What limits rate of SuperCharging?

[Cottonwood]

I agree. The probability that two cars will arrive simultaneously, both at very low state of charge, and both connect to the same Supercharger, is low. And even if it does happen, the effect is probably felt for only 10 minutes or so, as the power ramps down once the battery gets to half or so. 67.5 kW is still a pretty impressive power to be charging at.

Interesting thoughts. I will send the supercharger team a recommended fairness algorithm. What do you think about this as a strawman based on arrival time differences?

• 0-1 minute — Equal priority and sharing. If one car can't take 50%, then the other car gets whats left. We don't need rally races to see who can plug in first.
• 1-3 minutes — First to arrive gets 60% or most it can take, whichever is less; second to arrive gets rest.
• 3-5 minutes — First to arrive gets 70% or most it can take, whichever is less; second to arrive gets rest.
• More than 5 minutes — First to arrive gets 80% or most it can take, whichever is less; second to arrive gets rest. I believe this is the current split for first and second to arrive.

Of course we could avoid all hard boundaries and thresholds by the Supercharger cabinet doing a linear ramp in first to arrive priority from 50% to 80% over 5 minutes, or whatever time seems good for fairness... 5 minutes seemed pretty reasonable to me. In that amount of time the first to arrive has already been able to suck in 10 kWh and is on it's way to start tapering.

 

[brianman]

I agree. The probability that two cars will arrive simultaneously, both at very low state of charge, and both connect to the same Supercharger, is low.

In caravans for travelling, it definitely happens. This is one of the reasons that it's usually a bad idea for events like TMC Connect to travel in packs larger than 4. Also, it's trickier to keep a group of 5+ together anyway.

 

[PoweredByRain]

In caravans for travelling, it definitely happens.

Good point! Though if you're in a caravan you will hopefully be aware of the issue and not plug into the same Supercharger cabinet. (Hence your comment about max 4 cars at once, I presume. If we're talking about smaller Supercharger installations with as few as 2 Superchargers = 4 stalls, the limit becomes two...)

 

[mgboyes]

Interesting thoughts. I will send the supercharger team a recommended fairness algorithm. What do you think about this as a strawman based on arrival time differences?

• 0-1 minute — Equal priority and sharing. If one car can't take 50%, then the other car gets whats left. We don't need rally races to see who can plug in first.
• 1-3 minutes — First to arrive gets 60% or most it can take, whichever is less; second to arrive gets rest.
• 3-5 minutes — First to arrive gets 70% or most it can take, whichever is less; second to arrive gets rest.
• More than 5 minutes — First to arrive gets 80% or most it can take, whichever is less; second to arrive gets rest. I believe this is the current split for first and second to arrive.

Based on my (limited) understanding of how superchargers work I don't think this is possible. A supercharger is just a cabinet containing 12 of the same on-board chargers that work in the car, and those off-board chargers are controlled by the car as it charges (the signaling between car and supercharger is CAN) just like the onboard chargers would be. So the sharing between two cars at a supercharger can only be so granular, since each of the 12 charging units can only be controlled by one car at a time (and since they're fed from 3 phase power they may even only be switchable in groups of three). Otherwise how would two cars with different voltage needs (e.g. an empty S60 vs a nearly full S85) feed from the same supercharger? Check out the start and end of Tesla Model S / X: Supercharging a 60kW Battery from Dead, 105kW Charging Rate!!! - YouTube to see the variation in voltage being delivered to the car (from 300 to 350v) as the battery fills.

This is what leads to the commonly seen behaviour that when a second car joins the first drops to 90kW - because 3 of the charge units in the supercharger get reassigned to charge the second vehicle leaving just 9 charging the first vehicle.

Perhaps I've dreamt this, but it certainly makes a lot of sense to me.

