EV charger issue - Garo PME trips

[ElectricallyRandom]

Hi. Not sure if it's the right place to ask, so won't be upset if it gets removed. Apologies if the terminology (PEN/PME/etc) is used inaccurately.


I had an EV charger installed in December on my detached garage by a local electrician. The charger is one that doesn't have a PEN fault detection device (Tesla), hence we agreed I'd get it separately (Search Results - Detail - https://www.garo.co.uk/product-details/All/G6EV40PME).


Armoured 10mm^2 cable (from main circuit board to garage - approximately 32 meters) has been used to the garage where a smaller consumer board and the PME fault device was fitted. From there there's around 4-5m of 4mm cable to the actual charger.


After awhile I've noticed that the PME fault device trips every now and again - always late in the evening or at night, although that's mostly when I charge. It is quite sporadic, although in reality probably happens every 5-6 charges, which can be annoying in the morning if I miss the message on the car's phone app.


Interestingly the Tesla charger integrates with my Home assistant, so I can see the voltage at the charger, which seems to significantly drop on some evenings/nights (expected under load?), although never below the PME device's 207-253V range.


I have raised this with the electrician several times and he's not sure why this is happening. Also everything seemed fine on the 2 occasions he's tested it - once after installing and later when I raised the problem with him. However the test was never done by him at the time when voltage significantly dropped.

I did measure the voltage myself once when the voltage seemed low, soon after the PME device tripped and I reset it. In the house - ~232V, in the garage's consumer unit - ~230V, meanwhile Tesla charger in Home Assistant reported approx 222V while charging.



Question - where would I go from here ? Is there anything obvious another electrician would perhaps notice.. ?

Some photos attached.

Thank you

 

[Robson689]

Has the electrician tried replacing the GARO unit?

How come he decided to use this device vs an earth rod?

 

[ElectricallyRandom]

Has the electrician tried replacing the GARO unit?

How come he decided to use this device vs an earth rod?

No, not replaced (yet). Garo vs Earth rod - price and simplicity I suppose.

 

[timhoward]

My guess is either a faulty Garo unit or a regular supply voltage fluctuation that is too short to be picked up by the sampling the car charger is doing hence you can't see it on the app.
Leaving a decent multi-meter connected on min and then max might prove the point, failing that a voltage data logger could be hired and connected for a few days.
Even if you prove that, getting the DNO to do something about it can take effort and endurance.

To be honest I don't highly rate the methodology used by the Garo Pen protection devices as they aren't guaranteed to detect an open PEN conductor in all circumstances (subject to what loads your neighbours are pulling at the time) and they can be triggered by supply issues (as you may be seeing).

If it were me I'd be seriously considering banging an earth rod in and changing the earthing arrangements in the garage, removing the need for the PEN protection device completely.

I don't want to cause additional concern but a couple of unrelated things jump out at me too:
-you may have a PV system on a shared RCD in the house CU. This is a no-no as there's no point the RCD tripping to save your life if the PV system continues to deliver power until it notices it should shut down.
-the RCD topology isn't ideal - I think there's three 30ma type A RCD's in series. A Type B would normally be fitted for an EV charger.

 

[ElectricallyRandom]

My guess is either a faulty Garo unit or a regular supply voltage fluctuation that is too short to be picked up by the sampling the car charger is doing hence you can't see it on the app.
Leaving a decent multi-meter connected on min and then max might prove the point, failing that a voltage data logger could be hired and connected for a few days.
Even if you prove that, getting the DNO to do something about it can take effort and endurance.

To be honest I don't highly rate the methodology used by the Garo Pen protection devices as they aren't guaranteed to detect an open PEN conductor in all circumstances (subject to what loads your neighbours are pulling at the time) and they can be triggered by supply issues (as you may be seeing).

If it were me I'd be seriously considering banging an earth rod in and changing the earthing arrangements in the garage, removing the need for the PEN protection device completely.

I don't want to cause additional concern but a couple of unrelated things jump out at me too:
-you may have a PV system on a shared RCD in the house CU. This is a no-no as there's no point the RCD tripping to save your life if the PV system continues to deliver power until it notices it should shut down.
-the RCD topology isn't ideal - I think there's three 30ma type A RCD's in series. A Type B would normally be fitted for an EV charger.

