3phase lighting and shared neutral

[ColJack]

I knew this one would cause some debate . Its a funny one because in some ways its no different to other 3 phase neutral circuits imo, yet i do agree that it would be better off as 3 phases and 3 neutrals, i.e seperate circuits. Not an option to rip it out and start again though.

Regards.

just fit a TP breaker and then seperate switches ( bad design using a breaker as a switch.. not designed for use as such.. ), or a contactor and seperate / single switch if the current is too much for the switch itself..

 

[industryspark]

Coljack wrote


although it has a means upstream of the CU, you could likewise argue that the large isolator in the local substation makes it comply as that disconnects all the line conductors...
if it's not in the CU that houses the breakers, or is indeed the breakers itself, then it should be downstream..


So its the position of this 'device' that is the key issue here, agree?

I would say that reading the reg you qouted the circuit seems to comply to me, depends how you interpret.

Then again, common sense says that one phase could be isolated after this device leaving 2 connected so you could say it doesnt comply.



So, the best solution might be to change out the 20A sp mcb's for a TP one as suggested by RF, which brings in 314 as he also mentioned.

I believe 314 is division of installation so a TP mcb would take out all the lights in the event of a fault on one phase.

So what about fitting a linked fusecarrier where one fuse could blow yet leave other circuits uneffected? And we still have a device that is capable of isolating all 3 phases? And some seperate switches underneath of course!

Regards

 

[flameport]

if someone disconnected the neutral at the CU and switched one breaker off then yessyou would in effect have 2 ( sets of ) lights wired in series across 2 phases, giving you 400V across 2 lights in series, which is still only 200V across each ( set of ) light..

The lights would be fine.
The person with their fingers on the disconnected neutral might not be (depending on how well balanced the loads were).

The neutral might not be disconnected at the CU - someone might remove the live from one MCB, then climb up a ladder, test that the circuit was dead, then remove the neutral from a light fitting, and find one end of the disconnected neutral now had a significant voltage on it.

 

[ColJack]

care to try that again

see above re: switching..

is there no room in the mini CU to fit a TP isolator on the din rail next to the SP breakers?

 

[industryspark]

care to try that again

Dam quote button!

Yes i agree regarding the switching. Although the mcb's were confirmed to me by the manufacturer as suitable for functional switching so technically ok.

I hate to see them used as such though so seperate grid switches were in the mind to be fitted into the circuit.

No room in the enclosure to fit a TP isolator no-its a 4 module enclosure with 1 blank spare (never easy is it! )

Thanks for the inputs so far. Always good to see how others interpret the regs.

Regards.

 

[ricicle]

if 1 breaker trips it immediately inballances the load and the neutral current starts to climb... ( to what exactly I can't be bothered to work out.. need phasor diagrams etc.. )
same thing happens when lamps start to fail.. a bigger load on 2 of the phases than the other one..

switching on of the lamps.. ? where are the switches and can you guarantee that the user will switch all 3 sets of lights on?

The Neutral current will not rise if one breaker trips. The Neutral current will never be bigger than the biggest load on one of the phases. It will be the vector sum of the phase current of all 3 phases (Ignoring Harmonics)
The problem would occur if 3 circuits off the same phase shared 1 Neutral which wouldn't be a compliant circuit.

 

[industryspark]

Ricicle wrote

The Neutral current will not rise if one breaker trips. The Neutral current will never be bigger than the biggest load on one of the phases. It will be the vector sum of the phase current of all 3 phases (Ignoring Harmonics)
The problem would occur if 3 circuits off the same phase shared 1 Neutral which wouldn't be a compliant circuit.

That is how i understand it too, same as any other 3 phase system which might be unbalanced.

 

[333rocky333]

Am I reading this right the cable out to the 3 sets of lights is using one neutral also.
How would you connect the lights then,by having the other two phases passing through each light?

what about reg 521.8.2 would that not apply.

 

[ColJack]

run to a socket or similar and terminate L1 ph and N there.. then continue to next and do same for L2 and N, then L3 and N etc..

use socket or similar for fast easy repairs ( disconnect and bring it down.. maybe keep one as spare.. fast swap then.. )

 

[ColJack]

if 1 breaker trips it immediately inballances the load and the neutral current starts to climb... ( to what exactly I can't be bothered to work out.. need phasor diagrams etc.. )
same thing happens when lamps start to fail.. a bigger load on 2 of the phases than the other one..

switching on of the lamps.. ? where are the switches and can you guarantee that the user will switch all 3 sets of lights on?

