Installing a 3 phase rapid EV charger?

[Julie.]

So it's looking like this charger just requires a 5 pin 16A commando socket and that this scenario isn't part of section 722 as its just ELV on the vehicle.

I was hoping to run past you (and others, just happened to quote you) my final (hopefully) thoughts as this is my first delve into 3 phase.

Change the 3 pole mcb in the 3 phase board for a C20. Take from this to a separate small CU housing a 4 pole Type AC RCD. From there I go to the commando socket (or possibly bypass the CU and just have the required RCD combined in the commando socket)? Probably all in 5 core 2.5mm H07RN?

I have chosen a Type AC RCD for poor reasons. There is an upstream 300ma RCD which I was trying to avoid changing to a Type A. Also, type AC RCD's are 'usually' acceptable (i.e meet the regulations) to use to protect sockets.

I am unsure how to ascertain if a type A is required as I cant get anybody at Avant who understands what DC leakage is and how likely it is with the charger/loader.

Main question is the highlighted bit though.

I would suggest you fit type A rcd not ac

It doesn't matter that the upstream rcd is ac.

What does matter is the dc converter could inject dc into the lines, this may block an ac rcd - if you have two in series, neither would operate. If the downstream is type A then this would likely still operate.

There is usually no difference in cost, and it could be argued that since you have knowledge of the type of device intended for that circuit you ought to design for that, not fall back to the minimum standard which is likely unsuitable for this application.

It may have been different, if they had merely asked for a 16A socket outlet for general use, then a type ac whilst undesirable these days, would comply, had they subsequently used something that blocks the rcd, since you had no knowledge of the end device - no fault of yours.

However, you do actually know the intended end device. (for which ac type rcd is not recommended )

 

[Rockingit]

Agree with choosing an A type RCD for all the reasons above.......however, if you're fitting a 16a ceeform socket then you have to use a 16A CPD, not a 20A otherwise you fall foul of 433.1.1

 

[HappyHippyDad]

I would suggest you fit type A rcd not ac

It doesn't matter that the upstream rcd is ac.

What does matter is the dc converter could inject dc into the lines, this may block an ac rcd - if you have two in series, neither would operate. If the downstream is type A then this would likely still operate.

There is usually no difference in cost, and it could be argued that since you have knowledge of the type of device intended for that circuit you ought to design for that, not fall back to the minimum standard which is likely unsuitable for this application.

It may have been different, if they had merely asked for a 16A socket outlet for general use, then a type ac whilst undesirable these days, would comply, had they subsequently used something that blocks the rcd, since you had no knowledge of the end device - no fault of yours.

However, you do actually know the intended end device. (for which ac type rcd is not recommended )

Thanks Julie.

Could I ask why you don't feel it's a problem having a Type AC RCD upstream from the new Type A RCD that will be fitted? Could the Type AC RCD get blinded? The whole point in fitting a Type A RCD is because we fear we may get <6mA DC leakage. If we fear this, then how can we have a Type AC upstream as this could start to get saturated and potentially cease to operate?

Also, could you point me in the direction of any literature that suggests a type AC is not suitable for this intended use as it would be interesting to read? Or are you just going by your own knowledge?

Actually, I think I can answer my second question. I don't think a Type AC is generally suitable for items with electronic components, i.e the charger and loader! First question still stands though.

 

[Rockingit]

Thanks Julie.

Could I ask why you don't feel it's a problem having a Type AC RCD upstream from the new Type A RCD that will be fitted? Could the Type AC RCD get blinded? The whole point in fitting a Type A RCD is because we fear we may get <6mA DC leakage. If we fear this, then how can we have a Type AC upstream?

Also, could you point me in the direction of any literature that suggests a type AC is not suitable for this intended use as it would be interesting to read? Or are you just going by your own knowledge?

I assume the logic here is that any downstream requirements will be dealt with by the usual discrimination of other devices.

 

[partplonker]

Could I ask why you don't feel it's a problem having a Type AC RCD upstream from the new Type A RCD that will be fitted? Could the Type AC RCD get blinded?

Yes, you would have 2 type AC RCDs in series, so any DC leakage would blind both.

The Type A would trip on <6mA DC leakage, disconnecting the battery charger and protecting the upstream type AC.

You're right though, there is nothing in the regs that says you cannot do this. The only mention of using a particular type of RCD is when you get to special locations.

I do know that some European countries banned the use of type AC RCDs altogether in their wiring regs years ago if that tells you anything!

 

[pc1966]

Could I ask why you don't feel it's a problem having a Type AC RCD upstream from the new Type A RCD that will be fitted? Could the Type AC RCD get blinded? The whole point in fitting a Type A RCD is because we fear we may get <6mA DC leakage. If we fear this, then how can we have a Type AC upstream as this could start to get saturated and potentially cease to operate?