 

[mitch672]

Supercharger power is switched in (3) 10KW modules at a time, since each of the chargers are connected to a leg of a 3 phase feed, the switching between cars is 30KW (3 chargers), 60KW (6 chargers), 90KW (9 chargers), or 120KW (all 12 chargers dedicated to 1 charging stall). Those are the capabilities, as they can be switched today. So if two Model S's arrive at the same time with a low state-of-charge, they each get 60KW, that's the best that can be done. If one is further along charging and starts tapering (below 30KW), then just 30KW can be dedicated to it, and another Model S plugged into the same Supercharger can get 90KW.

 

[arg]

Supercharger power is switched in (3) 10KW modules at a time, since each of the chargers are connected to a leg of a 3 phase feed, the switching between cars is 30KW (3 chargers), 60KW (6 chargers), 90KW (9 chargers), or 120KW (all 12 chargers dedicated to 1 charging stall). Those are the capabilities, as they can be switched today. So if two Model S's arrive at the same time with a low state-of-charge, they each get 60KW, that's the best that can be done. If one is further along charging and starts tapering (below 30KW), then just 30KW can be dedicated to it, and another Model S plugged into the same Supercharger can get 90KW.

Is that definitely still the case with the latest superchargers? There was speculation that the step up from 120kW to 135kW superchargers was by changing from the original 10kW modules to the European 11kW modules - and the European modules are already 3-phase capable at the module level so would not need to be switched in threes.

But lately there seems to have been only speculation and very little hard info on the changes to supercharger internals.

 

[Cottonwood]

Based on my (limited) understanding of how superchargers work I don't think this is possible. A supercharger is just a cabinet containing 12 of the same on-board chargers that work in the car, and those off-board chargers are controlled by the car as it charges (the signaling between car and supercharger is CAN) just like the onboard chargers would be. So the sharing between two cars at a supercharger can only be so granular, since each of the 12 charging units can only be controlled by one car at a time (and since they're fed from 3 phase power they may even only be switchable in groups of three). Otherwise how would two cars with different voltage needs (e.g. an empty S60 vs a nearly full S85) feed from the same supercharger? Check out the start and end of Tesla Model S / X: Supercharging a 60kW Battery from Dead, 105kW Charging Rate!!! - YouTube to see the variation in voltage being delivered to the car (from 300 to 350v) as the battery fills.

This is what leads to the commonly seen behaviour that when a second car joins the first drops to 90kW - because 3 of the charge units in the supercharger get reassigned to charge the second vehicle leaving just 9 charging the first vehicle.

Perhaps I've dreamt this, but it certainly makes a lot of sense to me.

Supercharger power is switched in (3) 10KW modules at a time, since each of the chargers are connected to a leg of a 3 phase feed, the switching between cars is 30KW (3 chargers), 60KW (6 chargers), 90KW (9 chargers), or 120KW (all 12 chargers dedicated to 1 charging stall). Those are the capabilities, as they can be switched today. So if two Model S's arrive at the same time with a low state-of-charge, they each get 60KW, that's the best that can be done. If one is further along charging and starts tapering (below 30KW), then just 30KW can be dedicated to it, and another Model S plugged into the same Supercharger can get 90KW.

Where did this switching in groups of three come from? Is there a reference? There is no need to do that from a power feed point of view. The charing modules have a power factor of greater than 0.95 and as long as one phase is not overloaded, the utility won't complain.

My assumption from a reasonable engineering point of view was that each of the 12 Charging Modules had a Single Pole Double Throw relay on the output that was rated at 40 Amps or so. (135kW/360V/12modules=31.25A per Module) These relays don't even have to be switched under load, because the output can be ramped to zero current before switching. That way each module could be connected to either pedestal and be set to Voltage/current desired. This means that increments of shared power are quantized to 1/12 or 8.3% of total Cabinet power, but that is fine enough that any error is in the noise.

I have only observed shared, limited charging once, but my memory without any notes was that my car (second to arrive) ramped up in power pretty evenly as the first car tapered. We really need to get some people to do more measurements to see what the real behavior is.

