RCD Nuisance Trips - Which Test?

[ericmark]

One would hope that capacitors 'rated' for 230-250V AC would probably tolerate 500V DC without being damaged, but I'm not sure that one could necessarily rely on that. I personally would try to take any such capacitors out of circuit before undertaking a ≥500V IR test.

Kind Regards, John

I seem to remember regulations say they must, but not got them to hand.

 

[333rocky333]

What I am inclined to add is that I would have thought that an IR problem in the fixed wiring would be a very unusual/unlikely cause of occasional 'nuisance RCD trips' - and that, in ohe rare occasions it did happen, IR testing would quite probably not help, since one would probably not undertake the testing at a time of manifestation of an intermittent IR problem!

Kind Regards, John

 

[333rocky333]

When we did pubs we used to find low resistance on circuits, that were still running ok, but when large loads like the celler cooler kicked in the Rcds used to trip on the split load boards, I quess it was the current diverting elsewhere and not back through the relevant Neutral, bit like what Bernard often mentions, a lot of the other engineers used to blame the cooler and when there engineer came and checked there stuff, it was fine and sometimes even brand new.

Sometimes we used to use the Rcd ramp test to compare the loaded Rcd against it unloaded.

 

[JohnW2]

When we did pubs we used to find low resistance on circuits, that were still running ok, but when large loads like the celler cooler kicked in the Rcds used to trip on the split load boards, I quess it was the current diverting elsewhere and not back through the relevant Neutral, bit like what Bernard often mentions, a lot of the other engineers used to blame the cooler and when there engineer came and checked there stuff, it was fine and sometimes even brand new.

When you say 'low resistance', I presume that you mean at least 1 MΩ? If so, that would only make a very marginal difference to whether or not an RCD would trip when additional leaks came into play, since it would probably result in no more than about 0.2 mA of residual ('leakage') current at most - hence, if the RCD were going to trip at, say, 20 mA of leakage current, the cooler would have to result in about 19.8 mA of that for the RCD to trip.

Kind Regards, John

 

[nitro23456]

Ok, borrowed a clamp meter.....is there an acceptable earth leak mA figure for an individual circuit?

E.g. my oven/hob radial sits at 6.6mA but 0.3 with wall DP switch off. Seems quite high?

Downstairs lights circuit is at 1.2

Everthing going through the RCD adds up to 18, so not far off a potential trip threshold of 22-30.

Is this sort of low level individual leakage acceptable and if so i suppose my only option is to get an electrician to separate these circuits out by replacing with RCBO's?

Doesnt seem there would be much to gain from an IR test now? (from my point of view)

 

[DetlefSchmitz]

When you are evaluating the numbers, don't forget that a 30mA RCD is permitted to trip at any current between 15 and 30mA. I've seen figures for cookers quoted as 0.75mA per kW max.

 

[nitro23456]

Any figures for lights?

The oven circuit is about 11kw with the oven and induction hob so that would check out then.

I think a new RCBO board is probably due but i wouldnt want this to just mask an underlying problem.

 

[DetlefSchmitz]

Most figures available are from PAT sites, earth leakage being one of the required tests. So fixed lighting isn't something they give numbers for. However 1.2mA seems reasonable.

 

[bernardgreen]

The clamp meter will be measuring the AC leakage current and this will include the currents created by the capacitive coupling between Live and Earth and Live and Neutral. The amount of current depends on how much cable is involved.

Insulation resistance measures are performed with DC and therefore there is no continuous current created by the capacitive effect in cables.

 

[JohnW2]

The clamp meter will be measuring the AC leakage current and this will include the currents created by the capacitive coupling between Live and Earth and Live and Neutral. The amount of current depends on how much cable is involved.

Very true - but in this case it sounds as if many of the measurements were undertaken with loads connected, so the majority of the total (AC) 'leakage' current measured was probably attributable to filter capacitors etc., rather than the (small) capacitance of the cables.

Insulation resistance measures are performed with DC and therefore there is no continuous current created by the capacitive effect in cables.

Indeed so - and that's why the (AC) leakage currents observed above, and said to be 'expected'/acceptable' are far higher than would be acceptable were they resistive 'leakages' (i.e. as measurable at DC) of wiring. For example, the "0.75mA per kW max" quoted for cookers would, if measured at DC on wiring, be well under 1MΩ for even a small cooker, and far lower for large cookers.

That having been said, when investigating RCD trips measurements of AC leakage currents are clearly more valuable than DC 'insulation resistance' measurements, since it is imbalance of AC current which causes RCDs to trip - and, of course, one can measure the all-important imbalance when the circuit is energised and loads are connected. In fact, if an adequate means of measuring AC 'leakage' is available, I don't think that DC IR measurements really bring much, if anything, more to the table. Even if the (very rare) situation of RCD trips resulting from resistive ';leaks' within a cable, that will bee seen as an AC leakage current.

Kind Regards, John

 

[nitro23456]

Just to tie this off for me, with 18th ed whats are the options for CU changes..... does everything have to be on a RCBO now or a split load dual RCD boards still an option? Anything else?

 

[JohnW2]

Just to tie this off for me, with 18th ed whats are the options for CU changes..... does everything have to be on a RCBO now or a split load dual RCD boards still an option?

Most circuits now need RCD or RCBO protection (certainly all sockets circuits, lighting circuits and circuits supplying bathrooms), as do circuits with any ('new') ordinary cable buried ≤5cm deep in walls - so, essentially, everything. However, that can be done with a dual RCD board, not necessarily RCBOs (although there are obviously 'preference' arguments for the latter).

Anything else?

It has to be made of 'non-combustible material' (even though such a material probably doesn't exist) or enclosed in a 'non-combustible' (same comment) enclosure. In practice, that essentially means that it has to have a steel case - since that's probably all you can now buy.

Kind Regards, John

 

[nitro23456]

Ballpark figure for a sparky to supply and install an RCBO board?....

5x 32a rings
1x 32a radial
3x 6a light circuits.

I know itl vary, but just a very general ballpark figure would be most helpful