June 20th Speculation

[Yggdrasill]

See my post (#250) on:
http://www.teslamotorsclub.com/show...ll-be-accomplished/page25?p=362836#post362836

We're already talking about $6000 per vehicle that needs to be recovered for swapping. At some point it just starts being too many things to roll into the price.

Too many unknows.

There will actually be many hours between swapping. If you make swap stations too dense (too many cars / swap station) you won't have enough power to feed them. A swap station (and SuperCharger for that matter) needs to run at an average of 6 swaps per day to be energy neutral with 200 solar panels.

I can't say I care about this very much. The SC here won't have solar panels, instead they hook them up to the 11 or 22 kV grid, with a ~1 MW transformer. A transformer of that size costs maybe $100k and then you pay maybe $0.04/kWh in grid connection fees.

I don't know the specifics of the US, but I'm fairly sure it would be cheaper to hook the SS/SC up to the grid than to cover the entire usage through solar panels.

 

[WarpedOne]

Turns out that a Panasonic NCR18650A, which is what Tesla uses, has a maximum charge rate of 1C (3.1 Amps at 3.7V).

Well, you are correct in principle but details are much more involved than that. Charging rate of 1C just means the battery will be full in one hour.

Now, what is a full battery?
This isn't a bottle with a well defined volume but more like a spring - when is a spring fully stretched?
Definition of "full" is a trade off between desired life and capacity. Also there is nothing magical in one hour charging, it is an arbitrary classification that is easy to comprehend and is not affected by capacity, just chemistry.
Also, charging power is never constant. At low SOC levels, internal resistance rises roughly linearly with time, at higher SOC levels it rises super-linearly.
To charge with a constant current, charger must constantly increase the voltage to some predefined max voltage (again determined from desired life and capacity).
Internal resistance causes heating that effect resistance etc.
It is black art and rocket science. It is easy to charge a battery slowly with low currents. It takes a lot of knowledge and statistics to really determine where is the optimum.

Elon already spilled some interesting beens: SC doesn't effect the battery life enough to matter, high SOC does the damage.
Battery must already output up to 400kW or more than 5C. Granted, those are peaks, average output is much lower, under 1C and charging is a constant burden.

Do we know if MS85 charges faster than MS60?

MS85 should see 40% faster charging if charging was indeed limited by what cells can absorb.
If MS85 and MS60 get similar charging speeds (in mph) than cells in MS60 already see much higher C-rated charging than cells in MS85 and there is space for 40% faster charing for MS85.

 

[deonb]

Too many unknows.

I can't say I care about this very much. The SC here won't have solar panels, instead they hook them up to the 11 or 22 kV grid, with a ~1 MW transformer. A transformer of that size costs maybe $100k and then you pay maybe $0.04/kWh in grid connection fees.

I don't know the specifics of the US, but I'm fairly sure it would be cheaper to hook the SS/SC up to the grid than to cover the entire usage through solar panels.

Fair enough, but the cost basis used on the other thread is that SuperChargers are energy neutral (solar), and thus you don't charge for them. Hence, you don't need to charge for energy for swappers either, since a user will at any point either be a (free) SuperCharger user or a Swapper user, but it won't be both.

If you instead draw power from the grid, it's not a wash and you need to pass this on to users (which adds $10 to the per-charge/per-swap rate).


The SC math actually works out fine with solar over here. The $2000 per car more than covers what would be needed to build the SC network. At $300k/SC it means 150 cars is enough to pay for a single SC. And at 6 charges per day, you can cycle each car through every 25 days, or ~15 times per year without needing to even rely on the economics of grid storage.

 

[Yggdrasill]

The calculations should work quite well without solar as well.

Here, the average driver will need maybe 10 supercharges per year. And each car will probably be driven for an average of 15 years. That works out to around 150 supercharges per car.

Each supercharge will probably replenish in the area of 60 kWh, so that works out to around 9000 kWh over the life of the car.

Each kwh including grid connection fees costs around $0.14, so that works out to around $1260. Here you need to pay about $2100 for supercharger access, so that leaves about $900/car after electricity.