Thanks for this. Before we commit to an earth rod, are there any PEN devices you'd rate better than the Garo? Perhaps a simple swap would suffice?

As far as solar - that's the wiring the house came with (new build), I suspect the same as the rest of the large new development. But I take your point and may raise it with the builders.

Thanks

 

[marconi]

After awhile I've noticed that the PME fault device trips every now and again

What actually trips in the GARO? The contactor, the RCD or both?

 

[timhoward]

After awhile I've noticed that the PME fault device trips every now and again

What actually trips in the GARO? The contactor, the RCD or both?

That is a great question and I did make an assumption there.
I assumed the contactor tripped as there are 2 other 30ma Type A RCDs upstream and I suspected one or both of them would also trip if it was the RCD tripping.

Before we commit to an earth rod, are there any PEN devices you'd rate better than the Garo?

Unfortunately I'm not aware of any better standalone ones that aren't integrated into charge points.
I believe the Matt:e one works the same as the one you have. Some charge points also monitor current on the earth conductor as an additional factor in their decision making. @Gavin John Hyde is the resident expert.

( See

)

Ultimately the point of this thing is to make sure that if there is a supply fault your car body cannot become live. This method of protection was added to the regs based on the laws of averages to make it easier to fit charge points. My personal opinion is that it's better to make sure the car body isn't connected to supply earth in the first place, and bang a couple of £10 earth rods in.

 

[ElectricallyRandom]

That is a great question and I did make an assumption there.
I assumed the contactor tripped as there are 2 other 30ma Type A RCDs upstream and I suspected one or both of them would also trip if it was the RCD tripping.

Unfortunately I'm not aware of any better standalone ones that aren't integrated into charge points.
I believe the Matt:e one works the same as the one you have. Some charge points also monitor current on the earth conductor as an additional factor in their decision making. @Gavin John Hyde is the resident expert.

( See

)

Ultimately the point of this thing is to make sure that if there is a supply fault your car body cannot become live. This method of protection was added to the regs based on the laws of averages to make it easier to fit charge points. My personal opinion is that it's better to make sure the car body isn't connected to supply earth in the first place, and bang a couple of £10 earth rods in.

Might not be the answer to your question, but the only thing that trips is the switch on the Garo itself.

 

[EricMark]

I can understand not wanting to knock in a meter of earth rod, but to detect loss of PEN it really only needs to be 6 inches. However

Armoured 10mm^2 cable (from main circuit board to garage - approximately 32 meters) has been used to the garage where a smaller consumer board and the PME fault device was fitted. From there there's around 4-5m of 4mm cable to the actual charger.

If in a garage why does it need any loss of PEN detection?

 

[timhoward]

I can understand not wanting to knock in a meter of earth rod, but to detect loss of PEN it really only needs to be 6 inches. However

If in a garage why does it need any loss of PEN detection?

Are you suggesting an EVCP installed on a non-dwelling doesn't need to comply with BS7671?

 

[marconi]

ElectricallyRandom: please tell us the exact make and model of your ev charger. I want to check its maximum power and power factor and what type and amperage of rcbo is recommended.

 

[marconi]

http://www.electricneutron.com/wp-content/uploads/2012/07/Wire-sizing-_current-rating-table.pdf. Page 4 of 11 CCC of 4, 6 and 10mm2 cabling.

https://www.tesla.com/sites/default...ll-connector-installation-manual-en-EU-v2.pdf Page 9 minimum cable size

Your voltage readings and graph suggest the garo is working correctly. I think the voltage readings provide a clue to the problem as over current.

A look at the second link on page 9 at bullet 5 of Minimum Requirements says the minimum supply cable csa must allow for periods of sustained current of 40A. The same reference states the maximum output current of the charger is 32A. The difference between 40A on the input to the charger and 32A on the output is indicative of a power factor less than 1.

We have already noted significant voltage drops by measurement and in the App graph. Not surprising when circa 40A flows along a final circuit some 40m long and with a cascade of connections and wiring devices added in too contributing resistance.

The 10mm2 between CU and garage DB meets current and voltage drop requirements. The 5m of 4mm2 is inadequate - see first link page 4. A 4mm2 clipped direct just about meets minimum of 40A. For a margin above 40A when cables run in conduit as is the case for the OP and to minimise voltage drop the required cabling is 10mm2.

