Rcd tripping - very odd

[bri8384]

Hi all not posted for a long time but found this place very useful in the past so thought I would give it a try.

been working as a qualified sparky/maintenance engineer now for almost a year since I qualified and now have my first real problem, was hoping for some advice/reference to certain regs if possible.

so we have an American made packaging machine which is 3 phase + earth (no neutral) from the mains supply through a 3 pole isolator down to a transformer, it's as this point a neutral is picked up from the transformer star point, isolator is 32amp and all single core 6mm in wall and supply cable, this seems fairly common with a lot of new kit we are getting at the minute.

the problem with this particular machine seems to be their are also 9 servo drives within the machine and when powering up the machine it trips on the 100mA in the roof void, once reset a few times it will hold and run without any issue until the next time it's powered down.

As I'm still learning I don't fully understand why this is tripping an rcd when there's no neutral at this point in the circuit?

also as their is no neutral/ less than a 25m run of cable, machine used by trained operatives would it be acceptable to remove the rcd protection?

As best I can understand it, this trips due to the start up current and 9 servo drives 'fluttering' when booting up but as their is no neutral core at this point in the circuit an rcd seems pointless as it would never pick up a neutral fault in the first place.

Am I missing something obvious?

 

[Marvo]

So you've got a 3-phase supply from a DB which supplies a transformer which in turn supplies the machine via a 3/pole isolator? Is that right?

Is the RCD on the supply between the transformer and the machine? Maybe you could attach a sketch?

 

[bri8384]

So you've got a 3-phase supply from a DB which supplies a transformer which in turn supplies the machine via a 3/pole isolator? Is that right?

Is the RCD on the supply between the transformer and the machine? Maybe you could attach a sketch?

3 phase supply from sub board (with mcb and desperate rcd) to isolator on wall, then plug/cable to machine isolator, through this isolator (still no neutral at this point) then into transformer - at this point we get a neutral and goes through machines internal rcd. I'll try doing a sketch but drawing is not my strong point lol

 

[Marvo]

No probs, I've got the picture now but just to confirm it's the triple pole RCD in the SubDB that's tripping?

 

[bri8384]

No probs, I've got the picture now but just to confirm it's the triple pole RCD in the SubDB that's tripping?

Correct 100mA C type

 

[Marvo]

What tests have been done? RCD ramp test? Transformer IR test? Circuit tests?

The RCD will only see a leakage fault from the SubDB up to the transformer primary windings, and fault on the secondary or beyond won't be detected so the fauit(s) should be fairly straight forward to locate. You could also clamp the three phases simultaneously with an earth leakage clamp meter and monitor the leakage in real time. A clamp meter with a decent sample-rate and a max-hold facility would be best.

It's also possible that the transformer could be causing a brief phase angle anomaly on initial power-up which the RCD wouldn't be able to distinguish from a leakage fault. Just out of interest why is the circuit supplied via an RCD in the first place? I'm assuming the RCD isn't time delayed....maybe consider replacing with an S type....???.

 

[bri8384]

What tests have been done? RCD ramp test? Transformer IR test? Circuit tests?

The RCD will only see a leakage fault from the SubDB up to the transformer primary windings, and fault on the secondary or beyond won't be detected so the fauit(s) should be fairly straight forward to locate. You could also clamp the three phases simultaneously with an earth leakage clamp meter and monitor the leakage in real time. A clamp meter with a decent sample-rate and a max-hold facility would be best.

It's also possible that the transformer could be causing a brief phase angle anomaly on initial power-up which the RCD wouldn't be able to distinguish from a leakage fault. Just out of interest why is the circuit supplied via an RCD in the first place? I'm assuming the RCD isn't time delayed....maybe consider replacing with an S type....???.

