We installed this very charger and had many discussions with ICS and ECA technical as to the requirement for an upfront RCD as we had wired the charger in SWA, in the end we got sick of the doubt surrounding it and fitted a Proteus A type 4 pole RCBO in the 3 phase D.B. As you know these devices fit on the 3 phase bus bar like a 3 pole MCB with the N connection having a flying lead like single phase, the lack of a physical test button on the charger was our deciding factor.
That's interesting how different these systems work. In 2017 I installed a wall charger for my mum. The manufacturer also for a demanded type A RCD in the supply dedicated for that circuit and the charger offers the DC protection internally. It indicates an error (three red LEDs, it has no fancy display) and the type A RCD trips. Even with a DC fault. It probably creates a leakage current to trip the local RCD. And also uses this system to isolate in case of other errors like one of the three phases is missing, too high or too low voltages between the phases or between the phases and the neutral. I actually had the missing line error once. The heat pump of the car was running, and the door was opened, so the heat pump was increasing and tripping the selective main circuit breaker of one line in the house (50A). The 32A C32A didn't trip, the 30mA type A RCD did trip then. The 50A Neozed was not blown, but the selective 50A main circuit breaker before the meter tripped. These things are quite sensitive sometimes on some occasions. They are rated to be selective to gG/gL curve fuses of the same current rating, and these should actually blow first before that MCB trips. These breakers became popular since the regulations now demand that there must be a measure before the meter that an electric layperson can isolate the whole installation before (!!) the meter. So we now got these before the meter, and the regular isolator switch after the meter. And more upstream are still 63A NH fuses in the service head.
I've fit ICS's gear and have split the tails and fit a surge protected board with a type a 32 amp rcbo. The ICS has an electronic RCD that acts like a time delayed RCD, add this to the fact that the RDCDD is in circuit I really do rate this device.
Residual Current tested a Rolec EV Charger a while back and it looked like that had a separate built in RCD that then fed into their bespoke controller that also covered AC and DC RCD protection, but he could only get the controller to trip on the RCD testing from the EV Charger outlet, he didn't do as much testing as you though.
This is a very interesting subject. We've currently got a site where there's a dedicated 200Amp 6 way 3 phase board going in, with 5 MCB protected 3 phase Zappi chargers, all clipped direct. However the Zappi has an integral 30mA Type-A RCD with 6mA DC leakage detection in accordance with EN 62955. Now technically this complies and I haven't seen anything against it in either the manufacturers instructions or BS7671: Amd 2. I will call the NICEIC and MyEnergi and double check with someone more familiar with them than me.
Check the instructions mate. It is in there you need up front type A rcds or it's a deviation on the certificate. Can't send screenshots on here or I would share it here lol
@@electrician247 That's the trouble I've been going through the newest instructions I've found on their website and it doesn't say it is, just says "check if local regulations say an upstream is required". Which then reads back around into "if the equivalent is already present no need". Hence I will make a phone call because this will require a 28 way DB just for 4 chargers. (Schneider). Wish they just made a 3 way 4 pole RCBO by now, would of been a simple choice then.
@@electrician247 I think I've found the solution - can I drop you an email if I can find it and get your opinion. As honestly I don't need to call them, I'll take your word as gospel any day when it comes to stuff like this.
@@electrician247 I'd love to but their site is Schneider now with power tag monitoring for all the circuits and metering. With a view to expand it across their other sites.
You always explain these things well Mark I get what you say now about the RCD not being in the unit and at the supply. Like the way, you reach out to others (manufacturers) to give you feedback. I put your name forward last year for an award for the work with 121 something Paul Meadham E5 was involved in never heard anything after that did you get it?
Be interesting to see if you had just a MCB upstream and then did the 5x test if the charge point protection would operate. Or would it be saturated and not function. 🤔 Also you can’t install a EV charge point on a split load board, or so Paul Chaffers from NAPIT told me. I asked him personally at ELEX regarding 722.531.3.101 which says they have to be protected individually by an RCD. Which seems ludicrous if I rock up to a property with a brand new CU, type A RCD and only 1 spare way. I can’t use a MCB I have to put in a new CU, be it an additional or replace the existing unless I can create a non-RCD way and use a RCBO. 🙄
I expect it would trip quickly enough but this sounds like more excuses for some fun experiments! Will do that lol thanks! In terms of the split load rcd use. It really does come down to risk assessment based on what's sat on the adjacent circuits. Even more so if the charge point has type A rcd and DC leakage protection built in. Even if its not to our spec... I would always try to use rcbos but as you describe sometimes it's not necessarily the best value. As with all of these things guidance documents it's open to interpretation.