I suspect, I cannot be certain right now, that the consequence of excessive voltage drop is the EV charger draws a higher input current - more than 40A- to produce a 32A output current. What I think may be happening is the GARO 40A RCBO is tripping on sustained overcurrent in the first few hours of charging. This is worsened perhaps if the battery charging mode changes after some elapsed time - we can see a second significant voltage dip at 10pm on the graph when current increases. It looks like the charger began output at about 8pm where the first voltage dip is seen.

Quite why the 32A type B mcb has not tripped in the garage DB is unexplained. But it could be that there is more rapid temperature rise in the GARO since it contains a warm contactor coil.

So what would I do? I would swap the 4mm2 cabling for 10mm2 and see what happens. Actually I would measure the current first. Do you have an ampclamp with which to measure current to charger?

I could be wrong of course but this is a cheap and easy thing to do first in the absence of an ampclamp and should be done anyway to meet installation requirement. Your electrician could measure the current though.

 

[ElectricallyRandom]

ElectricallyRandom: please tell us the exact make and model of your ev charger. I want to check its maximum power and power factor and what type and amperage of rcbo is recommended.

Tesla Gen 3 charger

 

[marconi]

https://www.tesla.com/sites/default...ng/Gen3_WallConnector_Installation_Manual.pdf

Would you look at the link and page 36 - 'Wall Connector LEDs'.

When you first start up the charger from powered off what is the pattern of vertical green LEDs you see which records the mcb size the charger believes it is connected to? By selecting the mcb size the charger then knows at what level to limit the maximum output current and thus input current. This is done during first commissioning of the charger. Do you know if it was done?

When you next charge car would you feel the 4 mm2 cabling or the conduit in which it is contained. Do this after an hour and then two hours. Is the cable/conduit noticeably warm? Stick a thermometer on if you have one and take a reading. Also feel the surface temperature of the GARO RCBO and the 32A circuit breaker in the garage board.. You do not need to open up anything to do this and the previous test.

I am wondering if your Gen 3 charger has been incompletely/incorrectly commissioned such that it 'thinks' it has a power supply feed sufficient to deliver its maximum or near maximum current output of 48A. See page 6 of link above.

 

[timhoward]

I have to say I think I was barking up wrong tree - it certainly looks more like an earth leakage or over current issue than a PEN fault detector issue. It makes sense to follow the advice above.

 

[loz2754]

Are you suggesting an EVCP installed on a non-dwelling doesn't need to comply with BS7671?

If the car is inside the garage there is no need for PEN fault protection. (From the Code of Practice)

 

[ElectricallyRandom]

If the car is inside the garage there is no need for PEN fault protection. (From the Code of Practice)

The car is kept outside.

https://www.tesla.com/sites/default...ng/Gen3_WallConnector_Installation_Manual.pdf

Would you look at the link and page 36 - 'Wall Connector LEDs'.

When you first start up the charger from powered off what is the pattern of vertical green LEDs you see which records the mcb size the charger believes it is connected to? By selecting the mcb size the charger then knows at what level to limit the maximum output current and thus input current. This is done during first commissioning of the charger. Do you know if it was done?

When you next charge car would you feel the 4 mm2 cabling or the conduit in which it is contained. Do this after an hour and then two hours. Is the cable/conduit noticeably warm? Stick a thermometer on if you have one and take a reading. Also feel the surface temperature of the GARO RCBO and the 32A circuit breaker in the garage board.. You do not need to open up anything to do this and the previous test.

I am wondering if your Gen 3 charger has been incompletely/incorrectly commissioned such that it 'thinks' it has a power supply feed sufficient to deliver its maximum or near maximum current output of 48A. See page 6 of link above.

After it was installed, I did the commissioning myself - all it is is setting the charging current and a couple more basic settings. It is set to 32A. When I charge, the Tesla app does show 32A and my smart meter reading goes up by a little over 7kW. I will check the temps when I charge next.

 

[marconi]

Thank you for the information. A quick analysis.

