electricalforum.nz

There is a statement in the April 2019 EV WorkSafe guideline that I don’t understand. Page 8, “All RCds should” (a) ___________ ??? . I understand what (b) is saying.

Any thoughts?

Assuming you're referring to clause 1.4 if 2019 edition(please use clause number rather than page number);

a) & b) are two options, both relating to d.c in the fault current waveform.
Either the RCD must switch off if > 6 mA of d.c, or it must be of a Type that can continue to provide protection despite > 6 mA of d.c..

Basically d.c, even at such low levels, can mean that the RCD simply doesn't work when we would expect it to.
This is because RCDs are elctro-mechanical, and (simplistically) they use the current generated by the toroid to overcome a magnetic latch
With d.c., the current can either (depending on the polarity) reinforce rather than overcome the magnet ie locking the RCD "on"

I believe the 6 mA level is the limit for Type A, above that they can't be relied on.
Hence Type B being preferred for EV charging (and other cases where significant d.c. can be expected), but Type A plus additional residual d.c. detection device (RDC-DD) being acceptable.

there is some guidance on RCD selection in 2.6.2.2 of "3000" (Note 1)
As per Note 2; RCDs always have to be selected for the expected fault current waveform.

Thanks Alec

For years we in NZ have been a bit prideful that we use Type A rather than the Type AC that is acceptable in Oz.
Type AC only manages pure a.c waveforms; whereas Type A can handle pulsing d.c. as well.
For example, a simple hair dryer speed control using a diode (half-wave rectification) has a pulsing d.c waveform in both load current and earth fault current
Oz are now starting to catch up; and starting to specify Type A for some situations.
But at same time we're finding that widespread use of electronics in VFDs , switch mode power supplies, and LED lighting controls is reducing the effectiveness of Type A.
For EV charging in particular, there high chance of flat (as against pulsing) d.c. being superimposed on the (nominally but possibly severely distorted) a.c. waveform for an earth fault current, hence the increased requirement.

A Type A can handle up to 6 mA of d.c.; but with EV charging there could easily be more, meaning the Type A may not trip - and someone could die as a result.

There's a somewhat dated but very useful handbook published by Standards Australia (HB 113: 1998) that explains a lot about the internal workings of RCDs.
Most 'nuisance tripping" is due to poor selection and or incorrect installation.
But worse are the things that lead to the RCD not tripping when it should.
One of the common reasons for this type of failure is simply that the RCD hasn't had its button pressed for so long that the mechanical side has seized up.
Another is wrong type for the likely wave-form.

I was also in need of enlightenment regarding this. Thanks! Just a mediocre in our agency.







Frey Karts
North Shore
electrician north shore auckland

Interesting, NZ uses type A and AC. In the US we use level 1, 2, 3. Is that comparable? For example, the most common level 2 is 16 and 30 amps.

NZ requirements are based on IEC requirements. The performance requirements for various Types of RCD are set by AS/NZS 61008.1 & 61009.1 (RCCBs & RCBOs); which are based on same-number IEC Standards.
They specify parameters for each "letter" type WRT wave-form of the fault current.

Also AS/NZS 3190 brings in a number of other classifications; including "numeral" Types, WRT both nominal operating residual current and required speed of operation.
Eg we use 10 mA Type A for both schools and and medical; but the ones in medical must be Type I (roman 1) so not only can't be > 10 mA but also must operate faster than "general" 10 mA Type A as used for schools.

In NZ all mandatory RCDs have always had to be capable of operating on pulsating d.c - making Type AC non-compliant here.
Type AC has been the norm in Australia; they are now changing to make Type A their minimum specification - but it will take a while.

I'm unsure whether there's any direct relationship between US GFIs & IEC RCDs.
I presume those levels refer to load current; in which case there will be no comparison.
US practice generally has higher current for a given load due to lower nominal voltage.
That lower voltage will also affect the prospective fault current - as will the lower EFLI due to larger conductors.

But none of that affects how our RCDs work. They are designed for the expected waveform of the fault current; and (generally) to operate at low enough level of earth fault current , and quickly enough to avoid a shock causing heart fibrillation.
The 10 mA for schools is different; it's to avoid risk of suffocation die to inability to "let go" even at a level too low to cause fibrillation.