Isolation of TN supplies.

[nicebutdim]

I haven't seen a lot of discussion on this subject and have questions pertaining to isolation where TN earthing is present and any peculiarites which may exist between TN-S and TNC-S.

I see mostly 3 pole main switches on TN earthing systems, but do see quite a few instances where all live conductors are switched.

Furthermore, where loads are isolated what consideration should be given to this? I'm asking specifically in relation to the post linked below. It seems to be that the only time I see a 3P rotary isolator is on motor circuits, with 3P+N isolation being used in pretty much every other situation.

There's a gap in my learning here and it's one I'd like to resolve.

Low insulation resistance reading

4 pole main switches are fairly common here, but the point was more about the fact we're free to use them, whereas I believe they can not be used in IE. I wouldn't say free to use, you would only switch the neutral when absolutely necessary. Switching the neutral introduces a greater risk of...

www.electriciansforums.net

 

[pc1966]

Isolation and switching are different aspects. One example where you need 4P switching is a generator transfer switch where the two supplies are on differently earthed neutrals, so you need to break one while running off the other.

I'm not sure of why the UK regs differ in terms of mandating DP isolation for "ordinary user" SP supply isolation while being happy with 3P-only for a TP supply, maybe it is less about the risks of the DNO neutral being at elevated potential (open PEN faults a different story) and more due to the relatively specialised nature of TP installations where less DIY bodgery is anticipated, and where a mistake in swapping N for one of the L is more obvious to the point of equipment-damage-smokin' obvious.

 

[davesparks]

I've noticed a trend towards fitting 4 pole main switches in TPN boards where 3 pole was always the norm.

I've asked a few people why they are fitting them and often the answers are along the lines of "Just in case" or "cos it's safer" but they can never explain what it is incase of or how it is safer.

I would have to check but I'm sure that in the past the regulations absolutely forbade the use of a 4 pole switch and only permitted a bolted test link in the neutral of a 3 phase supply.

 

[pc1966]

Another practical aspect to using a 4P isolator is it simplified any IR testing (part of original thread I think...) as you are not likely to energise the load at all when doing a global IR check, or if you are testing the load on its own you can do so without needing to disconnecting a terminal somewhere (N link in isolator, or maybe at DB N-bar).

 

[davesparks]

I'm not sure of why the UK regs differ in terms of mandating DP isolation for "ordinary user" SP supply isolation while being happy with 3P-only for a TP supply,

I wouldn't be surprised if it's simply because we like to cling to the past and have just kept the spirit of the old rule which required DP switching and DP fusing in the days when the supply neutral was not guaranteed to be reliably earthed.

 

[pc1966]

My own experience is very limited compared to many folks on here, so I can only really say why I used 4P DB isolators and that is largely as we have up-front RCD for fire risk reasons on a TN-S supply (so similar to TT in that respect) and it allows me to work on a sub-DB without risking a N-E short tripping the incomer. Also it simplified SPD testing for the N-E module, although I could probably pull it and put the IR tester on the terminals directly.

Probably quite an unusual case!

 

[timhoward]

I would have to check but I'm sure that in the past the regulations absolutely forbade the use of a 4 pole switch and only permitted a bolted test link in the neutral of a 3 phase supply.

The earliest I have is 14th edition and it was either/or
For TP isolation it offers either all conductors or live conductors only and a bolted link for the 'conductor connected with earth'.
(in that edition an earthed neutral conductor was not defined as a "live conductor".)

 

[westward10]

Could be a polarity issue which is not uncommon for single phase whereby if it is incorrect then double pole isolates regardless. Don't think I've ever seen reverse polarity on a three phase board.

 

[pc1966]

Don't think I've ever seen reverse polarity on a three phase board.

There is not quite the same thing as "reverse polarity" as you can have:

• L-L swap leading to wrong rotation. This is not obvious without specific test (closer to SP reversal) but will likely trouble machines using 3P motors as fan direction reversed, etc, or
• N-L swap and then almost certainly see SP loads on the two other L-(L instead of N) voltages burning out!