All tests u mentioned come back ok, we have a clamp meter with monitoring/readout ability so will try this in the morning, my thinking is the same as yours (although I didn't know how to type it so well so thanks) yes s type would be the best option so will probably order one in the morning - the rcd was already in the cct with previous machine that was fed from this supply, from what your just said about why is their an rcd in cct does that mean we could get away without it bearing in mind regs/health and safety?

 

[Marvo]

Removing the RCD (and replacing it with an 'S' type for that matter) would really require confirming why it was installed in the first place followed by a risk assessment etc. I'm not sure about UK regs (my location is elsewhere) so I can't really comment on that aspect but there could also be in-house requirements of that particular company to consider.

 

[bri8384]

Removing the RCD (and replacing it with an 'S' type for that matter) would really require confirming why it was installed in the first place followed by a risk assessment etc. I'm not sure about UK regs (my location is elsewhere) so I can't really comment on that aspect but there could also be in-house requirements of that particular company to consider.

Sorry if I wasn't clear but the rcd was installed as standard, every cct on site has rcd protection, normally 30mA on the factory floor (integrated into the isolator) and then 100mA upstairs. I was considering just removing the rcd or just moving it downstairs as part of the isolator just so it could be reset easily until it holds in, I know this is not the best option and the regs I can see are fairly open to interpretation as per normal. I really do think your onto something with the anomalous phase angle tho, it's the only thing that makes sense to me.

my only real issue is that this is the only place I've worked as an electrician and have always known every cct to have rcd protection so not really sure if it can just be taken out/ other places don't use it.

as for risk assessments, my firm is crazy about them so naturally these are carried out for any job we don't already have a generic me in place for, can't even climb a ladder without risk assessment and a permit being completed lol

 

[mache]

Just an observation from a non electrician so this is not advice but just a query really. As I understand it company policy is for RCD protection on all circuits as mandatory safety measure. However my impression is that without a neutral the RCD would not function correctly as a safety device. If I am correct (and I'm happy for qualified people to correct this) then the RCD is giving a false impression of safety in that section of circuit and in such a scenario should be removed and some other method employed if possible. As a minimum clear and simple labelling should be employed.

Happy for people to correct my understanding of RCD's

 

[davesparks]

As long as the RCD is correctly connected for three phase no neutral operation then it will perform its function perfectly well. If it did not then three phase RCDs would be completely pointless unless the circuit was 100% out of balance (only one phase being used)

 

[Lucien Nunes]

my impression is that without a neutral the RCD would not function correctly as a safety device

I think you may be confusing two very different scenarios that sound similar:
*A 3-phase circuit without a neutral (i.e. the circuit under discussion).
*A 3-phase circuit where the supply neutral point is not earthed.

In the first case, all the circuit conductors (which happen to be three lines) go through the RCD and that is all an RCD requires. Had there been a neutral, that would have had to go through it as well. RCDs don't care whether conductors are lines or neutrals and how many of each there are, provided they all go through it. Any current that goes out through one wire but doesn't come back through another must be leakage which returns to the source of supply via the CPC or body of earth, bypassing the RCD, causing an imbalance and hence a trip. This return path is possible because the neutral is earthed at the source, regardless of whether it is required in this particular circuit.

In the alternative scenario, known as an IT supply, the entire system floats with respect to earth, i.e. the connection that is usually made between N & E at the substation transformer is not present. This configuration is only used for special circumstances and not for public supply. An RCD on an IT circuit would theoretically still detect leakage, but single faults to earth will not cause leakage (as there's no return path from earth to the circuit at the supply end) so the RCD won't detect those faults and won't trip.

 

[Bobster]

Why is there a need to fit an RCD for this installation?

Edit: Just seen you mention that it's company policy.

 

[snowhead]

My first thought would be to contact the manufacturer and see what the normal leakage is and whether they recommend an RCD and what type.

 

[Marvo]

The arrangement he's got at the moment, as I understand it, the RCD is only protecting the supply from the subDB and the transformer primary side so it's not fulfilling any manufacturers requirements for the machine on the secondary side of the Tx.