuracy which is not inherent in the equipment. People see, for example, "0.05" on a display and don't stop to think that although the resolution goes down to 0.01 the accuracy is poor at the limits of the instrument's measuring capability (i.e. in this case, just the 2-digit part could mean a true value of anything between 0.03 and 0.07 -- A very significant difference).
when reading 0 Ohms, I'd ask: How far and is the transformer and estimate the R of the loop. So is it realistic? In Germany you will always hear people say (literally): Who measures a lot, will measure a lot of crap.
I am in the situation that we have our own transformer (360kVA) in the building and nevertheless I have never read such values.
So this (low Ze) should rather be an industrial problem, where you will have to calculate short currents being unable to measure them correctly because of big or paralleled transformers and very short distances on strong copper busbars.
In Germany btw you usually never determine Ze , but only Zs for any circuit that is not protected by an RCD. Maximum short circuit current is defined to be 6kA for MCBs and 25KA for mains fuses at the entrance. To guarantee this is the job of the Poco, we don't care about it in domestic installations.
Some interesting comments there about instrument accuracy.
On the service where I couldn't obtain a meaningful Ze reading, I'm not quite sure where the transformer is located. There's the usual "middle of town" xfmr feeding the 3-ph network along the High Street, but that's about 200 yards away. I would have expected a slightly higher Ze if it's coming from there.
However, there is a new xfmr located right across the street, put there when a new supermarket was built two or three years ago. The service entrance I was measuring from is right by the front of the shop, and the xfmr is literally about 20 ft. away across the road. I'd always assumed it was only feeding the supermarket, but perhaps the units in the shopping center were also transferred to it to ease the load on the High St. sub-station. There might even be another xfmr somewhere around the back of the complex that I haven't noticed. I'll have to take a closer look.
Paul, I've come across this before with Ring style distribution mains. We use what is known as a Low Ohms Meter, it measures between 000.1000Ohms and 1 Ohm. It is important with Power Systems earthing. I would also tend to back up what Wolfgang is saying. The closer the transformer, the higher the fault current. I've seen switch-boards blown to bits, through MCB's that can't handle the fault current. Any of you guys ever used BKM Breakers?. IMO Don't!!.