The first could go undetected if you only have SP loads, but the 2nd is likely to be found out pretty quickly so not likely to be discovered on EICR, etc!

 

[SparkySy]

I had a situation not so long ago with a 4 pole main isolator where it was switched off for some equipment to be changed.
When it was switched back on the neutral did not close resulting in the voltage on the single phase control side going crazy as the voltage raised and cooking some of the single phase equipment.

 

[R-fur]

It seems to be that the only time I see a 3P rotary isolator is on motor circuits, with 3P+N isolation being used in pretty much every other situation.

3 Phase motors rarely have a neutral so 3P is fine.

 

[7029 dave]

When I was metering on 3 phase, We never used 4pole isolators.

 

[nicebutdim]

Isolation and switching are different aspects. One example where you need 4P switching is a generator transfer switch where the two supplies are on differently earthed neutrals, so you need to break one while running off the other.

I'm not sure of why the UK regs differ in terms of mandating DP isolation for "ordinary user" SP supply isolation while being happy with 3P-only for a TP supply, maybe it is less about the risks of the DNO neutral being at elevated potential (open PEN faults a different story) and more due to the relatively specialised nature of TP installations where less DIY bodgery is anticipated, and where a mistake in swapping N for one of the L is more obvious to the point of equipment-damage-smokin' obvious.

I'm thinking solely about isolation and you raise some of the points which brought this issue to the front of my mind.

I've noticed a trend towards fitting 4 pole main switches in TPN boards where 3 pole was always the norm.

I've asked a few people why they are fitting them and often the answers are along the lines of "Just in case" or "cos it's safer" but they can never explain what it is incase of or how it is safer.

I would have to check but I'm sure that in the past the regulations absolutely forbade the use of a 4 pole switch and only permitted a bolted test link in the neutral of a 3 phase supply.

Little irks me more than the "this is how we've always done it" response. I don't expect people to have all the answers, but "I don't know, let me look into it" would be better than folks blindly following rote.

Although this thread hasn't provided any empirical answers, I'm glad it's not just me that couldn't come up with a definite line of thinking.

Another practical aspect to using a 4P isolator is it simplified any IR testing (part of original thread I think...) as you are not likely to energise the load at all when doing a global IR check, or if you are testing the load on its own you can do so without needing to disconnecting a terminal somewhere (N link in isolator, or maybe at DB N-bar).

Testing is the one reason why I suspected 4 pole isolation is routinely used for equipment, although it could equally be a case of wholesalers keeping 4 pole isolators or people following what they've seen elsewhere. For any equipment which requires a neutral, I would never have considered not providing isolation of all live conductors.

I wouldn't be surprised if it's simply because we like to cling to the past and have just kept the spirit of the old rule which required DP switching and DP fusing in the days when the supply neutral was not guaranteed to be reliably earthed.

I'm inclined to consider 4 pole isolation useful for equipment and the convenience where testing is concerned probably outweighs any potential faults with the isolator. Of course I might now think differently where excceptionally expensive equipment is connected.

3 Phase motors rarely have a neutral so 3P is fine.

Yep, although I often see 5 cores used on motor circuits as the was on hand and neutral ends up parked in a connector.

I had a situation not so long ago with a 4 pole main isolator where it was switched off for some equipment to be changed.
When it was switched back on the neutral did not close resulting in the voltage on the single phase control side going crazy as the voltage raised and cooking some of the single phase equipment.

This is what I'm curious to know. Has anyone else encounted a similarly faulty 4 pole main switch or are such occurrences likely to be extremely rare?


An isolator can of course fail, but neutral links allow for the possibility of human error, should one need to be removed for testing...

 

[pc1966]

This is what I'm curious to know. Has anyone else encountered a similarly faulty 4 pole main switch or are such occurrences likely to be extremely rare?

I have seen a few occasions when a switch that had not been used for donkey's years was operated and failed as the contacts were poor/overheated due to dirt or tarnishing. I guess that might be a common risk for items that are hardly ever operated except for very rare jobs, like an isolator switch.