 

[widodh]

Where did this switching in groups of three come from? Is there a reference? There is no need to do that from a power feed point of view. The charing modules have a power factor of greater than 0.95 and as long as one phase is not overloaded, the utility won't complain.

There is a big label on the SuperChargers saying it's 3-phase input.

And on such big loads you always get 3-phase power. 3-phase is the best method of distributing such large loads.

 

[Cottonwood]

There is a big label on the SuperChargers saying it's 3-phase input.

And on such big loads you always get 3-phase power. 3-phase is the best method of distributing such large loads.

Three phase input does not require switching internal resources in groups of three or always keeping the three phases balanced. To obtain maximum power, the three phases need to be balanced, but that does not require switching internal loads in groups of three.

3-Phase power is a little more efficient in wiring than Split-Phase power, but not overwhelmingly so. In 3-Phase feeds, you need the three phase wires plus a neutral, in Split-Phase feeds, you need two line wires (~phases) plus a neutral, that is 3/2 the power for 4/3 the wire or a gain in efficiency of 12.5% in wiring efficiency. If you use only the phases or lines it is a sqrt(3) more power for 3/2 times the wire, that is a 15.5% gain in wiring efficiency. Gains or 12.5% and 15.5% are certainly worthwhile for large installations. Another reason to have 3-Phase power is to make motors and generators more efficient.

3-Phase power allows the Supercharger Cabinet to roughly keep the charging current constant, while maintaining a nearly perfect load factor of greater than 0.95, without huge 120 Hz components, but the battery already deals with 120 Hz charge current ripple at the 20 kW level when charging in a home with Split-Phase power and an HPWC.

Nothing about 3-Phase power input requires switching charging modules in groups of 3. Please give explain to me if you know why this is not so.

 

[mitch672]

for efficiency in building the SuperCharger, its very LIKELY the chargers are switched with 3 pole 40A contactor onto the one of the two charging busses (Simple 3 pole DPDT contactor, the chargers are connected to one car, or the other) for each vehicle. I highly doubt they are going through all of the added expense of wiring 3 phase power to each charger, and switching them to the vehicle buss individually. The cabinet would be even larger, and even more expensive with 12 contactors versus just 4, not to mention all of the extra AC wiring. They get to ~135KW by running each charger at 277V, and upping the limit to 40A (11080 Watts max), 11KW * 12 = 132KW, close enough for government work. All of the new greenfield installations run on 277V/480V directly.

Also running individual chargers, the 3 phase load will quickly become unbalanced, as most of the larger sites have (4) SuperChargers, serving up to 8 vehicles

 

[Cottonwood]

for efficiency in building the SuperCharger, its very LIKELY the chargers are switched with 3 pole 40A contactor onto the one of the two charging busses (Simple 3 pole DPDT contactor, the chargers are connected to one car, or the other) for each vehicle. I highly doubt they are going through all of the added expense of wiring 3 phase power to each charger, and switching them to the vehicle buss individually. The cabinet would be even larger, and even more expensive with 12 contactors versus just 4, not to mention all of the extra AC wiring. They get to ~135KW by running each charger at 277V, and upping the limit to 40A (11080 Watts max), 11KW * 12 = 132KW, close enough for government work. All of the new greenfield installations run on 277V/480V directly.

Also running individual chargers, the 3 phase load will quickly become unbalanced, as most of the larger sites have (4) SuperChargers, serving up to 8 vehicles

I sort of agree with what you said, but here are some details that I would like to clarify.

I agree that putting 4 charging modules on each phase, 3 phases total is a great way to distribute the 3-Phase power internally to the 12 charging modules and is the most probable design.

I agree that there is no need to do AC switching, only DC output switching.

Why not just put a single pole double throw (SPDT) relay on the output of each module. There is no need to gang into threes with 3-pole relays. The only goal that is needed is to roughly balance the load on each of the 3 phases of the AC in.