Hi mark, to eliminate the issue of the x5 test and achieve selectivity, could you install a time delayed rcd/rcbo at source (providing it is not being used for additional protection) ?
The RCD protection in the charger does NOT protect the cable from the distribution panel (CU) to the charger. Accordingly, there are three reasons why you might need a second RCD before the charger: 1. To make up for any lack of protection in the charger 2. To protect everything in the installation in the circuit from the distribution panel to the charger 3. As a backup in case the RCD in the charger fails.
Commercial installs need to be using time-delay upstream so a dodgy EV doesn't trip out every single charger (which then get left offline for days causing lots of problems)
My understanding, is a DP 30mA RCD has to be installed at source point (Id call that upstream?) But as mentioned when testing my Megger trips the upstream RCBO (DP 30mA) before completing the tests at the EV. Can we use a time delayed 61008 instead? Thanks
I am surprised they are mandating the use of a type A RCD before the charger. As the tests demonstrate, with a type A RCD before before the charger, there is a problem with selectivity. The way to force selectivity and ensure the RCD in the charger trips before the upstream RCD is to use a delayed type RCD. Selectivity is not achieved by using RCDs of different values of trip current. That's a myth that higher trip current devices are slower. Some in reality might be, but you shouldn't be relying on it. So I think they are addressing the selectivity issue by ensuring that there are no other accessories or loads on the EV charging circuit, and hence mandating that the EV charger be on its own dedicated final circuit *and* that has its own dedicated RCD which doesn't feed anything else. In this way, it doesn't matter which RCD trips first: there is no adverse impact on any other circuit in the property if the upstream RCD trips first.
For additional protection time delayed rcds are not suitable. That would be very poor design. Unless a TT system with more generous disconnection times for fault protection. These have to be 30mA as per BS7671. I would have perhaps thought a B type rather than A. However as a single circuit for only the EV I can see the logic.
We installed this very charger and had many discussions with ICS and ECA technical as to the requirement for an upfront RCD as we had wired the charger in SWA, in the end we got sick of the doubt surrounding it and fitted a Proteus A type 4 pole RCBO in the 3 phase D.B. As you know these devices fit on the 3 phase bus bar like a 3 pole MCB with the N connection having a flying lead like single phase, the lack of a physical test button on the charger was our deciding factor.
Really starting to appreciate your content lately mate, keep up the good work👍
Thanks, very kind.
That's interesting how different these systems work. In 2017 I installed a wall charger for my mum. The manufacturer also for a demanded type A RCD in the supply dedicated for that circuit and the charger offers the DC protection internally. It indicates an error (three red LEDs, it has no fancy display) and the type A RCD trips. Even with a DC fault. It probably creates a leakage current to trip the local RCD. And also uses this system to isolate in case of other errors like one of the three phases is missing, too high or too low voltages between the phases or between the phases and the neutral.
I actually had the missing line error once. The heat pump of the car was running, and the door was opened, so the heat pump was increasing and tripping the selective main circuit breaker of one line in the house (50A). The 32A C32A didn't trip, the 30mA type A RCD did trip then. The 50A Neozed was not blown, but the selective 50A main circuit breaker before the meter tripped. These things are quite sensitive sometimes on some occasions. They are rated to be selective to gG/gL curve fuses of the same current rating, and these should actually blow first before that MCB trips. These breakers became popular since the regulations now demand that there must be a measure before the meter that an electric layperson can isolate the whole installation before (!!) the meter. So we now got these before the meter, and the regular isolator switch after the meter. And more upstream are still 63A NH fuses in the service head.
Great video thanks Mark.
I've fit ICS's gear and have split the tails and fit a surge protected board with a type a 32 amp rcbo. The ICS has an electronic RCD that acts like a time delayed RCD, add this to the fact that the RDCDD is in circuit I really do rate this device.
Yeah I think they tick a lot of boxes.
Residual Current tested a Rolec EV Charger a while back and it looked like that had a separate built in RCD that then fed into their bespoke controller that also covered AC and DC RCD protection, but he could only get the controller to trip on the RCD testing from the EV Charger outlet, he didn't do as much testing as you though.
Yes rolec and podpoint are among the few brands that have a physical rcd built in.
Interesting 👍
This is a very interesting subject. We've currently got a site where there's a dedicated 200Amp 6 way 3 phase board going in, with 5 MCB protected 3 phase Zappi chargers, all clipped direct. However the Zappi has an integral 30mA Type-A RCD with 6mA DC leakage detection in accordance with EN 62955. Now technically this complies and I haven't seen anything against it in either the manufacturers instructions or BS7671: Amd 2. I will call the NICEIC and MyEnergi and double check with someone more familiar with them than me.
