UK After 3phase board, ideally some board wiring/testing too - LONDON/SE

mech and elec · Aug 21, 2021

I'm a mechanical manager with some electrical background (2330 L2 from college, 18th ed, 2392).

About to be an M+E manager so need to brush up on areas I'm unfamiliar with - 3 phase wiring and testing is the big one.

Is there anyone in London area with access to a (used) 3phase board who can guide me through install and testing for a day or two? Happy to renumerate. I have an MFT.

Failing that if anyone has a (preferably at least populated) 3phase board for sale, do get in touch.

Many thanks

timhoward · Aug 21, 2021

Do you have a copy of GN3? That explains 3 phase testing in detail and is an essential purchase if not.

3 phase testing has exactly the same principles as single phase testing, just with more permutations. Most tests are do it three times and record the biggest number. The main quirk is that as many MFT's can't measure Zs between phases we do a mathematical hack, see below.

Safe Isolation = 10 tests: each phase to earth, each phase to N, each phase to each other phase , and finally N to E.

Get used to writing out a 3 by 3 grid of 3 Ze's, 3 PEFC's and 3 PSCC's as follows:
Ze = Each phase to supply earth, exactly as single phase with main earth disconnected, highest number is final answer.
Record the three associated PEFC's at the same time on 2nd row.

For PSCC record each measurement from each phase to N. To allow for a dead short between phases double the largest reading. (Mathematically should be times root 3, the relationship between 230v and 400v, but GN3 says double it. )

As single phase, the PFC is the highest reading from the 2nd and 3rd rows, remembering each PSCC is doubled first, so invariably it will be one of the PSCC readings.

Using the arrows on a mid-range 2 pole tester, check phase rotation is consistent for the whole installation.

Dead tests: R1+R2 test is just repeated 3 times, each phase (L1 to CPC, L2 to CPC, L3 to CPC) with the highest reading recorded.

IR tests - basically all permutations tested of L1, L2, L3, N (where present) and CPC. See GN3

I may have a 3 phase board with some breakers that's missing a front cover, I can't remember if I chucked it or not. If I still have it I'll send a private message.

mech and elec · Aug 21, 2021

timhoward said:
Do you have a copy of GN3? That explains 3 phase testing in detail and is an essential purchase if not.

3 phase testing has exactly the same principles as single phase testing, just with more permutations. Most tests are do it three times and record the biggest number. The main quirk is that as many MFT's can't measure Zs between phases we do a mathematical hack, see below.

Safe Isolation = 10 tests: each phase to earth, each phase to N, each phase to each other phase , and finally N to E.

Get used to writing out a 3 by 3 grid of 3 Ze's, 3 PEFC's and 3 PSCC's as follows:
Ze = Each phase to supply earth, exactly as single phase with main earth disconnected, highest number is final answer.
Record the three associated PEFC's at the same time on 2nd row.

For PSCC record each measurement from each phase to N. To allow for a dead short between phases double the largest reading. (Mathematically should be times root 3, the relationship between 230v and 400v, but GN3 says double it. )

As single phase, the PFC is the highest reading from the 2nd and 3rd rows, remembering each PSCC is doubled first, so invariably it will be one of the PSCC readings.

Using the arrows on a mid-range 2 pole tester, check phase rotation is consistent for the whole installation.

Dead tests: R1+R2 test is just repeated 3 times, each phase (L1 to CPC, L2 to CPC, L3 to CPC) with the highest reading recorded.

IR tests - basically all permutations tested of L1, L2, L3, N (where present) and CPC. See GN3

I may have a 3 phase board with some breakers that's missing a front cover, I can't remember if I chucked it or not. If I still have it I'll send a private message.

Yes I have GN3. Encouraged to see it appears to be much of the same but more elaborate than single phase, as you point out.

Grid idea sounds sensible.

Do let me know about the old 3P board please.

Many thanks for your comments

pc1966 · Aug 22, 2021

timhoward said:
For PSCC record each measurement from each phase to N. To allow for a dead short between phases double the largest reading. (Mathematically should be times root 3, the relationship between 230v and 400v, but GN3 says double it. )

The doubling of your usual PSSC comes from the worst-case of a "three phase bolted fault" when all 3 lives are simultaneously shorted together and the resulting fault has no need of the neutral return, so the impedance of the neutral is unimportant. As most low-ish power systems have identical N and L conductors they ought to have similar impedance and so removing half of that results in your doubling of PSSC to get the 3-phase value used to determine if MCB/MCCB/switches/etc are able to survive the fault.

Of course the equivalence of N to L is not always true as some systems have a reduced size neutral as they assume a certain degree of 3-phase balance is always present, and also the errors can be more significant close to the transformer when its impedance becomes significant compared to the cable impedance. In those cases if you can measure the L-L PSSC values you should multiply the max found by 2/sqrt(3) ~= 1.155 to get the worst-case fault.

Not all domestic MFT can measure the L-L PSSC values, but even some reasonably priced ones can do so (like my DiLOG 9118) and typically they also offer other things you might need for 3-phase testing like checking the phase rotation direction.