They had to do something else to get to 135 kW output than just use 40 Amps at 277 Volts for each module. I agree that 40A*277V*12 = 132 kW, but that is AC input power. In every case when quoting Supercharger power, Tesla has quoted DC output power. I think that the numbers you use work for 120 kW DC out with 90% charger efficiency, 132kW*90% = 119kW, pretty close to 120 kW. To get 135 kW DC out at 90% efficiency requires 135kW/90% = 150kW AC input power. To do that at 277 Volts, means 12 charging modules need to accept 45 Amps in each or they need up the number of 10 kW modules to 15.

If each Supercharger Cabinet takes in 3-Phase power and keeps the load roughly balanced across its internal charging modules, then the entire Supercharger Site stays roughly balanced for 3-Phase power loading. Nothing about the site gets way out of balance.

 

[arg]

Nothing about 3-Phase power input requires switching charging modules in groups of 3. Please give explain to me if you know why this is not so.

It would do if they were taking 3-phase delta with no neutral and still trying to hang 277V loads off it.

It is quite clear that they are NOT doing that, but such thinking could be where this rumour originated?

 

[WarpedOne]

The correct unit of "charging rate" is not kW (10, 20, 120, 135kW, ...) nor it is time (4 hours for full charge, half a charge in half an hour, ...)
The correct unit of charging rate is mph - how many miles it pushes into the battery in some amount of time.

Current superchargers manage up to 300mph charging, 120kWh battery of same existing cells as in 85kWh would manage 420mph charging. Bigger battery can charge faster. Of course you have to feed it more kW, but that is only technical detail that don't mean much in everyday life.
In every day life we operate on distance and time.

 

[Cottonwood]

The correct unit of "charging rate" is not kW (10, 20, 120, 135kW, ...) nor it is time (4 hours for full charge, half a charge in half an hour, ...)
The correct unit of charging rate is mph - how many miles it pushes into the battery in some amount of time.

Current superchargers manage up to 300mph charging, 120kWh battery of same existing cells as in 85kWh would manage 420mph charging. Bigger battery can charge faster. Of course you have to feed it more kW, but that is only technical detail that don't mean much in everyday life.
In every day life we operate on distance and time.

kW and charging mph are interchangeable. For DC charging of an 85, the conversion is 300 Wh/mi. With a maximum charging power of 120 kW from an existing Tesla Supercharger into a single Tesla Model S, that is a charge rate of 400 rated miles per hour. The "A" batteries like my sig 85 are limited to 90 kW charging power or 300 mph maximum charge rate.

 

[mgboyes]

The correct unit of "charging rate" is not kW (10, 20, 120, 135kW, ...) nor it is time (4 hours for full charge, half a charge in half an hour, ...)
The correct unit of charging rate is mph - how many miles it pushes into the battery in some amount of time.

IMO mph is a cute way of representing charge rate, but not at all useful in these technical discussions because it varies depending on whether you have your car configured to rated or ideal miles (and it varies between countries, because for example in the UK we have rated and typical miles where rated is 312 and typical is 245 for an 85).

In the UK and EU our onboard chargers each consume 3 phase directly. Since that capability exists in the chargers, it is quite possible that Tesla could build a supercharger cabinet for any country where each of the 12 modules was connected to all 3 supply phases, and therefore each of the 12 could have its output directed individually to one of the two connected cars without ever drawing unbalanced AC power. Whether they are doing that or not, I do not know.

Clearly it is possible to have a 3 phase supply and to draw uneven loads from it, but I'd guess that the cost to Tesla is lower if they run in a mode where they guarantee to draw power evenly across the phases since it requires less complex load planning for the power company.

 

[Cottonwood]

Clearly it is possible to have a 3 phase supply and to draw uneven loads from it, but I'd guess that the cost to Tesla is lower if they run in a mode where they guarantee to draw power evenly across the phases since it requires less complex load planning for the power company.

