EU spec Supercharger

jkirkebo · Feb 5, 2015

mod note: This conversation moved out of CHAdeMO adapter thread in Model S forum

mgboyes said:
Sorry but that bears no resemblance to what I have seen in the UK, and goes against all the arguments/logic about how SC cabinets are configured in the US and elsewhere that I have seen. Do you have a source/reference for this? Examples of such transformers in use? Which country have you seen this setup in?

US SC units are a stack of twelve x 10kW charge units (same as the ones in US cars) hence 120kW.
EU SC units are a stack of twelve x 11kW 3 phase charge units (same as the ones in EU cars) hence 132kW which is rounded up to 135kW for labeling.
There are no step up transformers at any UK supercharger site that I've visited or seen photos of.
In countries where 400V 3 phase power is common why on earth would you go to the huge expense and hassle of stepping up slightly to 480V?

The rating label on a Gen II 135kW supercharger says input of 380-480V 3phase at up to 192A. 192A is exactly what you need to drive 12 European charge modules at 16A per phase each. That's 133kW total at an EU standard 400V (230V L-N on each phase).

Nevertheless, all newer Norwegian SC sites have the step-up transformers. Here is a picture of the label:

480V 3-phase 192A is 160kW, same as the transformer rating, and the neccessary input power to achieve 150kW output (since chargers are not 100% efficient).

If you still don't believe me about the transformers I'll take an overview picture of Solli SC. I also have the building permits which goes into details about the transformers, but they are in Norwegian.

As for the internal SC configuration, my guess is 15x10kW US units. 480V is what they use in the US so it makes sense to do the same here.

The maximum current output of a Gen II 135kW supercharger is 330A continuous so it could not possibly charge two cars at a total of 150kW between them. When you and an adjacent car saw a total that exceeded 120kW are you sure that you were both connected to the same SC cabinet. At some large SC sites the paired bays aren't adjacent - they sometimes install them as e.g. 1A 2A 3A 4A 1B 2B 3B 4B.

The 330A limit is for one connector, not the total power shared. Yes, I am sure we shared. I checked the labeling, we were the only two cars there and the other car was nowhere near full battery. 120kW for me and 30kW for him. If he was on another SC he would get way more than 30kW. Also there was a tech on site working on the SCs, he said he was upgrading them. I do not know what he was upgrading, but he confirmed 150kW per cabinet rating.

My guess is the newer US Superchargers also have 15 10kW units, while the old ones had 12 units. Instead of continuing with two lines they are probably standarising on the US configuration.

arg · Feb 5, 2015

mgboyes said:
US SC units are a stack of twelve x 10kW charge units (same as the ones in US cars) hence 120kW.
EU SC units are a stack of twelve x 11kW 3 phase charge units (same as the ones in EU cars) hence 132kW which is rounded up to 135kW for labeling.

I don't think this is quite accurate. The original US units were 120kW, more recent ones "135kW". I strongly suspect that the more recent US installations are using the same charger modules as used in the European cars, though that is just speculation of course.

In countries where 400V 3 phase power is common why on earth would you go to the huge expense and hassle of stepping up slightly to 480V?

It does seem strange. Possibly since the transformers seem Norway-specific, there's some other reason why they have to have a transformer (earthing, star vs delta or somesuch) and if they are having a transformer anyhow, why not get the voltage right?

Or possibly the supply voltage there is lower (although we are all happily 'harmonised' on 400V, the tolerances are quite wide and supplies are typically still at their historic voltages - we normally get more than 400V in the UK)?

Here's the spec label on one of the SC cabinets at the Maidstone charger in the UK:

The labels on the back of the supercharger cabinets have long defied analysis (often being out of step with proven reality), though this one seems closer than the early ones.

However, there's some better data available - coincidentally also from Maidstone! Follow this link to the planning application for that supercharger, and download the 5th document listed - it's a datasheet for the superchargers.

