Tesla Gen.3 Wall Charger

[Slovett]

Hi,

I've just installed a Gen.3 TESLA Wall Charger.

32A Circuit with Type A RCD Protection at SwitchBoard (Wall Charger has Built in RDC-DD Protection)

I know everything about EV chargers are just guidance at this stage, but the way WorkSafe seems to be investigating these days, it's better to be safe than sorry.

During the Test to insure the RDC-DD operates - it didn't trip until it reached 12mA DC current.

3000:2018 - 7.9.3.3 (c) NOTE says that the RDC-DD disconnects the supply if currents of 6mA or greater are detected. But as far as I know there is no mandate or guidance as to what a Fail or Electrically unsafe test result is.

The Wall charger comes with a SDoC but does not state the IEC62955 Standard for RDC-DD requirements.

What results are others getting?

Im guessing the RDC-DD tripping @ 12mA is still safe, but unsure why 3000 has stated 6mA in the first place?

[AlecK]

All Notes to clauses are intended to provide additional information (as explained in the Foreword).
They do not set requirements.
This one relates to the reason that we don't use 'normal' Type A RCDs for EV supplies;
which is that with a rectifying load such as an EV charger there are very likely to be d.c. currents above 6 mA in the waveform of the residual (fault) current.
It was added to explain why adding an RDCDD to a Type A is deemed acceptable, while a Type A on its own isn't.

Notes to Clause 2.6.2.2.1 provide a (very) basic overview of the capabilities of various Types of RCD.
A Type A is designed to operate on sinusoidal waveform, and on pulsing d.c. waveform (as results from half-wave rectification).
It's known that Type A cannot be relied on to operate cirrectly if there's more than 6 mA of constant d.c in the residual current waveform.
The operating mechanism includes a magnetic coil. Basically the a.c. residual current results in a magnetic force to operate the trip mechanism. But if there's significant d.c this can create an opposite magnetic force.
My understanding is that a Type A will generally be reliable up to 6 mA of (non-pulsing) d.c component;
but I can't be 100% on that because operating in presence of straight d.c components is not a requirement for Type A.

Hence the requirement, for EV circuits, to either
a) use a Type B RCD (which is designed to be reliable in presence of d.c. components in the waveform); or
b) add an RDCDD in conjunction with a Type A. The operating parameters for the RDCDD will be set by the product Standard.
Note that using an RDCDD is only valid if all conditions for the Exception are met. Including that the RDCDD complies to the IEC Standard cited.

Even if the SDoC for this charger did specify compliance if the RDCDD with that IEC 62955, it still wouldn't meet the requirement of the clause.
The clause requires the subcircuit to be protected ; and you can't provide RCD protection at the load end of the circuit - it has to be at origin.

If you look at the EV charging guidelines published by Worksafe (2019); clause 1.4 similarly requires TCD protection of the subcircuit;
and advises that the RCD should either:
a. incorporate the ability to continue to provide protection in the event of above 6 mA of DC fault current, including leakage current, or
b. isolate the supply in the event or more than 6 mA of DC fault, including leakage, current.

My interpretation:
If we take the Type B approach, of Clause 7.9.3; then we're following option a) of the guidelines.
If we use an RDCDD + Type A RCD; we're following option b) - and the EDCDD should operate as soon as the d.c component rises above 6 mA (the point at which a Type A RCS becomes unreliable).
I'd need to read to the RDCDD Standard to be absolutely sure; but the wording of the Guidelines is pretty clear.

Bottom line is you need to install compliant RCD protection at the switchboard; and can't rely in whatever protection may be built into the charger itself.

[PeteRig]

Hi Alec, are you saying if the type A RCD is built into the charger with RDC-DD an A type RCD is still require at the origin end?
So testing would be for a type A RCD, no testing of the RDC-DD?
Thanks
Peter

[AlecK]

Correct on both points.

Both 7.9.3 (domestic) & 7.9.4 (other) specify that the "subcircuit" shall be protected;
and you can't provide this protection at the downstream end.


And the verification specified for newly-installed RCDs [8.3.10] is exactly the same for all types.
So yes the test for a car-charging circuit is the same as we would aply for a Type A on a normal socket circuit, or for a Type AC we might use as an "optional extra".

We are not required to test the DC-sensing function of either a Type B or an add-on RDCDD.
Just as we don't have to test the pulsing DC operation of a Type A, nor how fast it operates.
Nor do we have to test what current an mcb trips at, nor test its breaking capacity.
Nor do we test the IP ratings of enclosures.
All these things are part of product Standards.

The testing we have to do, as specified in Section 8, is NOT testing of the products used; it's testing of our work.
The point is to prove that we have installed these products correctly.

With reference to RCDs, that's that they have been installed in a manner that allows them to do what they are intended to do.
Note that the required result is not that it operates within a specified time, or at a particular value of leakage current, or on a particular waveform;
but simply that it interrupts all required poles of the circuit it's supposed to protect.


Other types of testing may be useful for fault-finding.
And some special applications , eg medical, require additional testing (not just of RCDs); both when commissioning and when carrying out periodic assessments.

[PeteRig]

Thanks Alec, why do the guidelines in 2.6 (f) say "use a purpose built RCD Tester...."?

[AlecK]

Since I had no involvement in writing Worksafe's Guidelines; I can only speculate as to why those words are there.
The wording "purpose build RCD tester" doesn't specify any particular functionality.
But it suggests not the RCD's integral push-button. Only suggests; because it can be argued that the integral test is a 'purpose built RCD tester".