I'm currently looking at ATS switches and some recommend you test every 3 months, not sure if this is due to risk of control failing or to keep the contacts clean, either way you find out if it is working before you need it for real!

An isolator can of course fail, but neutral links allow for the possibility of human error, should one need to be removed for testing...

Both have risks to them, I guess it is hard to know what is more likely, but folks getting distracted and forgetting to fit (or properly tighten) a N link is one I would bet on.

For normal switching then if you have no reason to go over 3P you avoid any risk of momentary open-N over-voltage (or finding the rarer switch that has a make-first, break-last N contact) and save a little on cost as well.

 

[7029 dave]

The brand was wylex cant find any on the internet, must of made them exclusive to the company's specification.
LLL switched with N being of solid link

 

[MDJ]

I think switching 3 phase and neutral at a rotary isolator controlling an outside air con unit for example is perfectly normal, but fitting a 4 pole isolator or main switch in a DB which is 3 phase but has single phase circuits a no, no, I have personally seen a neutral break on a 3 phase DB which was controlled via a contactor and had a stop circuit controlling the coil and when the circuit energised so the board would become live the neutral hung and 400 volts went across all circuits in the board and tens of thousands of pounds of connected equipment burned out, when I was an apprentice it was not the norm to switch a neutral in a 3 phase system with single phase circuits connected so that is how I was taught and learned. I however love to hear other sparks opinions and thank the OP for staring this thread.

 

[nicebutdim]

This is certainly a broader conversation than I'd anticipated.

I've never been involved in wiring any transfer switches for generators, but have been involved with a few fairly sizeable manual UPS changeover switches. The auto side of things I know little about as it was handled within the UPS, but from memory the main changeover switched all 4 poles and the UPS output side did the same, but I think that neutral was monitored. On the other side I'm fairly certain the rectifier and bypass isolators switched only the three lines and had links botled through neutral.


Edit: We can probably all agree that where functional switching is concerned it would generally be a bad idea to switch neutral.

 

[pc1966]

I think larger & good quality UPS link the output N & E when on-battery, so they would need 4P switching as input N and output N cannot be assumed to be the same. However, I guess the batteries and inverter are always referenced to the output side N so only need the L switched to safely isolate for any maintenance work.

The ones I have used in the 5kVA region usually have both + & - of the battery pack on a connector so you isolate both before withdrawing the tray of batteries, then they are all isolated from the metalwork and a bit less dangerous to handle. I still don't like handling them!

 

[nicebutdim]

I had been thinking about these after you mentioned generator transfer switches and posted from memory, but since found an image which confirms those memories.

Not sure what size each UPS was. Cable size I think was 70 or 95mm2.

 

[pc1966]

Not sure what size each UPS was. Cable size I think was 70 or 95mm2.

Sound like something in the 100kVA range.

 

[westward10]

There is not quite the same thing as "reverse polarity" as you can have:

• L-L swap leading to wrong rotation. This is not obvious without specific test (closer to SP reversal) but will likely trouble machines using 3P motors as fan direction reversed, etc, or
• N-L swap and then almost certainly see SP loads on the two other L-(L instead of N) voltages burning out!

The first could go undetected if you only have SP loads, but the 2nd is likely to be found out pretty quickly so not likely to be discovered on EICR, etc!

Not really sure how this is overly relevant to what I suggested.

 

[timhoward]

Yep, although I often see 5 cores used on motor circuits as the was on hand and neutral ends up parked in a connector.

In my first sparking era (late 90's) I was always told to run a neutral to the isolator, as the chances are whoever is using the machine will want a socket and a work light next week, especially in the case of lathes.
But if a socket was added and the N came into play we certainly didn't swap to 4 pole isolators!
I'd imagine RCD requirements prevent this arrangement these days anyway.

The only real benefit to TP+N isolation seems to be easier testing. At least I think it's easier to know that everything upstream is isolated and you aren't finding upstream N-E issues, especially when going through the IR permutations.