Even or balanced loading of a 3-Phase connection to the utility is not black or white, but a matter of degree. Clearly, a Supercharger Cabinet does not want to load up one phase to its maximum then on to the next phase, etc. If the cabinet balanced the 3-Phase loading within the resolution of 1 in 12 charging modules, then the worst case imbalance is 1/12 or 8.3%. That is not bad. On top of that, by definition, the Cabinet will draw a balanced load when it is operating at maximum output. That is a modest amount of imbalance, but well within what the utility can easily accommodate and typical of many commercial loads.

 

[WarpedOne]

kW and charging mph are interchangeable

Not really, because:
a) 120kW says nothing about what it means in real life. 100mph says everything - you wait one hour and you just got 100miles of range
b) 120kW does not take into account cars efficiency, 100mph does. 120kW can mean 300mph charging for one car and 350mph for a different car
c) 120kW knows nothing about my exact average energy use. 100mph already knows how much energy I use. 120kW may mean 300mph for me and 330mph for you. How many miles will you get from SS in one hour - 330 "your" miles, and I will get 300 "my" miles.

IMO mph is a cute way of representing charge rate,

Not quite, "mph charging rate" is the only really meaningful one that can take into account all there is to take into account. If one uses kW he must than 'translate' what it really means for him and his case. This translation is conversion of those kW into mph...

 

[sigurdi]

Who uses mph when filling diesel or gas?

When writing about superchargers and rate delivered kW is the right measure to use.
It wil be the same everywhere EU, UK, USA.

mph is also a average of the charge, it is based on default values not what you actualy is using.

And if I wrote that I charged on the 135kW EU SC at 500km/h it will say you nothing, if I told you I got 122kW at peak you could compare that to the maximum you got on your SC.

Here is a picture of one of the Gen II, 135kW EU SC chargers. (Åmot, Vinje Norway)

It is feed with 480v, highest charg rate seen on the new 135kW chargers, so far in Norway is 116kW (328A x 355v)

Midlefart in denmark have documentation on 122kW (341A x 358v) - not sure if this is the same 135kW SC charger.

 

[WarpedOne]

Who uses mph when filling diesel or gas?

Who uses gallons/minute when filling diesel or gas? Irrelevant questions.

Explain to any nonelectrical Joe what 120kW means for chargning and try not to use words miles and hour.
Not a single person had any idea what 120kW or 20kW or 10kW or 15kW actually ment for them.
All of them understood immediately what "200 miles in half an hour" means for them. Those are terms they understand.

ALmost noone understands 120kW. Even most of Model S owners claim they have 85kW battery ...

 

[sigurdi]

It is better you share the right knowledge with Joe, that kW is the power delivered at one moment in time and kWh is the amount of energy stored or used.
Kilowatt hour - Wikipedia, the free encyclopedia

I always use a electrical heater as a example, most people know what a 1000W or 1kW heater delivers and that is uses 1kWh if you leave it one for one hour and also what they pay for their electricity.

The big problem when using miles/h, km/h when explaining energy that it is in conflict with the measurement of speed and it depends on the car used, the driver, and can make people more confused in the long run.
After giving the right information like in 30 minutes I have filled 50kWh into my battery and it is enough energy to drive 200 miles with my car. ;-)

Example why miles/h is a bad idea and gievs more confusion than information (numbers is just from the air since I do not use miles at all (only km)

You have just told Joe that you charged your Tesla with 200 miles in half an hour, he then walk over to a leaf that is charing on a Chademo 50kW charger.
The leaf owner explains that he is charging with 175 miles in half an hour.
Joe comes back to you and says that the leaf charger is almost as powerful as the Tesla charger. ;-)

Or in a year you takes Joe on a trip with your Model X, stops on a SC and you get 180 miles in half an hour, why do the charger deliver less energy to the Model X is Joes question.

When writing about charging in forums like this, everybody should use kW and kWh since it is the same all over the world and you do not have to explain if it is rated, range, what car you have........

Instead of sharing misinformation, educate people. :biggrin: