3 phase at work - 1 phase down - questions?

[mattylad]

Hello, at work in an old victorian building we have a 3 phase input.

This morning several sockets are off, we think its because a phase has gone down. I'm not investigating it further than looking at the breakers (all on) and someone has been called.

However, in the server room we have 1 dbl socket that works and one next to it does not. Some rooms on one floor are on another is not etc.

How should 3 phases be wired around the building?
It seems weird that some sockets work - some dont so if they are on different phases then that to me sounds dodgy.

BTW - we dont actually use and 3ph equipment.

 

[ericmark]

It is considered as good practice where 3 phase is used as a single phase supply to only have one phase to each room so maximum voltage in any room is 230 volt but I have not seen any regulations that says this must be the case.

The main problem is to balance the phases and there is a trade off between keeping each room with only one phase and balancing the load on each phase.

In days gone by I have seen UPS run on a three phase supply powering a whole suit of rooms with single phase which is maintained with a power cut. However UPS units for a single PC station are now quite cheap and the last time I looked into using UPS I found cheaper to put a UPS at each desk than one central unit. As to server room much would depend on when it was installed as to if three or single phase supply.

Previous use will also dictate how split. I remember one office with three phase as originally a Lab and they had one or two machines which needed three phase.

Lighting is another thing today mainly HF units in offices so stroboscopic effect no longer an issue but years ago it was common to split across phases to reduce the stroboscopic effect.

As offices are refurbished and RCD's are added the dangers of 400 volt rather than 230 volt are much reduced and so today I would really only worry about balancing load.

Oddly I was told as an apprentice it was not allowed to have two phases in the same room but I have been in latter years unable to find a regulation to support that. It may have been in an early edition before 16th Edition when it became BS7671 the regulation book also had what is now only in the guides with rules like no socket within 3 foot of a sink. These rules have been dropped and replaced with more general rules like socket must be suitable for environment.

Will be interesting so see what others say on this subject.

 

[plugwash]

IIRC there used to be a requirement that if accessories on different phases was within some set distance there had to be a 415V warning label but that requirement was later removed.

 

[Risteard]

It is considered as good practice where 3 phase is used as a single phase supply to only have one phase to each room so maximum voltage in any room is 230 volt

By some. Others think it far better practice to try to balance the phases equally so that there is as little neutral current as possible.

 

[JohnW2]

It is considered as good practice where 3 phase is used as a single phase supply to only have one phase to each room so maximum voltage in any room is 230 volt

By some. Others think it far better practice to try to balance the phases equally so that there is as little neutral current as possible.

Does the balance really interest or concern anyone other than the DNO? If, as per what we're presumably talking about, the phases are all used separately within the installation (which is the situation I have in my house), then all the installation's (single-phase) neutral conductors (as far back as the DB) will be carrying 'whatever currents they are carrying', dependent only on the loads. Even if the phases are very well balanced, the only place where neutral current will be very low (or zero) will be on the DNO's side of the DBs neutral bar.

Kind Regards, John

 

[mfarrow]

As a commercial user I believe (or rather have been told, but not seen the evidence) that we are charged for neutral imbalance.

The only way I can think of to truely guarantee neutral current being low is via 3-phase loads or single-phase loads in series across a split-phase supply (i.e barn heaters in series fed from remotely from farm premises).

In terms of use in a socket or home, I applaud the use of sockets spread across multiple phases, as phase outages do happen and it's good that this office can keep going despite the inconvenience. Ideally this would also be across multiple protective devices with a single circuit, though I'm not sure whether you could have separate RCBOs in an RCD-orientated age. I'll have to look that one up.

I remember something about labelling in the 16th, although I thought it was misinterpreted. Agaiin, that's one for later.

 

[Risteard]

I remember something about labelling in the 16th, although I thought it was misinterpreted. Agaiin, that's one for later.

It was misinterpreted as it was intended to relate to voltages to Earth (Uo) exceeding 230V. This has been stated explicitly since Amendment No. 1 to the 17th Edition.

 

[ericmark]

As I expected there are different views. Also as I expected no Current rules as to what should be done. So as far as you are concerned there is nothing to be done.

 

[JohnW2]

As a commercial user I believe (or rather have been told, but not seen the evidence) that we are charged for neutral imbalance.

Interesting. I can certainly tell you that, as a domestic user, that simply wouldn't be possible - I just have a single 3-phase meter, and the (single) reading on the meter obviously gives no indication as to how balanced, or otherwise, the load has been.

As I said (and unless there is sometimes a financial incentive such as you mention), I can't really see what incentive there would ever be for a user of separated phases of a 3-phase supply to have any interest in whether the phases were balanced or not.

In terms of use in a socket or home, I applaud the use of sockets spread across multiple phases, as phase outages do happen and it's good that this office can keep going despite the inconvenience. Ideally this would also be across multiple protective devices with a single circuit, though I'm not sure whether you could have separate RCBOs in an RCD-orientated age. I'll have to look that one up.

I have certainly occasionally benefitted from having all three phases when there has been a 'phase outgages'. As for you latter point, if one is simply splitting the phases into separate single-phase 'sub-installations' (which is the situation in my house), one obviously can use RCDs and RCBOs 'in the usual way', and what one obviously does not want is a 3P+N RCD anywhere in the upstream part of the picture. Despite all final circuits being RCD protected, I have to have (or so I believe) up-front TD RCDs in my TT installation to protect long distribution cables of at least two of the phases - but I deliberately have separate SP RCDs for each phase, for the above reason.