If charger input terminal voltage is 222 V as you measured and output is 32A delivering 7.6kW from the table of page 6 of manual and assumed power factor is 0.9 (the tesla Gen 3 includes some power factor correction technology) then apparent input current is:

I = 7600/(0.9 x 222) = 38A

The apparent input current waveform is not a pure sinewave and will be peaky ie have a high crest factor which the ratio of its peak value to its rms value. A higher crest factor is indicative of the presence of harmonics and distortion.

See: Crest Factor - https://www.sunpower-uk.com/glossary/crest-factor/#:~:text=Crest%20Factor%20(peak%2Dto%2D,has%20a%20CF%20of%201.414.

At face value 38A is higher than than the rating of the type B 32A mcb in the garage DB. For the GARO 40A RCBO, 38A is 95% of its rated amperage - good practice is to derate the amperage rating of an mcb when it passes a continuous current such as several hours of high current ECV charging. For one MCB in an enclosure with a continuous load a typical derating is 80%. Thus the GARA 40A has a derated rating of 40 x 0.8 = 32A which is below the apparent input current of 38A.

Further derating may be required - I cannot be certain - because the GARO 40A RCBO is next to a warm contactor which would derate the continuous current handling further to below 32A.

See the IET article you should find if you google 'Derating grouped circuit breakers within a consumer unit'. I cannot post a link because the EF forbids it.

All in all then it seems to me that a maximum EV charging current of 32A is too high. A figure of 24A for maximum charging current is within the continuous current carrying capacity of the 32A mcb and 40A rcbo when each is derated. (32A mcb derated by 80% is 26A).

You can see now why I would like you to feel and if possible measure the temperatures of the rcbo and wiring. So, once you have done this, could you change the maximum charging current to 24A by telling the charger the mcb size is 30A - check the LED lights show this change has been made. And then see if the GARO trips over say the next 10 charges. Hold fire for now on changing the 4mm2 cabling as I suggested earlier.

Where are you in the country - just for my interest please?

 

[ElectricallyRandom]

Thank you for the information. A quick analysis.

If charger input terminal voltage is 222 V as you measured and output is 32A delivering 7.6kW from the table of page 6 of manual and assumed power factor is 0.9 (the tesla Gen 3 includes some power factor correction technology) then apparent input current is:

I = 7600/(0.9 x 222) = 38A

The apparent input current waveform is not a pure sinewave and will be peaky ie have a high crest factor which the ratio of its peak value to its rms value. A higher crest factor is indicative of the presence of harmonics and distortion.

See: Crest Factor - https://www.sunpower-uk.com/glossary/crest-factor/#:~:text=Crest%20Factor%20(peak%2Dto%2D,has%20a%20CF%20of%201.414.

At face value 38A is higher than than the rating of the type B 32A mcb in the garage DB. For the GARO 40A RCBO, 38A is 95% of its rated amperage - good practice is to derate the amperage rating of an mcb when it passes a continuous current such as several hours of high current ECV charging. For one MCB in an enclosure with a continuous load a typical derating is 80%. Thus the GARA 40A has a derated rating of 40 x 0.8 = 32A which is below the apparent input current of 38A.

Further derating may be required - I cannot be certain - because the GARO 40A RCBO is next to a warm contactor which would derate the continuous current handling further to below 32A.

See the IET article you should find if you google 'Derating grouped circuit breakers within a consumer unit'. I cannot post a link because the EF forbids it.

All in all then it seems to me that a maximum EV charging current of 32A is too high. A figure of 24A for maximum charging current is within the continuous current carrying capacity of the 32A mcb and 40A rcbo when each is derated. (32A mcb derated by 80% is 26A).

You can see now why I would like you to feel and if possible measure the temperatures of the rcbo and wiring. So, once you have done this, could you change the maximum charging current to 24A by telling the charger the mcb size is 30A - check the LED lights show this change has been made. And then see if the GARO trips over say the next 10 charges. Hold fire for now on changing the 4mm2 cabling as I suggested earlier.

Where are you in the country - just for my interest please?

Hi. I really appreciate such a comprehensive reply - must say it was a very interesting read (my inner geek approves ). Preferably I'd really like to charge at a higher current than <32A in the future if possible, but I'll try it for testing purposes.
As I can change the charging current in the Tesla app, I may just do that - it normally slowly ramps up the current to whatever it is set to.

I'm in Hampshire (are you anywhere close?).

Thanks!

 

[marconi]

Not far from Dartford Crossing.