Pictures of V3 Supercharger cabinets

[juliusa]

There will still be a bottleneck, but instead of it being at every pair of stalls, it will be the total site capacity. In reality, how often will an 8-stall site be trying to charge 8x cars capable of 250 kW, and every single one of them will be at the optimal SOC and temperate to receive 250 kW? I would expect Tesla are sizing the grid connection such that someone would never know there is a bottleneck.

That makes a lot of sense. In the few cases where it is limited by the site capacity, so be it - it will avoid a large demand charge for that month where in one hour of the month there was higher usage.

Sharing the capacity across the whole site is a much nicer user experience - you don't have to try to find an unpaired stall, and if there are no unpaired, playing stall roulette trying to guess which open stall will give you the best charge.

I do question the ones in Fairfield CT though - 8 talls/500kVA and 12 stalls/750kVA? That is a mere 60kW per stall. Could be that is just a start - when there are 250kW cars out there, perhaps they will upgrade the feed at that time.

 

[mociaf9]

There will still be a bottleneck, but instead of it being at every pair of stalls, it will be the total site capacity. In reality, how often will an 8-stall site be trying to charge 8x cars capable of 250 kW, and every single one of them will be at the optimal SOC and temperate to receive 250 kW? I would expect Tesla are sizing the grid connection such that someone would never know there is a bottleneck.

Plus, the cars can't take 250 kW for very long. You'd pretty much have to intentionally engineer it to try and get more than 2 or 3 cars at a single installation all pulling the max rate at the same time.

 

[Big Earl]

It’s possible that 750 kVA is all the utility is able to provide in that location and Tesla chose to maximize the number of stalls rather than maximizing power output per stall. I need to look into this location some more. I was there a few days ago but it was raining and I didn’t get any good pictures. The pics in the Fairfield thread don’t show enough conduit and wiring detail... can someone local get some better shots of the sites?

 

[Chuq]

Any idea what the usage pattern of that location is? Do vehicles typically move on after charging or sit there for ages when done? Perhaps if Tesla has identified that half the cars sitting there at any one time are idle (completed charging) or trickle charging to 100% then 60 kW per stall averaged across the whole site may be sufficient.

 

[Big Earl]

It’s a service plaza, so more people will want to keep moving. That said, getting food and going to the bathroom will put you well into the charge speed taper.

 

[juliusa]

It’s possible that 750 kVA is all the utility is able to provide in that location and Tesla chose to maximize the number of stalls rather than maximizing power output per stall.

Could be. Although it is odd that they did 8 stalls/500kVA on one side, and across the highway 12 stalls/750kVA.

The pics in the Fairfield thread don’t show enough conduit and wiring detail... can someone local get some better shots of the sites?

I could - though the Darien V3 sites are easier (closer), which is also a V3 site (no transformers yet though).

As best I can tell, the sites would look the same regardless of the size of the service feed, right? For example, I see no difference between the two Fairfield CT sites, other than the number of stalls and size of the transformer the utility put in. I think the transformer pad is the same size, and the conduit sizes look the same.

 

[Big Earl]

Could be. Although it is odd that they did 8 stalls/500kVA on one side, and across the highway 12 stalls/750kVA.

I could - though the Darien V3 sites are easier (closer), which is also a V3 site (no transformers yet though).

As best I can tell, the sites would look the same regardless of the size of the service feed, right? For example, I see no difference between the two Fairfield CT sites, other than the number of stalls and size of the transformer the utility put in. I think the transformer pad is the same size, and the conduit sizes look the same.

They should be similar, yes. I’m most curious to see how many charging cabinets they install. Should be three cabinets for twelve stalls if they’re V3 sites.

 

[juliusa]

They should be similar, yes. I’m most curious to see how many charging cabinets they install. Should be three cabinets for twelve stalls if they’re V3 sites.

These pictures show the charger slabs they poured. The top one is a little hard to see, but the bottom one (8-stalls, 500kVA transformer) shows two sets of conduit presumably for two charger cabinets.

Supercharger - Fairfield

 

[Cuttin]

So I just looked at the covered up new SC in cherry hill NJ at wawa on rt 73. Might be called Marlton not cherry hill. Wawa is almost open.

The cable is very thin which is sign of v3. And they are keeping them covered so more of a clue.

Also there didn't seem to be numbers on bottom. Which would make sense bc there is no sharing so who cares what number it is.

Heres a pic of the cabinet. There are 8 chargers. And what looks like to me two cabinets both a bit different looking than I'm used to. One of them said 480volt. Not sure what that means.

 

[qdeathstar]

480v refers to the AC voltage in the panel. 277/480 3 phase.

 

[Cuttin]

480v refers to the AC voltage in the panel. 277/480 3 phase.

Lol. In layman terms.... Does that point to v3 or v2?

 

[Big Earl]

Lol. In layman terms.... Does that point to v3 or v2?

Could be either. 480 volts is what supplies all Tesla Supercharger locations regardless of type.

The cement pads and electrical conduit layout shown in some of the pictures indicates that this will be a V3 location.

 

[Chuq]

Also there didn't seem to be numbers on bottom. Which would make sense bc there is no sharing so who cares what number it is.

Interestingly enough some of the other V3 sites have had conduits labelled "1A, 1B, 1C, 1D" and "2A, 2B, 2C, 2D" - even though there is no pairing they probably still need labels of some sort so that faulty stalls can be identified. However I don't think I've seen pictures of V3 stalls with labels?

 

[Cuttin]

Sorry. I gave wrong name for location. I think it's called maple shade, not cherry hill or Marlton.