I remember something about labelling in the 16th, although I thought it was misinterpreted. Again, that's one for later.

I think you're right. If there is any question of two different phases which may require separate isolation (as with any 'dual circuit' situation) existing within the same enclosure, then a warning label to that effect is clearly appropriate. However, I never really understood the labels warning about 400V (or 440V) within enclosures or between things - what precautions would one take knowing that 400/440V pds exist which one wouldn't take if they were 'only' 230/240V?

Kind Regards, John

 

[AndyPRK]

A 415v yellow sticker.

 

[mfarrow]

I have to have (or so I believe) up-front TD RCDs in my TT installation

If you mean 'TP' as well as TD then yes, I've confirm my own suspicions:

531.2.1 An RCD shall be capable of disconnecting all the line conductors of the circuit at substantially the same time.

So if you have a circuit which requires an RCD for Automatic Disconnection of Supply (section 531) then it needs to disconnect all line conductors at the same time. There is no stipulation of RCDs provided for Additional Protection, so one assumes having one per phase on a 3-phase circuit, where ADS is provided by fuses/breakers, is acceptable!

 

[JohnW2]

I have to have (or so I believe) up-front TD RCDs in my TT installation

If you mean 'TP' as well as TD then yes, I've confirm my own suspicions:

No, I don't, and deliberately not - for the reasons I explained. There are there separate TD DP RCDs.

]531.2.1 An RCD shall be capable of disconnecting all the line conductors of the circuit at substantially the same time.

So if you have a circuit which requires an RCD for Automatic Disconnection of Supply (section 531) ...

Which I do.

...then it needs to disconnect all line conductors at the same time.

Assuming you mean all four of the conductors (3P+N), that is not my interpretation - and I would say my view represents common sense. There are no 3-phase 'circuits', either distribution circuits or final circuits, in my installation, and nothing beyond the meter is 3-phase. There are essentially three separate single-phase 'sub-installations' ('circuits' if you wish) originating from the meter, and I can see no reason why an RCD installed solely to give fault protection to one pair of conductors in a distribution circuit needs to disconnect conductors associated with totally different 'sub-installations'/'circuits'. Put in more regulatory terms, my DP RCDs do "disconnect all live conductors of the circuit" (at substantially the same time), as required by 531.2.1.

Do you disagree?

Kind Regards, John

 

[mfarrow]

No, I don't, and deliberately not - for the reasons I explained. There are there separate TD DP RCDs.

Fair enough, I was confused by your point of distributing "two phases". I thought you meant 'together'. As you don't then you don't need an RCD.

I was not attempting to start a debate around your own installation. If you have split your phases at the origin into three circuits then, they are 3 circuits.

Assuming you mean all four of the conductors (3P+N), that is not my interpretation

Then you assume incorrectly. Neutral is not a line conductor.

My main point was the observation of the IETs regulations, which I was trying to remember earlier, and also their (deliberate?) lack of requirement of requirements for RCDs provided for Additional Protection.

I'm still trying to decide why the disconnection of all lines is required for an RCD anyway, instead of just the phase affected. Perhaps it's in the lack of available distinction in a 3-phase RCD regarding which phase is leaking. Answers on a postcard?

 

[JohnW2]

No, I don't, and deliberately not - for the reasons I explained. There are there separate TD DP RCDs.

Fair enough, I was confused by your point of distributing "two phases". I thought you meant 'together'. As you don't then you don't need an RCD.

I'm glad you agree. Sorry if I confused you by saying that I needed TD RCD protection "for at least two of my phases". The point is that one of the phases goes to a CU only a couple of metres from the meter, whereas the other two have long (separate) distribution circuits, heading in different directions, which need protecting. In reality, I've got TD RCDs on all three ('unnecessarily' on one), not the least because I use them as isolators!

I was not attempting to start a debate around your own installation. If you have split your phases at the origin into three circuits then, they are 3 circuits.

Quite so.

Assuming you mean all four of the conductors (3P+N), that is not my interpretation

Then you assume incorrectly. Neutral is not a line conductor.

Apologies, I misread the reg - I read it as 'all live conductors'. However, that doesn't alter the point I was making.

I'm still trying to decide why the disconnection of all lines is required for an RCD anyway, instead of just the phase affected.

Are they not thinking of '3-phase circuits', with 3-phase loads? In that case, one can't really talk about faults affected just one phase, or for it to be safe to disconnect just one phase in the case of a fault, can one?

Kind Regards, John

 

[mfarrow]

Are they not thinking of '3-phase circuits', with 3-phase loads? In that case, one can't really talk about faults affected just one phase, or for it to be safe to disconnect just one phase in the case of a fault, can one?

Yes, in the case of a 3-phase load, it seems quite logical.

However there are instances, because of budget/time where this ruling does not make sense, and can impede the integrity of the design.

A job I worked on awhile ago required the installation of lighting in 3-phase circuits (to reduce cabling) off a TT supply. This was an industrial installation covering several hundred meters and a proprietary cable harness system. Each row of lights was split and staggered into the 3 phases. We had installed an isolator for each row (to disconnect all line conductors [reg somewhere]), and an MCB for each phase of each circuit. Thus if there was a L-N fault only one phase, of one circuit, in one row would be affected. However 531.2.1 would not allow the same logic to be applied to the RCDs, so under earth fault conditions an entire row of lighting would be lost.

Now, I'm not going to go into what we actually did (there are some politics involved) but suffice to say it's difficult to get a contractor to agree to ignoring a regulation without raising the CDM 'who is the designer now?' type discussion.