RCD Tripping Causes?

[Paulusgnome]

Greetings all,

I am engineer for a distributor of switchboards and components for same. One of my functions is QC.

We have accumulated a pile of RCDs (GFIs if you are American) which have been returned to us as faulty by numerous electricians who have found that they are mostly nuisance-tripping.

When I test these with a tester, I find that nearly 3/4 are not faulty at all.

Most strange. My thinking at present is that the predominant cause must be related to the load or to a particular type of load.

My question then, to all you fine experts, is what load types are known to cause nuisance tripping of RCDs, particularly in domestic applications?

We know about refrigerators and freezers. How about HID or fluoro lamps? Computer zap-catchers? EMC by-pass caps in electronic appliances?

 

[Bung]

Electric stove plates are also a problem (in those areas where stove circuits are on RCD). This is particularly true in areas with high humidity levels. I guess that wiring made up of a conductor inside a copper tube with a powder insulation (I forget the name of the stuff) might also be an issue if it gets damp. I remember it giving problems in my (long distant) past in industry, especially if bent too sharply.

Bung
Life is non-linear...

 

[imok2]

My thoughts were focused on a reliable voltage supply.

 

[jbartos]

Suggestion: Electronic devices that have grounded chassis may cause the higher ground leak. Also, an electrical installation itself may cause a ground leak, if it is old one.

 

[rbulsara]

Misapplications include, but not limites to, the following:

1. electronic devices as jbartos suggested.
2. Long circuits (circuit's capacitive leakge currents trip personnel GFIs - 3ma to 6 ma range).
3. GFIs applied to HID lighting circuits.
personnel GFI applied to equipment or items like heat tracing (where a less sensitive 30-300mA GFIs are more suitable).

4. It would be interesting to gather the info of where the 'good' GFI reported as bad ones were applied!

 

[sylim]

In the case of domestic, the sudden switching on(energization) of a great number of fluroescent lights, electronic devices with switched-mode power supplies are especially prone to nuisance tripping in the aftermath of a power failure.

Also any appliance that generates higher order harmonics have a tendency to cause nuisance tripping.
Remember, the RCD normally work well in 50/60Hz power system, once the load current contains of higher order harmonice, it tend to increase its sensitivity, ie causing unwanted tripping.

 

[Skogsgurra]

Paulusgnome,

I have measured ground currents in frequency inverters and I have found up to 500 mA RMS currents flowing. Most of the current is PWM edges that go through motor capacitance to ground.

The motor capacitance is usually between 1 and 10 nanofarads and since the high frequency components in the PWM motor voltage go well above 1 MHz a lot of current goes that way.

A screened motor cable will also produce ground currents - more if the cable is long.

A du/dt filter that diverts the high frequency components to ground will also increase ground fault current.

It is more and more obvious, and many people already has made it a house rule, that RCDs cannot be used when there are variable frequency drives in the installation. And do not try to use a 300 mA device! It will not save lives and it will not prevent fires. Only 30 mA devices (or lower) save lives.

 

[Ahr35181]

Paulusgnome,

Getting back to the root problem, i.e your pile of returned RCD's. Over 25% failure of the units tested by your own company apears very high for a life saving device. When you tested them, did they fail on the 'Nuisance Tripping Test'? i.e half times rated tripping current? How did they fare on the 'Rated tripping current' test & the '5 Times tripping current.The ones that past the nuisance tripping test, by which tolerance did they pass? (e.g did they trip at 52% rated tripping current)

Just as a matter of interest, before your RCD's go on the market do you test them?

Alan

P.S What make are the RCD's?

 

[Paulusgnome]

I have now tested 38 of approximately 120 returns, all of which have supposedly been found to be faulty by the electrician who installed them and then tested the installation.
12 of these were found to be actually faulty when tested.
4 were physically damaged - broken toggle, case etc.
5 would not latch on.
3 would not trip.

We do not test each RCD before it is sold, but our manufacturer certainly does. Site testing before commissioning by a licensed electrician is mandatory here.

I would love to disclose the identity of said manufacturer, but my desire to remain employed is somewhat stronger.

Thanks to all those who have replied, the info is very helpful.

Later,
Mark aka Paulus

 

[jbartos]

Suggestion: The good GFIs may have been tripped by transients. There are transient ground currents. Also, the ground currents having a large harmonic content can trip the GFI.

 

[jghrist]

A 25% failure rate of returned devices does not necessarily mean a high rate of failure overall. It depends on how many of the devices sold were returned.

 

[Shortstub]

Bung,
You are thinking of MI or Mineral Insulated. Invented by the French it is also known as Pyrotenax!