I know we measure the prospective short circuit current between L1,L2,L3 and N and then double the highest reading

Or, more accurately, establish Ipf between lines and divide by 0.87. Mind you, from my experience, its all fruit machine numbers anyway. As I understand it, the standard MFT instrument simply divides the measured resistance (rather than impedance) into 230. If you are out even by a small margin in the measurement of resistance, the resulting difference in KA can be substantial.

A candidate on the 2391 last week measured 0.33 ohms and 700A on his brand new Megger X, (the new tester that has little electronic diagrams to tell the operator how to conduct each test, heaven help us!) The new 1741 owned by the centre showed 0.26 ohms with an Ipf of nearly 900A. The impedance difference was only 0.07 ohms but the resulting current difference was substantial. The test rig is well into the installation, so you would expect a reasonable degree of accuracy at the higher resistance values than you would at the real intake position with a 1MVA tx in the next field. 

Why should we waste time with prospective earth fault current?

The Ipef should be taken with the earthing system connected so should be greatest at intake. So as you move to a single-phase board downstream and you know the PN and PE KA values at intake, you may choose to set aside duplicate re-testing at the SP board providing the Icn values of the OCPDs are grater that the values obtained at intake..

I know we measure the prospective short circuit current between L1,L2,L3 and N and then double the highest reading

Or, more accurately, establish Ipf between lines and divide by 0.87. Mind you, from my experience, its all fruit machine numbers anyway. As I understand it, the standard MFT instrument simply divides the measured resistance (rather than impedance) into 230. If you are out even by a small margin in the measurement of resistance, the resulting difference in KA can be substantial.

A candidate on the 2391 last week measured 0.33 ohms and 700A on his brand new Megger X, (the new tester that has little electronic diagrams to tell the operator how to conduct each test, heaven help us!) The new 1741 owned by the centre showed 0.26 ohms with an Ipf of nearly 900A. The impedance difference was only 0.07 ohms but the resulting current difference was substantial. The test rig is well into the installation, so you would expect a reasonable degree of accuracy at the higher resistance values than you would at the real intake position with a 1MVA tx in the next field. 

Why should we waste time with prospective earth fault current?

The Ipef should be taken with the earthing system connected so should be greatest at intake. So as you move to a single-phase board downstream and you know the PN and PE KA values at intake, you may choose to set aside duplicate re-testing at the SP board providing the Icn values of the OCPDs are grater that the values obtained at intake..

lyledunn said:

The Ipef should be taken with the earthing system connected so should be greatest at intake. So as you move to a single-phase board downstream and you know the PN and PE KA values at intake, you may choose to set aside duplicate re-testing at the SP board providing the Icn values of the OCPDs are grater that the values obtained at intake

Important to note, though, that unlike Ze, in most installations you can't calculate the prospective earth fault current at a distribution board in the same way, because of parallel earth paths.

Any calculation would have to ignore R2 (or Z2 if CSA of cpc > 16 sq mm, or SWA cables are used) to give you "worst-case conditions".

The resulting calculation will give you a reading higher than the actual prospective earth fault current ... sometimes much higher.

The alternative is measurement, but that's not accurate either !

0.33 ohms and 700A on his brand new Megger X, .. snip...

1741 owned by the centre showed

0.26 ohms with an Ipf of nearly 900A.

That is not a million miles out from identical ;-) 

230/0.33 is 696

230/0.26 is 885

0.33 and 0.26 are  both 0.295 +/ 0.035

i.e. within 12%.

Actually at least for the few MFT designs I have seen inside, because really they look at voltage droop during application of a test load to deduce supply impedance, it is closer to a reading of absolute magnitude of total Z rather than resistive part - measuring that requires accurate I/V phase comparison and that is not so easy.. A

The situation is not helped,  because the test currents are  set small enough not to trouble the lightest ADS (or RCD) it is trying to deduce things by multiplying up from very small changes in the measured voltage numbers, so if other loads are also making the supply voltage fluctuate, the readings can be quite amusing.

A mechanical equivalent is something like trying to open the door from the hinge side, where a very small change in fulcrum position alters the forces massively because the lever length is too short.

And then there is the problem of extra contact resistance in the test loop - quite often a scale and polish of the 'nana' plugs and the probe ends can find a few tens of milliohms.

'Fruit machine' can indeed be an accurate description. In some ways a test lamp and an 'is it earthed at all' test finds  the most serious errors just as reliably, it's just not a very safe test to perform.

Mike

lyledunn said:

The impedance difference was only 0.07 ohms but the resulting current difference was substantial.

Not in relative terms.

I am inclined to wonder how often the PFC gets anywhere near the rating of the devices in a DB (as opposed to heftier panels).