If Tesla sells 2000 cars heare each year, that means that Tesla has $1.8 million per year to spend on superchargers. Considering that Norway only needs around 15 SC locations to be quite well covered, the economics should work out just fine.

I haven't looked at the figures for the US, but I'm sure it's better to have solar than to not have it. Even so, I'm also fairly sure that connecting to the grid isn't disasterous for the economics.

 

[deonb]

I haven't looked at the figures for the US, but I'm sure it's better to have solar than to not have it. Even so, I'm also fairly sure that connecting to the grid isn't disasterous for the economics.

I think it's about even. When SolarCity installs panels on their lease basis, they would only do that in areas that charge more than 11c per kWh for grid connections. Below that the grid becomes cheaper than what SolarCity can provide, even amortized over lifetime.

So solar really just brings the cost of charge down to 11c per kWh, which you can get in some places anyway. What it DOES do however is to create hundreds upon hundreds (or thousands) of big advertising boards that stares people in the face who still want to still use the long chimney argument against Tesla.

I would think that is less of a concern in Norway since you don't constantly have to battle Fox news and the GOP over there just to even start to have a factual debate.

 

[Yggdrasill]

I think it's about even. When SolarCity installs panels on their lease basis, they would only do that in areas that charge more than 11c per kWh for grid connections. Below that the grid becomes cheaper than what SolarCity can provide, even amortized over lifetime.

So solar really just brings the cost of charge down to 11c per kWh, which you can get in some places anyway. What it DOES do however is to create hundreds upon hundreds (or thousands) of big advertising boards that stares people in the face who still want to still use the long chimney argument against Tesla.

I would think that is less of a concern in Norway since you don't constantly have to battle Fox news and the GOP over there just to even start to have a factual debate.

I wish we could have a factual debate here, but it's not very easy. I'm constantly running into the argument that any power we consume could have been exported to replace coal somewhere else. So, even though the power we actually use is 95% hydropower, there are people saying electric cars run on coal power (not even grid-mix!).

I agree that it's good to combat the long tail pipe argument - but it shouldn't be combated at any cost. You said in the other thread that 12,000 SC locations would be needed for 2 million cars. That would be a ridiculous waste of money. Each station would be woefully under-utilized.

12,000 stations with 10 spots each would be 120,000 SC spaces. Each SC space should on average per day be able to serve at least 10 cars for 20 minutes. If each SC-capable car needs 10 supercharges per year, 12 000 SC stations should be able to serve in the area of 44 million cars. So basically we disagree by a factor of about 20.

Given the figures in the other thread, you basically think that each SC spot should on average serve around 0.5 cars per day for 20 minutes...

 

[deonb]

I wish we could have a factual debate here, but it's not very easy. I'm constantly running into the argument that any power we consume could have been exported to replace coal somewhere else. So, even though the power we actually use is 95% hydropower, there are people saying electric cars run on coal power (not even grid-mix!).

I agree that it's good to combat the long tail pipe argument - but it shouldn't be combated at any cost. You said in the other thread that 12,000 SC locations would be needed for 2 million cars. That would be a ridiculous waste of money. Each station would be woefully under-utilized.

12,000 stations with 10 spots each would be 120,000 SC spaces. Each SC space should on average per day be able to serve at least 10 cars for 20 minutes. If each SC-capable car needs 10 supercharges per year, 12 000 SC stations should be able to serve in the area of 44 million cars. So basically we disagree by a factor of about 20.

Given the figures in the other thread, you basically think that each SC spot should on average serve around 0,5 cars per day for 20 minutes...

It's not me saying that, it's Elon... The SC's during the week is expected to be "empty". And on weekends they'll be "sized for peak usage" for Friday's and Sundays (i.e. you would never have to wait for a spot even on those days). So even with your 10 cars for 20 minutes per spot hypothesis (replace 'spot' with 'charger' and I'm with you), it would only be for 2 days per week.

We also know they're meant to be solar positive using 300W max cells. So we don't really have to guess much further than that.

