RCD Types for EV Chargers on TT systems

Assuming we are talking about Mode 3 chargers, then upstream RCDs don't need to be Type B as a result of the charger in all cases … provided of course the electronics in the EVSE itself doesn't generate DC residual currents.

The issue that requires DC residual fault current protection to be in place, is a fault between a live conductor and PE downstream of the charger (e.g. in the charging cable or car), because of the pilot currents returning via PE from the vehicle to the EVSE.

If the EV charger is protected by a Type B RCD, or a Type A or F with an RDC-DD, any DC residual fault currents downstream of that protection will be disconnected by those devices, and therefore upstream RCDs not affected.

One issue regarding RCD protection upstream of EV charging equipment, is the requirement for all live conductors to be disconnected, as there is often no guaranteed selectivity between cascaded RCDs, even of higher residual current rating - the only way to ensure selectivity is to use an S-Type time-delay upstream of a non-delay RCD.

Thank you for the reply gkenyon. 

I thought this was the case but I had been informed by Siemens and Deopke that both charger and upstream RCD had to be Type B.

Thanks again for the guidance.

CameronElmtronics: 
 

Thank you for the reply gkenyon. 

I thought this was the case but I had been informed by Siemens and Deopke that both charger and upstream RCD had to be Type B.

Thanks again for the guidance.

That interpretation can't be true, otherwise BS 7671 would not permit a Type A or Type F RCD in cases where the EVSE has built-in protection against DC residual fault currents (effectively equivalent to an RDC-DD or Type B RCD internally). See Regulation 722.531.3.101.

The only time a Type B would be required upstream of the EVSE, is where the electronics in the EVSE itself somehow generates conditions where DC residual fault currents are propagated upstream. In general, that should not happen.

I thought the problem was that 30mA B-type RCDs might not trip until the d.c. residual current reaches 60mA - whereas 6mA d.c. might be enough to interfere with the operation of A-type RCDs.

In some ways the RDC-DD solution is better as it limits the d.c. residual current (from that one charge point at least) to 6mA rather than 60mA - thus allowing A-type RCDs to be used upstream.

Hence the general rule that if you need a B-type at some particular point, then likely you'll need B-types upstream of that.

It gets even more complicated where you have multiple charge points fed from a common upstream RCD (individually each charge point might have a d.c. residual current <6mA but that might add up to something well above 6mA). Also the 6mA and 60mA figures might not hold true for higher rated RCDs … but what the equivalent numbers are seems to be hard to find.

   - Andy.

AJJewsbury: 
 

I thought the problem was that 30mA B-type RCDs might not trip until the d.c. residual current reaches 60mA - whereas 6mA d.c. might be enough to interfere with the operation of A-type RCDs.

Yes, this is the usual claim when equipment poses a risk of giving smooth dc residual (because of not either having a design to avoid it or having its own detection to limit it to a few mA).  The type-EV RCD has a tighter upper limit on what smooth dc residual it will permit, intended to make it ok for upstream type-A.

722.531.3.101 doesn't seem to care whether the EVSE has a Type B RCD or RDC-DD within it, except that you might be able to get away without the Type A RCD upstream if there's a Type B within it.

So, is the wording in 722.531.3.101 wrong?

gkenyon: 
 

722.531.3.101 doesn't seem to care whether the EVSE has a Type B RCD or RDC-DD within it, except that you might be able to get away without the Type A RCD upstream if there's a Type B within it.

So, is the wording in 722.531.3.101 wrong?

722.531.3.101 seems reasonable for its scope of protecting the EV side of the RCD.

It ignores possible effects on upstream RCDs that could be protecting other parallel circuits as well as the cables to the EV installation. If the regulations are to be taken as a whole, it's reasonable to leave other parts of the regulations to deal with ensuring that upstream RCDs are suitable types to handle whatever the EV installation (with its local RCDs or built-in detection) might throw at them. The general one seems to be 531.3.3, which wants the 'appropriate RCD' to be selected for each case. If a new load is connected downstream, that appropriate choice may have to be reassessed for upstream RCDs.

Realistically, it might be good to have a note near 722.531.3.101 to remind that the installation of a new piece of power electronics could require upstream RCDs to be checked for suitability. After all, the 531.3.3 is vague, and not very helpful by giving short descriptions and then referring to IEC/TR 62350 [which would cost about two copies of BS7671].

Practically, of course, we come down to the usual arguments about how worthwhile the effort and cost is to try to avoid reduced protection in a rather special combination of circumstances - e.g. significant dc from EV system, but not enough to trip type-B, and also a pulsating dc fault in the same direction elsewhere in the system. But it seems that the intention of the regulations is for RCD protection not to be lost for any within-spec behaviour of equipment in the system.  (Mild changes of threshold are a different matter: e.g. an RCD in a 3-phase system could theoretically have its threshold pushed up to nearly twice its rated operation level if there's near-threshold leakage on the two other phases. No one seems to worry about this mild change that's unlikely actually to happen to the full extent.)
 

gkenyon: 


One issue regarding RCD protection upstream of EV charging equipment, is the requirement for all live conductors to be disconnected, as there is often no guaranteed selectivity between cascaded RCDs, even of higher residual current rating - the only way to ensure selectivity is to use an S-Type time-delay upstream of a non-delay RCD.

Graham, can I ask where the requirement for all live conductors of the upstream RCD to be disconnected comes from if the RCD in the EVSE does that?

lyledunn: 
 

gkenyon: 


One issue regarding RCD protection upstream of EV charging equipment, is the requirement for all live conductors to be disconnected, as there is often no guaranteed selectivity between cascaded RCDs, even of higher residual current rating - the only way to ensure selectivity is to use an S-Type time-delay upstream of a non-delay RCD.

Graham, can I ask where the requirement for all live conductors of the upstream RCD to be disconnected comes from if the RCD in the EVSE does that?

Selectivity (or lack of it) may mean that if there's an issue downstream of the "EV RCD", the upstream RCD will trip first. If the RCD operation is required to disconnect all live conductors, where there is no selectivity, surely all upstream RCDs should meet the requirement?

Whilst I accept this point of view could be a link back to the “general rules” as Andy has pointed out, in the discussion in this thread, I believe there is a world of difference … purely because the words regarding “protection against DC residual fault currents” is wide open to broad interpretation. That is, no performance requirement as to the level or type of “DC residual fault currents” is stated regarding the criteria for permitting Type A or Type F upstream … whether that's the right approach or not, is yet another question.

So, is the wording in 722.531.3.101 wrong?

I wouldn't say it's wrong, but it's only concerning itself with the RCD protection required by the EVSE outlet (which would need to be part of the installation if not incorporated into the EVSE) - it simply doesn't comment on the requirements for any further upstream RCDs.

   - Andy.