The Neutral current will not rise if one breaker trips. The Neutral current will never be bigger than the biggest load on one of the phases. It will be the vector sum of the phase current of all 3 phases (Ignoring Harmonics)
The problem would occur if 3 circuits off the same phase shared 1 Neutral which wouldn't be a compliant circuit.

it WILL rise, but I never said it will rise above the load of one phase..
what I meant was that in a ballanced system, the sum of the neutral current will be 0.. ..
if you lose a phase then the sums no longer equal 0..

 

[Spark123]

In a bog standard ballanced 3 phase circuit the vector sum will equal zero, when there are a number of unballanced phases the resulting neutral current will be no larger than that of the largest single phase.
However as Ricicle mentioned before I believe SONs do introduce harmonics hence these be taken into account when working out the resulting neutral current, even with 3 ballanced phases of SONs the neutral may not be at zero amps.

 

[tim west]

I was on a job where the neutral on the interlock failed on a genny test and the lighting ended up across phases, they coped ok for a few seconds then one after another the fabled trapped smoke within the ballasts was released along with the pop of windings gradually getting faster and faster as several floors of control gear were trashed, expensive genny test , earnt us a bonus weekend of replacement though

 

[bernardgreen]

Are we getting confuced between "star point" of the supply system and the "neutral" of the final circuit.

Current in the Neutral conductor of a final circuit should be equal and opposite to the current in the accompanying Live (or line or phase ) conductor. If it is not then there is a fault and an RCD will detect it and isolate the Live to the defective circuit. Neutral current drops to zero as well

The sum of all neutral currents should in a well ballanced system result in no current in or out of the star point of the supply system at the sub-station transformer. That zero is the algerbraic and vectored ( phase angles ) sum of all the currents in all the final circuit neutrals connected to the star point.

Motors have a shared neutral for all three phases ( or is it really a star point ) but sometimes this neutral is not connected to the supply. But in a motor loss of one phase means the other phases will be removed by protective devices as the 3ph motor will not work properly with only 2 phases supplied. That is not the case with three sets of lighting where loss of one phase will not be harmful to the lighting PROVIDED the neutral to each light unit is permanently connected to the star point of the supply and is capable of safely carrying the current of two phases of lighting load.

 

[ericmark]

This is an interesting post. I have fitted lighting track many times and it would be impossible to run separate neutrals for each phase so since it had at the BS numbers on the track I would assume there is no problem using a single neutral to feed the three phases.

In all the systems I have worked on we had a contactor doing switching so all three phases were switched together but often they were fused rather than using a circuit breaker so they could have lost a phase.

The same could also have been said for many of the three phase motors used and what ever we did the sub-station still used fuses so we could lose one phase.

To counter the problem of losing one phase we had motor overloads and fuses including the fuse in every lamp plug that was connected to the lighting track.

With the ability to unplug every lamp we never worked on a fitting aerial and they were always replaced and worked on back in workshop so there was little to no danger to anyone maintaining the system. This was one of the main reasons for using the track system since the lights did not need to be switched off to work on a single faulty unit.

On the older systems we were slowly fitting Clik ceiling roses again so single lights could be isolated.

So long as a safe procedure is in place for maintenance I see no real problem in the system as it is related to by "industryspark" however this does not mean it complies with the regulations. But we know there are times where the writers of the regs have missed things like for example cables buried in a wall where they are part of a SELV circuit so we must use some common sense rather than trying to become lawyers.

Looking at 521.8.2 for example it says "line and neutral" not phase and neutral so must refer to single phase circuit. And if you look at circuit definition it says "device(s)" so does not need to be a single device but could be a set of 3 fuses.

If it were my job I would do a risk assessment and file it to show I had considered the problems and had introduced a safe method of work with a method statement. I would not worry too much about following the letter of the regs. The permit to work would require the method prescribed to be followed so I can see no real problem?

 

[bernardgreen]

One solution to the problem of damage to units could be to trip all three phases to OFF if the current in the shared neutral (where in connects the supply neutral bar) rises above a preset limit. This limit would sensibly be the maximum safe carrying current of the weakest part of the shared neutral.

To counter the problem of losing one phase we had motor overloads and fuses including the fuse in every lamp plug that was connected to the lighting track.

On motors where there was a star point that was not connected to supply neutral a voltage operated 3 pole trip was connected between motor star point and supply neutral. This picked up phase failure and a few motor faults as well.