Darn it, I love her for years and mix it up. Whole bunch of little towns in a small area. Sorry.

 

[Chuq]

One thing that I've noticed is that some regions seems to be getting V3 superchargers with 3 stalls to a cabinet.

Check out all the V3 sites in China at supercharge.info - many are 3, 6, 9, 15 or 21 stalls.

In Norway there's also a small number of 3, 6, and 9 stalls.

Most of the Trans Canada Highway sites are 6 stalls.

We've only just got our first V3 sites in Australia - actually four of them! And they are all 3 or 6 stall.

This is a photo of the cabinet at one of the Australian ones. I notice it is labelled as "CHARGER, 350 kVA, 480VAC, 3 OUTPUT"

Am I right in assuming this takes 350 kW of grid input and splits it between 3 stalls (each of which is individual capable of up to 250 kW)?

Does anyone have an equivalent image from one of the 4 output units?

 

[mociaf9]

Interesting, that's definitely different. And as expected it also has a different part number. Here's one with "4 Output", but I don't recall which specific location it comes from:

 

[Chuq]

Interesting, that's definitely different. And as expected it also has a different part number. Here's one with "4 Output", but I don't recall which specific location it comes from:

Interesting that apart from the number of outputs every single other value is 100% identical!

I've never really paid attention before, but the "AC input - maximum continuous power" is 350 kVA - which, I believe, is (closely enough) equivalent to 350 kW. So each cabinet can only use a grid supply of 350 kW for the 3 or 4 stalls?

The "DC input power" is 575 kW but I always thought that was for any stationary battery or solar input.

There must be more to it than that or I must be misunderstanding something...

 

[mociaf9]

The "DC input power" is 575 kW but I always thought that was for any stationary battery or solar input.

There must be more to it than that or I must be misunderstanding something...

The supercharger cabinets are also connected to each other via DC. So, in addition to using DC from a connected battery they can also share power between cabinets at a site.

And the 350 kVA is the maximum continuous input power from the grid. Which means that for short periods the value can be exceeded.

 

[miimura]

Let's say you have an 8 stall site in the USA and a 9 stall site in China. The USA site will have two 4 stall cabinets with 700KVA total input power. The China site will have three 3 stall cabinets with 1,050kVA total input power. In China, 4 empty cars showing up at that same time can all draw the 250kW max power, no matter which stall they connect to. The DC bus will take power from the cabinets that have extra power available. The worst case is 3 full stalls on one cabinet and one car on another cabinet. 750kW-350kW=400kW will be required from the DC bus to satisfy the demand on the full cabinet. Meanwhile, in the USA, the 4 cars could connect to one cabinet and 1,000kW-350kW=650kW would be required from the DC bus, but it can only supply 575kW. Those 4 cars would be throttled to about 230kW.

Another way to look at this is the average power across all stalls. 350/4=87.5kW per stall versus 350/3=116.7kW per stall. The chances of anyone receiving noticeably throttled Supercharging at a site with 3 outputs per cabinet is vanishingly small. It's still not very likely in the States, but it's definitely possible when a site is full.

I have to wonder if Tesla will start installing the 3 output cabinets in the States too. It would be interesting to see a statistical distribution of charging sessions versus % throttled. I define % throttled as P(delivered)/P(request)

 

[Chuq]

Thanks for the insightful posts @mociaf9 and @miimura. I made a comment elsewhere based on what you've said here. If I've interpreted what you've said correctly. I thought I'd share the comment here in case I had misinterpreted anything and corrections were necessary:

General AC / DC info:

• Supply from the grid is AC
• Supply from battery storage or solar is DC
• Cabinets are connected to each other via DC

Specs of the units:

• The line "AC input - Maximum continuous power" where it says 350 kVA, means the grid connection *per cabinet* is approx 350 kW. Note there are some losses, but peak power can also slightly exceed this, so I'll use 350 kW as an approximation.
• The line "DC input / output" where it says 575 kW means that each cabinet can accept 575 kW from DC sources, whether that be solar, battery, or other cabinets, and can also supply the same to other cabinets.

Examples - these are based on the specs above. If my understanding/assumptions are wrong the examples will be wrong - please let me know if that's the case!

• A 1 cabinet site with no solar or battery has no DC supply, so the 350 kW is split between the three stalls (average of 116 kW each). For four stalls it would be an average of 88 kW each. I guess that's why there aren't that many of them.
• A 2 cabinet site with no solar or battery can only supply 350 kW between cabinets because that's all the 2nd cabinet grid connection can provide. So if all four stalls are packed (and the other cabinet's stalls are empty) that's an average of 175 kW each. For a three stall per cabinet site, it's 233 kW.
• A 3 cabinet site with no solar or battery means one high demand cabinet (and the other two being empty) can get 350 kW from the grid and the full 575 kW from the other two cabinets (average ~287.5 kW from each cabinet). So that's 925 kW total (average of 308 kW for 3 stalls-per-cabinet, or 231 kW for 4 stalls-per-cabinet).

Note in all cases, the average being lower than 250 kW doesn't affect the peak speed of individual stalls, and in many cases wouldn't be noticeable by the driver. The average numbers would only matter in practice if you had multiple Model 3 LR/Performances, all at about 5-10%, all with their battery pre-heated, all arriving at the same time, and all going to stalls connected to the same SC cabinet.

I mention 308 kW above, which is of course more than what the stalls will supply, but I don't know if the current limitation of 250 kW is the car or the stall. Maybe the existing hardware, without any changes, will be able to charge a S/X Plaid, Cybertruck or Roadster at greater than 250 kW?