The key difference between using the integral test circuit and an external tester is that the integral circuit doesn't use the PEC when creating an imbalance in the current, while most external testers do.
Both types of test apply a suitable resistor to approximate rated residual current (RRC) when connected at specified voltage.

The integral circuit places it between the load side Active and the line side Neutral, so that a small current bypasses the sensing part of the RCD ( for RCDs with NZ Approval, generally this is a toroid coil).
Thus the "supply" & "return" currents will be out of balance by approx RRC, causing the RCD to operate.
The exact test circuit, which is often printed on the device, nay include aux contacts to ensure that when the RCD is tripped, supply is removed from the test circuit - thus avoiding burn-out of the test circuit.

An external tester applies the resistance between active and PEC, which again creates an imbalance between "supply" & "return" current through the sensing part of the RCD.

It may be that whoever write that part of the Guidelines doesn't trust integral test circuits, or just thinks it's better in some way to use an external 'purpose built RCD tester".
Which could be as simple as a resistor and a pair of leads.

I am not aware that Worksafe has suggested - let alone actively sought - any amendment of clause 7.9.

[Tiger]

Thank you, Alec, for all the information.

my understanding to this,

"RCD protection at the load end of the circuit - it has to be at origin"

is trying to protect people when getting electric shock from the non-Rcd protected circuit from switchboard to the load. But in many cases, the Tesla Gen3 chargers are installed right under the swithboard and cable are cancelled behind the wall. The chances that having access to these cables are extremly low and in this case, it is really a waste of money to have RCD type B or RCD type A + RDC-dd at origin giving Gen3 has already built-in RCD Type A + DC 6mA.

All Notes to clauses are intended to provide additional information (as explained in the Foreword).
They do not set requirements.
This one relates to the reason that we don't use 'normal' Type A RCDs for EV supplies;
which is that with a rectifying load such as an EV charger there are very likely to be d.c. currents above 6 mA in the waveform of the residual (fault) current.
It was added to explain why adding an RDCDD to a Type A is deemed acceptable, while a Type A on its own isn't.

Notes to Clause 2.6.2.2.1 provide a (very) basic overview of the capabilities of various Types of RCD.
A Type A is designed to operate on sinusoidal waveform, and on pulsing d.c. waveform (as results from half-wave rectification).
It's known that Type A cannot be relied on to operate cirrectly if there's more than 6 mA of constant d.c in the residual current waveform.
The operating mechanism includes a magnetic coil. Basically the a.c. residual current results in a magnetic force to operate the trip mechanism. But if there's significant d.c this can create an opposite magnetic force.
My understanding is that a Type A will generally be reliable up to 6 mA of (non-pulsing) d.c component;
but I can't be 100% on that because operating in presence of straight d.c components is not a requirement for Type A.

Hence the requirement, for EV circuits, to either
a) use a Type B RCD (which is designed to be reliable in presence of d.c. components in the waveform); or
b) add an RDCDD in conjunction with a Type A. The operating parameters for the RDCDD will be set by the product Standard.
Note that using an RDCDD is only valid if all conditions for the Exception are met. Including that the RDCDD complies to the IEC Standard cited.

Even if the SDoC for this charger did specify compliance if the RDCDD with that IEC 62955, it still wouldn't meet the requirement of the clause.
The clause requires the subcircuit to be protected ; and you can't provide RCD protection at the load end of the circuit - it has to be at origin.

If you look at the EV charging guidelines published by Worksafe (2019); clause 1.4 similarly requires TCD protection of the subcircuit;
and advises that the RCD should either:
a. incorporate the ability to continue to provide protection in the event of above 6 mA of DC fault current, including leakage current, or
b. isolate the supply in the event or more than 6 mA of DC fault, including leakage, current.

My interpretation:
If we take the Type B approach, of Clause 7.9.3; then we're following option a) of the guidelines.
If we use an RDCDD + Type A RCD; we're following option b) - and the EDCDD should operate as soon as the d.c component rises above 6 mA (the point at which a Type A RCS becomes unreliable).
I'd need to read to the RDCDD Standard to be absolutely sure; but the wording of the Guidelines is pretty clear.

Bottom line is you need to install compliant RCD protection at the switchboard; and can't rely in whatever protection may be built into the charger itself.

[AlecK]

It may be that the requirement for the RCD protection to be at origin of circuit is excessive.
But - unless & until it is changed - it's still the requirement.
And complying with a requirement is never a "waste of money"

[JamesM]

Both 7.9.3 (domestic) & 7.9.4 (other) specify that the "subcircuit" shall be protected;
and you can't provide this protection at the downstream end.

Hi AlecK – I’m really curious as to why protection can’t be provided at the downstream end as long as it’s a single circuit without branches – along the same logic that 2.5.3.3 provides for with reference to overload protective devices. After all, the same current would flow at every point of the conductor, and that same flow would be interrupted regardless of where the device was located.

[AlecK]

Overload protection can be provided at any point on a circuit.
However short circuit protection can only be provided for points downstream of the protective device.
Similarly for RCD protection, which is about a curent imbalance in the part of the circuit downstream of the protective device.
Imbalances upstream of the RCD cannot be sensed, so protection can't be provided upstream of the RCD.

In this case, the requirement is not just RCD protection of the EVSE, but RCD protection of the (entire) subcircuit that supplies the EVSE.
If the RCD is within the EVSE, then it can't protect the subcircuit.
If it's part way along, it can protect part of the subcircuit, but not all of it.
The only way to provide RCD protection for the entire subcircuit is to have the RCD at the origin of the circuit.