This clearly shows that on the input side they are current-limited, so take higher input power at higher voltage (always 192A, but 126kVA at 380V rising to 160kVA at 480V). Continuous DC output power is specified as 115kW at 380V, 120kW at 400V and 145kW at 480V. All totally rational up to that point; the outputs correspond to inputs with a reasonable efficiency factor. Unfortunately, the continuous DC current is specified as 305A/310A/335A for the three voltages, which doesn't entirely make sense - particularly the 335A figure to go with 145kW, which would imply a battery voltage of 432V. The other two figures correspond to more reasonable battery voltages; possibly the 335A limit is the max per output connector rather than the max for the unit as a whole.

The maximum current output of a Gen II 135kW supercharger is 330A continuous so it could not possibly charge two cars at a total of 150kW between them. When you and an adjacent car saw a total that exceeded 120kW are you sure that you were both connected to the same SC cabinet. At some large SC sites the paired bays aren't adjacent - they sometimes install them as e.g. 1A 2A 3A 4A 1B 2B 3B 4B.

Well, at 480V it's specified to deliver 145kW continuous. Somehow that 335A max limit needs to be explained away - either its a per-output limit as I speculate above, or the 335A can be exceeded short-term even if it can't be sustained continuously.

mgboyes · Feb 5, 2015

jkirkebo said:
Nevertheless, all newer Norwegian SC sites have the step-up transformers. Here is a picture of the label:
View attachment 71375

Fascinating - you learn something every day. My apologies for doubting you!

arg said:
I don't think this is quite accurate. The original US units were 120kW, more recent ones "135kW". I strongly suspect that the more recent US installations are using the same charger modules as used in the European cars, though that is just speculation of course.

Yes I'd agree with your speculation. 135kW superchargers seemed to arrive around the same time as EU cars in general.

arg said:
It does seem strange. Possibly since the transformers seem Norway-specific, there's some other reason why they have to have a transformer (earthing, star vs delta or somesuch) and if they are having a transformer anyhow, why not get the voltage right?

Or possibly the supply voltage there is lower (although we are all happily 'harmonised' on 400V, the tolerances are quite wide and supplies are typically still at their historic voltages - we normally get more than 400V in the UK)?

The labels on the back of the supercharger cabinets have long defied analysis (often being out of step with proven reality), though this one seems closer than the early ones.

However, there's some better data available - coincidentally also from Maidstone! Follow this link to the planning application for that supercharger, and download the 5th document listed - it's a datasheet for the superchargers.

This clearly shows that on the input side they are current-limited, so take higher input power at higher voltage (always 192A, but 126kVA at 380V rising to 160kVA at 480V). Continuous DC output power is specified as 115kW at 380V, 120kW at 400V and 145kW at 480V. All totally rational up to that point; the outputs correspond to inputs with a reasonable efficiency factor. Unfortunately, the continuous DC current is specified as 305A/310A/335A for the three voltages, which doesn't entirely make sense - particularly the 335A figure to go with 145kW, which would imply a battery voltage of 432V. The other two figures correspond to more reasonable battery voltages; possibly the 335A limit is the max per output connector rather than the max for the unit as a whole.

Well, at 480V it's specified to deliver 145kW continuous. Somehow that 335A max limit needs to be explained away - either its a per-output limit as I speculate above, or the 335A can be exceeded short-term even if it can't be sustained continuously.

The fixed 192A input current limit does scream "stack of twelve EU charge modules" to me I have to say. They're equally able to consume 16A per phase each, but their output power varies based on input voltage.

Which in turn begs the interesting question of "If I had a 3 phase EVSE at home and I bought a 480V step-up transformer, could I charge at 26.5kW?"

wk057 · Feb 5, 2015

If they're anything like the chargers in the US, they limit to 10kW regardless of input voltage and will drop amperage downward to do so as input voltage rises to stay <= 10kW.

FlasherZ · Feb 5, 2015

mgboyes said:
Which in turn begs the interesting question of "If I had a 3 phase EVSE at home and I bought a 480V step-up transformer, could I charge at 26.5kW?"

We don't know if the EU-spec chargers in the car can take 277V, or if the switching power supply in the wall connector would be able to handle it. We generally believe that the US chargers can, as Tesla has said those chargers are used in the supercharger and we know that the superchargers use the L-N 277V voltage of a 480V AC service. In theory, on a US dual-charger car, you could then get 22 kW charging, although I think someone who had played with it said that the car actually caps total charge power at 20 kW by reducing amps in the face of a higher voltage. The US charger labels show 85-265V (so maybe superchargers are running them out of spec, or the labels are being conservative). The EU chargers have a lower tolerance for voltage, if I recall correctly.

arg said:
I don't think this is quite accurate. The original US units were 120kW, more recent ones "135kW". I strongly suspect that the more recent US installations are using the same charger modules as used in the European cars, though that is just speculation of course.

There's no reason for Tesla to use the EU chargers in the supercharger cabinets, even in EU environments; it would be far less efficient to use them. Each of the sub-chargers in the EU modules is a discrete charger capable of only 16A. A collection of them would have more overhead and cost with no benefits to be gained when compared to using the larger 40A single-phase US models, since the goal is to parallel them to create a large DC bus.

The reason they use the step-up to 480V is so that they can maximize the power from the charging units. At 230V (the L-N voltage of a 400V 3-phase system), 12 chargers in parallel can only achieve 110 kW (230 * 40 * 12). At 277V (the L-N voltage of a 480V 3-phase system), 12 chargers in parallel @ 40A can achieve 133 kW. This is all prior to overhead, of course, which is why the 135 kW units got another 3 chargers - for a total input load of 166 kW.

(Note: that planning site is giving me an error so I can't pull the spec sheet in question)

dpeilow · Feb 5, 2015

Why not just put a couple more charger modules in the supercharger rather than go to the extreme of fitting a large, bulky, expensive transformer at each site?

wk057 · Feb 5, 2015

dpeilow said:
Why not just put a couple more charger modules in the supercharger rather than go to the extreme of fitting a large, bulky, expensive transformer at each site?

The transformer is probably cheaper.

dpeilow · Feb 5, 2015

I looked at various transformers for a previous project. At those power levels I doubt it.

Mario Kadastik · Feb 6, 2015

Guys, there are various supercharger topics, this is about the CHAdeMO adapter...

arg · Feb 6, 2015

FlasherZ said:
We don't know if the EU-spec chargers in the car can take 277V, or if the switching power supply in the wall connector would be able to handle it. We generally believe that the US chargers can, as Tesla has said those chargers are used in the supercharger and we know that the superchargers use the L-N 277V voltage of a 480V AC service. In theory, on a US dual-charger car, you could then get 22 kW charging, although I think someone who had played with it said that the car actually caps total charge power at 20 kW by reducing amps in the face of a higher voltage. The US charger labels show 85-265V (so maybe superchargers are running them out of spec, or the labels are being conservative). The EU chargers have a lower tolerance for voltage, if I recall correctly.

Well, my speculation was based on:

The early superchargers were definitely based on 12 off US-spec chargers (Tesla said so, and the EU modules didn't exist at that time)
The EU modules have a higher power rating at nominal voltage. As used in the cars, US chargers accept 40A, while EU chargers accept 48A. At nominal voltage they are rated at 10kW vs 11kW; if the US ones really cap at 10kW that's a further difference from the EU ones which, in the car application, will take 3*16A up to at least 245V = 11.7kW (observed here in the UK where our '230V' is usually on the high side for historical reasons).
Around the same time the EU cars started to be delivered, GenII superchargers appeared with a higher rating, almost exactly in proportion to the rating difference in the chargers (120 * 11/10 = 132)
The GenII cabnets appear from photos to be the same size as GenI (though I haven't measured them - if they are indeed bigger that would blow my theory out of the water).

There's no reason for Tesla to use the EU chargers in the supercharger cabinets, even in EU environments; it would be far less efficient to use them. Each of the sub-chargers in the EU modules is a discrete charger capable of only 16A. A collection of them would have more overhead and cost with no benefits to be gained when compared to using the larger 40A single-phase US models, since the goal is to parallel them to create a large DC bus.

Not necessarily. It's evident from teardown photos of the US version that it's a digtally controlled regulator with multiple IGBTs in parallel to achieve the rating, so there's no obvious efficiency disadvantage in splitting them into three groups. Probably the EU version costs a fraction more to build, but not a significant amount in the scale of things.

(Note: that planning site is giving me an error so I can't pull the spec sheet in question)

Strange - the link works for me (even from a different PC to the one I first posted it). The download links off that page do time out (which is why I linked to the surrounding page rather than directly to the document).

rossifan46 · Feb 6, 2015

specs for a 400 to 480V transformer

mgboyes said:
Sorry but that bears no resemblance to what I have seen in the UK, and goes against all the arguments/logic about how SC cabinets are configured in the US and elsewhere that I have seen. Do you have a source/reference for this? Examples of such transformers in use? Which country have you seen this setup in?

US SC units are a stack of twelve x 10kW charge units (same as the ones in US cars) hence 120kW.
EU SC units are a stack of twelve x 11kW 3 phase charge units (same as the ones in EU cars) hence 132kW which is rounded up to 135kW for labeling.
There are no step up transformers at any UK supercharger site that I've visited or seen photos of.
In countries where 400V 3 phase power is common why on earth would you go to the huge expense and hassle of stepping up slightly to 480V?

The rating label on a Gen II 135kW supercharger says input of 380-480V 3phase at up to 192A. 192A is exactly what you need to drive 12 European charge modules at 16A per phase each. That's 133kW total at an EU standard 400V (230V L-N on each phase).

The maximum current output of a Gen II 135kW supercharger is 330A continuous so it could not possibly charge two cars at a total of 150kW between them. When you and an adjacent car saw a total that exceeded 120kW are you sure that you were both connected to the same SC cabinet. At some large SC sites the paired bays aren't adjacent - they sometimes install them as e.g. 1A 2A 3A 4A 1B 2B 3B 4B.

Here's the spec label on one of the SC cabinets at the Maidstone charger in the UK:

View attachment 71049

Here is a picture of the specs for a 400 to 480V transformer used for the SC in Vinje, Norway.

jkirkebo · Feb 6, 2015

FlasherZ said:
The reason they use the step-up to 480V is so that they can maximize the power from the charging units. At 230V (the L-N voltage of a 400V 3-phase system), 12 chargers in parallel can only achieve 110 kW (230 * 40 * 12). At 277V (the L-N voltage of a 480V 3-phase system), 12 chargers in parallel @ 40A can achieve 133 kW. This is all prior to overhead, of course, which is why the 135 kW units got another 3 chargers - for a total input load of 166 kW.

I have another theory here.

Using EU chargers, each can consume 48A. At 230V (phase-N voltage for a 400V system) that is 11kW, or 132kW for a stack of 12.
Stepping up to 480V results in a phase-N voltage of 277V. 48A*277V=13.3kW per charger. A stack of 12 would draw a total of 159.6kW, almost exactly the rating of the transformer.

For an input of 159.6kW an output of ~150kW is quite realistic.

This also means that it's the input stage that limits the EU chargers. Charging speeds of 23,5kW have been witnessed in EU cars when voltage was high (like 245V).

The question then remains, if such a transformer could be used at home for AC charging

arg · Feb 6, 2015

jkirkebo said:
I have another theory here.

Using EU chargers, each can consume 48A. At 230V (phase-N voltage for a 400V system) that is 11kW, or 132kW for a stack of 12.
Stepping up to 480V results in a phase-N voltage of 277V. 48A*277V=13.3kW per charger. A stack of 12 would draw a total of 159.6kW, almost exactly the rating of the transformer.

That's not just a theory, it's established fact - per the Tesla document I linked to above.

For an input of 159.6kW an output of ~150kW is quite realistic.

145kW according to the document.