12 parking spots can host 200 solar panels. 200x 300W solar panels can deliver 10x 35kWh charges per day (at most), which is 70 per week. Meaning you have 35 charges on Friday and 35 on Sunday. With 12 parking spots (meaning 3 chargers using the new 4-1 ratio) it means you shouldn't have to wait, but it's actually right on the cusp - not terribly oversized. (Even you used the number of 10 per spot. This is just 10-per-charger. The extra spots are mostly there to support the Solar panels, and for the convenience of not having to "idle" behind someone that's currently charging at a low rate or ran off shopping).

But you can actually only have 10% of the days quota arriving within the same 20 minute window before a wait / slow charge experience will happen.

So whether or not it's a good idea to do things this way vs. cheaper & smaller but as a result having queues on peak times and drawing power from the grid is left up to you to decide (for your own company ). But Elon decided he doesn't want queues during peak times, and he wanted Solar positive. And it all works out (very) favorably financially, so why not.

I wish the battery swapping had numbers this cut-and-dry.

 

[JRP3]

The point about storing a swap pack at full SOC is another reason I don't see dedicated swapping stations loaded with fully charged packs waiting to go. It would make sense for a call ahead reservation setup as I've suggested, though I would still assume a "full" pack would be at standard charge not range charge, or maybe somewhere in between. I just don't see a good case to have a bunch of fully range charged packs sitting around degrading faster than they need to.

Regarding increasing charge speeds, as I've posted elsewhere I believe pulse charging could substantially increase charge rates but still not faster than filling up with gas.

 

[arg]

Swapping vs Supercharging

One point I haven't seen in these analyses is the question of how much charge people actually take on (or want to take on) at an SC site.

Packs coming out of swapping will obviously be 'full' (or some fixed level like 90% that''s part of the definition of what the service offers). This is both a benefit and a disadvantage:

• With supercharger, filling the last 20% is time-consuming, so the swapper offers an increased advantage if that is what you need to do.
• Swapping has a fixed time penalty, and we are guessing a monetary cost, regardless of how full the pack that you are giving up is. So if your pack was already fairly full, or if you don't need it full when you leave, then the swap option isn't a great advantage over 5 or 10 mins of supercharging.

It is hard to say how much charge people typically want from a Supercharger. For a site in the middle of nowhere, your pack is by definition fairly depleted when you get there, and you need a fair amount of charge to get to anywhere when you leave. But as the network gets built out, quite a lot of the sites are not as remote as that: a lot of the demand will be from people doing trips that are only slightly beyond their available range (or would be in range but they want a bit of comfort margin). These people are adequately served by a few minutes of supercharging and likely wouldn't use the swapper. Likewise, at present a lot of people need to leave the supercharger with close to maximum charge because the next one (or their destination) is a long way away, but as the network builds out then that is less of an issue and more trips can be made without going outside the 10%-75% state of charge where supercharging is most effective.

I'm not sure how to model this demand, but possibly it says that swapping only makes sense in the most remote locations, or perhaps on a few strategic routes.

Maybe in Europe you just need swap stations in Germany where people empty their packs in an hour of flat-out driving?:wink:

Alternative models

All of this assumes that swapping is indeed an alternative to supercharging, (ie. swap regularly for continued high speed driving). There seems to have been less analysis on this thread of the other possible interpretation - that swapping is something done only occasionally to swap in some different kind of pack. Various possibilities here:

• Swapping 85 packs into 60 cars. This would be a calculated risk on cannibalizing 85 sales: do you gain more sales from people who previously thought they needed an 85 but couldn't afford it and now buy a 60 rather than nothing at all, compared to those who downgrade from 85 to 60. This wouldn't kill 85 sales altogether, since having an 85 all the time is much more convenient than having to go to a swap station to get one: those that can afford to just say 'fully loaded' without worrying will still do so, and there are some people who genuinely need an 85 for their daily driving.
• An even higher capacity pack. Possibly one can be built from cells that won't have enough lifetime to make them economic to sell with the car, but to rent one for a couple of weeks could make sense.
• Similarly, a pack supporting faster/deeper charging so that you can make more efficient use of Superchargers on a trip. Could even be simply one of the current packs that software permits you to 'thrash', against a payment that covers refurbishing them regularly.