Check the instructions mate. It is in there you need up front type A rcds or it's a deviation on the certificate. Can't send screenshots on here or I would share it here lol
@@electrician247 That's the trouble I've been going through the newest instructions I've found on their website and it doesn't say it is, just says "check if local regulations say an upstream is required".
Which then reads back around into "if the equivalent is already present no need".
Hence I will make a phone call because this will require a 28 way DB just for 4 chargers. (Schneider). Wish they just made a 3 way 4 pole RCBO by now, would of been a simple choice then.
@@electrician247 I think I've found the solution - can I drop you an email if I can find it and get your opinion. As honestly I don't need to call them, I'll take your word as gospel any day when it comes to stuff like this.
Proteus do a 4 pole rcbo that fits in a tpn spine. Ditch the schneider
@@electrician247 I'd love to but their site is Schneider now with power tag monitoring for all the circuits and metering. With a view to expand it across their other sites.
You always explain these things well Mark I get what you say now about the RCD not being in the unit and at the supply. Like the way, you reach out to others (manufacturers) to give you feedback. I put your name forward last year for an award for the work with 121 something Paul Meadham E5 was involved in never heard anything after that did you get it?
The esr awards? Yes I recently got a runners up award of highly commended in the change makers category. Thanks 🙌
Good 👍
Be interesting to see if you had just a MCB upstream and then did the 5x test if the charge point protection would operate. Or would it be saturated and not function. 🤔
Also you can’t install a EV charge point on a split load board, or so Paul Chaffers from NAPIT told me. I asked him personally at ELEX regarding 722.531.3.101 which says they have to be protected individually by an RCD.
Which seems ludicrous if I rock up to a property with a brand new CU, type A RCD and only 1 spare way. I can’t use a MCB I have to put in a new CU, be it an additional or replace the existing unless I can create a non-RCD way and use a RCBO. 🙄
I expect it would trip quickly enough but this sounds like more excuses for some fun experiments! Will do that lol thanks!
In terms of the split load rcd use. It really does come down to risk assessment based on what's sat on the adjacent circuits. Even more so if the charge point has type A rcd and DC leakage protection built in. Even if its not to our spec...
I would always try to use rcbos but as you describe sometimes it's not necessarily the best value.
As with all of these things guidance documents it's open to interpretation.
Hi mark, to eliminate the issue of the x5 test and achieve selectivity, could you install a time delayed rcd/rcbo at source (providing it is not being used for additional protection) ?
It's a cracking idea but in this application we can't. The type A 30mA rcd is for additional protection.
The RCD protection in the charger does NOT protect the cable from the distribution panel (CU) to the charger.
Accordingly, there are three reasons why you might need a second RCD before the charger:
1. To make up for any lack of protection in the charger
2. To protect everything in the installation in the circuit from the distribution panel to the charger
3. As a backup in case the RCD in the charger fails.
It's swa, no rcd protection is needed. Unless TT of course.
1. Is exactly why it's needed in most cases.
Commercial installs need to be using time-delay upstream so a dodgy EV doesn't trip out every single charger (which then get left offline for days causing lots of problems)
time delay is not an option. Either needs rcd or doesnt.
My understanding, is a DP 30mA RCD has to be installed at source point (Id call that upstream?) But as mentioned when testing my Megger trips the upstream RCBO (DP 30mA) before completing the tests at the EV. Can we use a time delayed 61008 instead? Thanks
It is frustrating isn't it. But sadly not we can't use time delay in this application.
Gooder
I am surprised they are mandating the use of a type A RCD before the charger.
As the tests demonstrate, with a type A RCD before before the charger, there is a problem with selectivity.
The way to force selectivity and ensure the RCD in the charger trips before the upstream RCD is to use a delayed type RCD. Selectivity is not achieved by using RCDs of different values of trip current. That's a myth that higher trip current devices are slower. Some in reality might be, but you shouldn't be relying on it.
So I think they are addressing the selectivity issue by ensuring that there are no other accessories or loads on the EV charging circuit, and hence mandating that the EV charger be on its own dedicated final circuit *and* that has its own dedicated RCD which doesn't feed anything else.
In this way, it doesn't matter which RCD trips first: there is no adverse impact on any other circuit in the property if the upstream RCD trips first.
For additional protection time delayed rcds are not suitable. That would be very poor design. Unless a TT system with more generous disconnection times for fault protection.
These have to be 30mA as per BS7671.
I would have perhaps thought a B type rather than A. However as a single circuit for only the EV I can see the logic.
Those fooking EV charger manufacturer just don't equip their stupid charger with a 6mA DC leakage device.