What's puzzling is why they consider the transformer worth the money (given that even 380V can very nearly charge a single car at full rate), compared to, for example, spending that money on an extra supercharger cabinet. And why particularly in Norway, when we don't seem to be seeing them elsewhere in Europe?

jkirkebo · Feb 6, 2015

arg said:
What's puzzling is why they consider the transformer worth the money (given that even 380V can very nearly charge a single car at full rate), compared to, for example, spending that money on an extra supercharger cabinet. And why particularly in Norway, when we don't seem to be seeing them elsewhere in Europe?

Because each SC cabinet feeds two cars. 400V was enough when the cap was 290A, no single car can draw more than approx 105kW from those older SCs. When the limit was raised to 335A or so, more capacity was needed. A step-up transformer was probably the easiest and cheapest way to achieve that. The other options was to put more chargers in each cabinet, or keep the 105kW limit per car.

Why the transformers are not used outside of Norway (do we know that for sure?) I do not know. Maybe because of low utilisation so far?

FlasherZ · Feb 6, 2015

Looking at the document, I'm coming around to agree with you. It does all line up to being 36 chargers (12 modules x 3 sub-chargers), each at 16A, 12 sub-chargers per phase. All the numbers line up.

It just seems like more components to fail, but I can see it now.

It also has another side benefit, in that the failure of a sub-charger module means that only 4 kW is lost instead of 11 kW.

As for why the step-up is there, remember that you also have to deal with overhead from the AC->DC conversion as well, which is as much as 10% according to measurements that people have done at home.

And finally, remember that Tesla doesn't use the 400V or 480V L-L voltages, they use the L-N voltages (230V, 240V, 277V).

andrewas · Feb 6, 2015

The statement that a supercharger consists of a stack of the same chargers as in the car is true and confirmed by Tesla leadership in a talk. The number of chargers in the stack per supercharger has been growing over time as they have "upped" the supercharger capacity. It also means that a supercharger can loose capacity by one or more of the chargers in the stack being faulty. This is monitored by Tesla who will dispatch repairs as needed. The good side of it is the supercharger will continue to be working - only less efficiently as long as there are working chargers left in the superchargers stack.

Mark Petersen · Feb 9, 2015

if I remember correctly the 400v network in norway is different than the rest of Europe
I believe they primary uses a Delta configuration where most on Europe uses WYE configuration
If I remember correctly tesla 3phase 400v onboard charger is designed to uses a WYE configuration
http://ecomento.com/2014/01/24/tesla-motors-blames-norwegian-electricity-grid-cold-weather-model-s-charging-failures/
so my guess the reason that we see the 3 phase autotransformer is that it is used as a Δ-Y transformer
Y-Î” transform - Wikipedia, the free encyclopedia

sjokomelk · Feb 9, 2015

Mark Petersen said:
if I remember correctly the 400v network in norway is different than the rest of Europe
I believe they primary uses a Delta configuration where most on Europe uses WYE configuration
If I remember correctly tesla 3phase 400v onboard charger is designed to uses a WYE configuration
http://ecomento.com/2014/01/24/tesl...-grid-cold-weather-model-s-charging-failures/
so my guess the reason that we see the 3 phase autotransformer is that it is used as a Δ-Y transformer
Y-Î” transform - Wikipedia, the free encyclopedia

All Tesla Superchargers in Norway have new, dedicated transformers from the power grid, and they are all 400V WYE.
The 400V grid in Norway is exactly the same as in any other country. It's the 230V three phase grid that is our "special" delta configuration.

Search

EU spec Supercharger

jkirkebo

Model S P85+ VIN 14420 EU

arg

Active Member

mgboyes

Member

wk057

Former Tesla Tinkerer

FlasherZ

Sig Model S + Sig Model X + Model 3 Resv

dpeilow

Well-Known Member

wk057

Former Tesla Tinkerer

dpeilow

Well-Known Member

Mario Kadastik

Active Member

arg

Active Member

rossifan46

New Member

jkirkebo

Model S P85+ VIN 14420 EU

arg

Active Member

jkirkebo

Model S P85+ VIN 14420 EU

FlasherZ

Sig Model S + Sig Model X + Model 3 Resv

andrewas

Member

Mark Petersen

Model S EU P71

sjokomelk

Member

Similar threads