You will have less chance of the upstream one being blinded by using a type A as it will trip on partial DC, the very thing that blinds a type AC in large amount!

Really, there is little or no reason to chose type AC now, type A ought to be the minimum!

 

[pc1966]

I do know that some European countries banned the use of type AC RCDs altogether in their wiring regs years ago if that tells you anything!

A lot of EU is on TT systems so the RCD is often the only means for earth fault disconnection so the risk from electronic faults blinding it are far more serious!

Here we are mostly TN earth so most faults would take out the OCPD aspect, so presumably our regs lagged as not such a high risk. However, today there is no excuse for type AC being made and as other have already said, cost difference to type A is negligible and they are just variations of the same differential transformer approach.

Type B are another kettle of fish, they have much more complicated internals to detect "pure" DC faults and are much more expensive as a result.

 

[brianmoooore]

Agree with choosing an A type RCD for all the reasons above.......however, if you're fitting a 16a ceeform socket then you have to use a 16A CPD, not a 20A otherwise you fall foul of 433.1.1

Something changed in the regs here? Maximum CPD for a 16A Ceeform/Commando type socket was 20A.

 

[Rockingit]

Something changed in the regs here? Maximum CPD for a 16A Ceeform/Commando type socket was 20A.

I got caught out on this on an inspection years ago.... 433.1.1 (ii) says "the rated current or current setting of the protective device (In) does not exceed the lowest of the current-carrying capacities (Iz) of any of the conductors of the circuit.." So as the conductors of a 16A socket are only nominally rated for 16A, the CPD has to be also 16A or (whatever the socket size is).

 

[DPG]

I got caught out on this on an inspection years ago.... 433.1.1 (ii) says "the rated current or current setting of the protective device (In) does not exceed the lowest of the current-carrying capacities (Iz) of any of the conductors of the circuit.." So as the conductors of a 16A socket are only nominally rated for 16A, the CPD has to be also 16A or (whatever the socket size is).

The conductors of the circuit surely though. You don't have 'conductors of the socket'.

 

[Julie.]

Yes, you would have 2 type AC RCDs in series, so any DC leakage would blind both.

The Type A would trip on <6mA DC leakage, disconnecting the battery charger and protecting the upstream type AC.
..

This.

It's similar to having a 10A mcb downstream of a 40A mcb, any fault causing sufficient current to operate the 10A would trip it, it doesn't matter that the 40A mcb may not provide sufficient protection downstream (of the 10A mcb).

 

[Rockingit]

The conductors of the circuit surelythough. You don't have 'conductors of the socket'.

You'd think so, but no. 433.1.204 is the follow-on exemption clause for RFC's that references compliance to 433.1.1

Also are the requirements of 512.1.2 - "Every item..."

I had an inspector make me swap a 10a switch for a 20a one as it was downstream of a 13a FSU!!

 

[Julie.]

...
You're right though, there is nothing in the regs that says you cannot do this. The only mention of using a particular type of RCD is when you get to special locations.
..

Yes there is, in 531.3.3 it states that the type of rcd appropiate to the application shall be selected.

It also notes guidance on the correct selection to an iec document.

 

[HappyHippyDad]

This.

It's similar to having a 6A mcb downstream of a 40A mcb, any fault causing sufficient current to operate the 6A would trip it, it doesn't matter that the 40A mcb may not provide sufficient protection downstream (of the 6A mcb).

I think I am getting confused about 'when' a type A RCD actually trips.

I thought that it simply had to capability to withstand <6mA DC pulsating current leakage without getting blinded. I now realise that it trips when it detects <6mA DC, but when does it choose to trip? At 1mA, 2mA etc etc?

What happens if the DC leakage is >6mA? Why would the Type A not trip? Is this because it would then be saturated? Still doesn't make sense though as it would have to have gone through a lower amount of DC current first (i.e<6mA) and therefore trip the RCD before it reaches the higher levels. Unless DC leakage occurs in larger 'bursts'!

 

[Julie.]

I think I am getting confused about 'when' a type A RCD actually trips.

I thought that it simply had to capability to withstand <6mA DC pulsating current leakage without getting blinded. I now realise that it trips when it detects <6mA DC, but when does it choose to trip? At 1mA, 2mA etc etc?

What happens if the DC leakage is >6mA? Why would the Type A not trip? Is this because it would then be saturated? Still doesn't make sense though as it would have to have gone through a lower amount of DC current first (i.e<6mA) and therefore trip the RCD before it reaches the higher levels. Unless DC leakage occurs in larger 'bursts'!

Your initial understanding is correct.

The rcd only operates on ac (type ac and A) , however if you haves some dc, then the ac type will fail to see any ac fault/leakage current ; the type A can however see the ac fault/leakage current providing the dc value is less than 6mA or pulsating.

Hence a type A can be used on devices such as rectifiers etc whilst a type ac cannot.

Type B and so on can operate on ac and dc

My analogy was just meant to show that providing you have a device which will operate downstream of another device which won't, the situation is acceptable.

I think the fact I picked a 6A mcb may have confused the issue. EDIT I have changed it to a 10A mcb so it should be less confusing.


A type ac rcd could be blocked by a dc current (of any value), or by pulsating dc

A type A will not be blocked by dc current less than 6mA or by pulsating dc - but could be blocked by dc greater than 6mA

A type B will not be blocked by dc

 

[pc1966]

What happens if the DC leakage is >6mA?

Here is one I prepared earlier, in true Blue Peter style:

RCD "blinding" by DC - in - Electrical Wiring, Theories and Regulations

, RCD "blinding" by DC, Electrical Wiring, Theories and Regulations, ElectriciansForums.net Est.2006 | Free Electrical Advice Forum and page_number.

www.electriciansforums.net

 

[partplonker]

I think I am getting confused about 'when' a type A RCD actually trips.

I thought that it simply had to capability to withstand <6mA DC pulsating current leakage without getting blinded. I now realise that it trips when it detects <6mA DC, but when does it choose to trip? At 1mA, 2mA etc etc?

What happens if the DC leakage is >6mA? Why would the Type A not trip? Is this because it would then be saturated? Still doesn't make sense though as it would have to have gone through a lower amount of DC current first (i.e<6mA) and therefore trip the RCD.

Edit: Removed all my nonsense because I missed that the AC RCD is 300mA for fire protection.

 

[Julie.]

Perhaps it is useful to see why a type ac gets blocked, whilst type A doesn't, so I have attached a scrawled diagram, and will try to explain.

Firstly the primary current (30mA?) Induces a magnetic flux in the core, and it is this flux which induces the current in the secondary winding, which is detected by the rcd electronics.

Normally you select a core material which has low losses - so a narrow BH curve as per the first (LH diagram). When there is primary current, if you trace the ac waveform up, since it hits the B-H curve in the 'working' bit - it produces a decent flux which produces a decent secondary current, so the rcd operates well on an ac fault/leakage current.

But if you have a dc offset (see RH diagram) then the ac current isn't in the good part of the B-H curve - instead it is near the flat portion, hence the primary ac current produces virtually no flux change, and hence no secondary current - this is due to the dc offset.

Now take type A (very bottom drawing).

In this case you use more material (gives more 'good' bit of the characteristic), and a different material with higher losses - this gives a wider B-H curve, and providing the dc offset is less than 6mA the primary current can still produce sufficient magnetic flux to give sufficient secondary current to operate the rcd.

Sorry for the c**p drawings!

 

[marconi]

Julie: I must write to say your drawings were very clear and illustrative of what you set out to show. I think such 'doodles' are both helpful in explaining things, enjoyable to digest and akin to one's handwriting say something good and informing about the drawer's character. More of the same please.

 

[HappyHippyDad]

If there are three spare ways in the Merlin board then yes take the circuit from there but you will have to fit a separate rcd as you won't get a four pole rcbo for that board.

How about a 3 pole RCBO Westward?
Haven't found one yet though!

 

[westward10]

How about a 3 pole RCBO Westward?
Haven't found one yet though!

You won't get three or four pole rcbos for that board.

 

[Risteard]

How about a 3 pole RCBO Westward?
Haven't found one yet though!

A triple pole RCBO would be no use for an EVSE installation.

 

[Aaron b]

If there are three spare ways in the Merlin board then yes take the circuit from there but you will have to fit a separate rcd as you won't get a four pole rcbo for that board.

I have a job using these boards currently. Do they take the Schneider multi 9s? I assume they do but would rather be sure before buying ten mcbs.

 

[HappyHippyDad]

A triple pole RCBO would be no use for an EVSE installation.

Completely agree Risteard, but in this case it is just feeding a commando socket so would be ok. However, I've bought a 4 POLE RCD combined with a commando socket and will use a 3 pole MCB in the CU.

 

[Risteard]

Completely agree Risteard, but in this case it is just feeding a commando socket so would be ok. However, I've bought a 4 POLE RCD combined with a commando socket and will use a 3 pole MCB in the CU.

Ideal.

Generally an MCB feeding an RCCB is going to be the most realistic option here.