I don't know which model is more likely - on the one hand, my suggestions for the 'occasional swap' model don't sound very compelling, but on the other hand it feels to me more like there's scope for a single 'aha' moment where you come up with a combination that does make sense. In comparison, the straightforward swapping-to-charge model seems to need incremental improvement on a lot of fronts and offer less scope for a single stroke of innovation.

 

[deonb]

I'm not sure how to model this demand, but possibly it says that swapping only makes sense in the most remote locations, or perhaps on a few strategic routes.

^^^ this. (But I don't know how.)

If we go back to fundamentals, (aka Twitter), we are focusing over and over on the following aspect and debating every possible point of it ad nauseam:
"Elon: There is a way for the Tesla Model S to be recharged throughout the country faster than you could fill a gas tank."

But what if we're asking the wrong question? What if the real question, and the issue that swapping truly means to address, is this?
"Elon: There is a way for the Tesla Model S to be recharged throughout the country faster than you could fill a gas tank."


That is the one thing that is the most lacking in Supercharging. The charge rate as it is actually fine. Incremental improvement and getting it to 10 minutes over 5 years and it's not even a big issue anymore (just double the battery size and you've got your 150 mile charge in 10 minutes). But what it does not and will not ever do is to provide service throughout the country.

I think we're not even touching this aspect is because we're taking it for granted that battery swapping is more expensive and more difficult to deploy than superchargers.

But what if its not?

 

[Yggdrasill]

It's not me saying that, it's Elon... The SC's during the week is expected to be "empty". And on weekends they'll be "sized for peak usage" for Friday's and Sundays (i.e. you would never have to wait for a spot even on those days). So even with your 10 cars for 20 minutes per spot hypothesis (replace 'spot' with 'charger' and I'm with you), it would only be for 2 days per week.

We also know they're meant to be solar positive using 300W max cells. So we don't really have to guess much further than that.

12 parking spots can host 200 solar panels. 200x 300W solar panels can deliver 10x 35kWh charges per day (at most), which is 70 per week. Meaning you have 35 charges on Friday and 35 on Sunday. With 12 parking spots (meaning 3 chargers using the new 4-1 ratio) it means you shouldn't have to wait, but it's actually right on the cusp - not terribly oversized. (Even you used the number of 10 per spot. This is just 10-per-charger. The extra spots are mostly there to support the Solar panels, and for the convenience of not having to "idle" behind someone that's currently charging at a low rate or ran off shopping).

But you can actually only have 10% of the days quota arriving within the same 20 minute window before a wait / slow charge experience will happen.

So whether or not it's a good idea to do things this way vs. cheaper & smaller but as a result having queues on peak times and drawing power from the grid is left up to you to decide (for your own company ). But Elon decided he doesn't want queues during peak times, and he wanted Solar positive. And it all works out (very) favorably financially, so why not.

I wish the battery swapping had numbers this cut-and-dry.

It may be that Elon wants to achieve fully solar-powered superchargers long term, but in the next few years I think it is unrealistic.

Tesla currently has 8 stations, and some 37 SC spots. With 0.5 charges per spot per day and 10 charges per car per year, that works out to a capacity of 675 cars. So, currently the SC network can serve about 6.75% of the Teslas made. That's assuming all of the SC stations are powered by sunlight, which they are not.

 

[deonb]

It may be that Elon wants to achieve fully solar-powered superchargers long term, but in the next few years I think it is unrealistic.

It is known (from the SuperCharger announcement) that the solar installation will lag the charger installation by at least 18 months, with grid storage 6 month after that. Until then it will use grid power. (I'm missing the point of the question?)

Tesla currently has 8 stations, and some 37 SC spots. With 0.5 charges per spot per day and 10 charges per car per year, that works out to a capacity of 675 cars. So, currently the SC network can serve about 6.75% of the Teslas made. That's assuming all of the SC stations are powered by sunlight, which they are not.

We also know that the current system is already over-utilized. (Elon: "Avoid Gilroy"). There is already waiting, and there shouldn't be. And the SC network only covers around 40% of owners as is, and for those owners probably less than 50% of intended routes.

 

[TD1]

I think we're not even touching this aspect is because we're taking it for granted that battery swapping is more expensive and more difficult to deploy than superchargers.

But what if its not?

Thats actually a good point.
For example Elon stated that each super charger uses 12 Converter (the same converter that are build in the Model S) afaik these Converter are quite expensive and there isn't much room to cut the price on those.
So the additional cost vs. SC wouldn't be that much higher.

 

[Palpatine]

Remeber, the wording from Tesla has been pretty clear: A full battery in less time than it takes to fill a tank of gas. I know that the time to fill a tank of gas is debatable. The full battery however is not. It means full. And the way I see it only swapping your battery to another battery, which has been pre-charged, can accomplish that at this point in time.

If they have a way to do that which we have not considered, great. A full battery really quick sounds wonderful.

However if this solution is battery swap, that brings a huge group of negatives to the table. Battery swap had better be free or most people would pass and just plug into the free supercharger.
Even if Battery Swap is free, I would be hesitant to risk an error in the system leaving me stranded. I would rather just recharge for free while I use the bathroom, stretch my legs and grab some food for the next leg of the trip.

 

[deonb]

If they have a way to do that which we have not considered, great. A full battery really quick sounds wonderful.

However if this solution is battery swap, that brings a huge group of negatives to the table. Battery swap had better be free or most people would pass and just plug into the free supercharger.
Even if Battery Swap is free, I would be hesitant to risk an error in the system leaving me stranded. I would rather just recharge for free while I use the bathroom, stretch my legs and grab some food for the next leg of the trip.

From a practical perspective I agree. I don't see the big deal in adding a 20 minute stop halfway through a 4 hour drive.

HOWEVER, it makes it much much easier to sell the car. People don't internalize what it would be like to own an EV and "never have to go to the gas station", and they instead focus on the "it's much slower to charge than a gas station" part.

If you can point at a thing and say: "No way, it's actually faster", you're over a significant sales hurdle.

I have very smart people who work around me. But still, virtually everybody that don't own an EV gets tripped up by that initially.

 

[joefee]

For a 1000mi trip a swap makes sense durning a leg when you can also swap drivers and keep on going with no need for eating, etc. Pick up your charged pack in the same manner on the return trip. A great option IMO:smile:

 

[stopcrazypp]

I have very smart people who work around me. But still, virtually everybody that don't own an EV gets tripped up by that initially.

This is what I've been trying to say too. Battery swapping, like supercharging, is in the marketing budget precisely to address the general public's concerns about charging speed. You can argue until you are blue in the face that the speed of supercharging is "good enough" (which works for people already convinced about the viability of EVs) but you can never say it's faster than filling up a gas tank. And addressing this sticking point opens up a much larger market than any other move Tesla can make.

 

[tdiggity]

Just dreaming things up:

Tesla Supercharger stations area also hyperloop stations. You don't get in a train or a pod or anything like that. You just use your Model S/X or Gen III and the hyperloop takes you to the next station..

 

[kenliles]

but you can never say it's faster than filling up a gas tank. And addressing this sticking point opens up a much larger market than any other move Tesla can make.

on the other side of that trade, it openly admits 'you can never say charging is faster than filling up the gas tank' - and that admission means battery swapping had damn well better suffice, otherwise we share your range anxiety. It's a dangerous marketing game to play and reeks of some desperation that charging will never be the universal answer.

 

[roblab]

I was reading of the Indy Pit Stop Competitions where they fuel a car, AND replace tires, AND wipe the window, AND give the driver refreshment.

The record is something like 8.6 SECONDS.

Since we are comparing apples to kiwi fruit, the fastest charge to the fastest gas fill up, I don't think EVs can compete. Yet. Even Battery Swap can't be done in less than 10 seconds. Or, maybe June 20 will be REALLY world breaking news.