Supercharger congestion - a modest proposal

[dgpcolorado]

So with about 70,000 MS's sold in the US through Q1 2016, some SCs are already full. Let's say MS + MX double that through 2017. Then let's say M3 double's that in 2018. Two and a half years from now there are 4x as many potential SC users as today. If the current utilization model is representative of the future, then Tesla has a couple of years to quadruple the number of SCs that it has deployed in the last four years.

I just don't see how the current model will work without some reasonable rationing (in the economic use of the word) to support the original and ongoing intent of the SC network.

Not so, that's not how queuing theory works. At the present time a huge majority of Supercharger Stations can be considered "underutilized". (For example, here in Colorado and Utah I have yet to see a station with more than two cars, one of which was mine.) The total population of Tesla cars could increase substantially and they would need little, if any, expansion.

This means that once the general Supercharger network along long distance travel routes is built out, the continued building or expansion of stations can be focused on areas that are seeing congestion. Therefore, a quadrupling of cars does not need a quadrupling of Superchargers. There are models that can be used to predict the most efficient way to expand the network, but it certainly won't need to be increased at the same rate as sales of cars.

 

[calisnow]

So with about 70,000 MS's sold in the US through Q1 2016, some SCs are already full. Let's say MS + MX double that through 2017. Then let's say M3 double's that in 2018. Two and a half years from now there are 4x as many potential SC users as today. If the current utilization model is representative of the future, then Tesla has a couple of years to quadruple the number of SCs that it has deployed in the last four years.

I just don't see how the current model will work without some reasonable rationing (in the economic use of the word) to support the original and ongoing intent of the SC network.

FWIW, in your example I don't think Tesla would need to quadruple the number of superchargers overall because many superchargers don't see much use. However some of them they may need quadruple (or even quintuple) - I was in Rancho Cucamonga last night around 6:00 pm - 13 of the 14 stalls were occupied (including me). In that particular location there appears to be plenty of room to quadruple the spots if they need to.

 

[Boatguy]

FWIW, in your example I don't think Tesla would need to quadruple the number of superchargers overall because many superchargers don't see much use. However some of them they may need quadruple (or even quintuple) - I was in Rancho Cucamonga last night around 6:00 pm - 13 of the 14 stalls were occupied (including me). In that particular location there appears to be plenty of room to quadruple the spots if they need to.

Point taken. Perhaps it's 4x the number of stalls as opposed to SC sites.

 

[Boatguy]

Not so, that's not how queuing theory works. At the present time a huge majority of Supercharger Stations can be considered "underutilized". (For example, here in Colorado and Utah I have yet to see a station with more than two cars, one of which was mine.) The total population of Tesla cars could increase substantially and they would need little, if any, expansion.

This means that once the general Supercharger network along long distance travel routes is built out, the continued building or expansion of stations can be focused on areas that are seeing congestion. Therefore, a quadrupling of cars does not need a quadrupling of Superchargers. There are models that can be used to predict the most efficient way to expand the network, but it certainly won't need to be increased at the same rate as sales of cars.

I think it's reasonable to assume that with the MX and then the M3 the distribution of owners and driving will be different than today. Could you please expand upon the model that should be applied and what it tells us about the necessary expansion of the network?

And the thinking which dismisses the assertion raised by the OP and myself that some rationing system will need to be applied? It's certainly an assertion on which I would like to be horribly wrong. I don't look forward to seeing SC's look like gas station queues of the 70's.

 

[David_Cary]

As someone who has SCed 6 times in a my first year of ownership, I couldn't give a crap with whatever cost model they went with as long as I didn't have a line. You could charge me $1 a kwh (and I pay $.05 at home) and I would be okay with that. And I paid "$2000" when I bought my car for the privilege - no I didn't.

I do think a very reasonable scenario is $.50 a minute in your 50 or 100 mile home radius. Of note, I have 3 SCs about 50 miles a way in 3 directions and I've hit all of them.

One of my SCs (120 miles from home near the in laws) has a free Chademo about a mile away. So that adds a little computation. The Chademo has much better food options for a sitdown.

All the Cali people need to come over here and see how many free options we have and life goes on - including my free Chademo 3 miles from home that I've never been blocked from (but it has been down). I know a Leaf owner who didn't install a 240V at home because this was nearby.

4 of 5 Tesla owners I know off the top of my head - make more than $250 an hour. Seems to me that would be expected given the cost.

 

[henderrj]

May I add another wrinkle? We have a patent pending on a device that will allow vehicles to plug themselves into a charge port. Maybe more to the point, unplug themselves. With summon already in place it would be a pretty simple task to set some stalls apart as "no human zones" - you pull into a queue and get out (after touching the "connect" button to confirm). Then the car would pull up to the first available stall and, after charging, move to a parking area where you would pick it up (a pull through spot seems advised). Since it seems clear the 3 will have an even better autonomous driving suite this might make a considerable difference in congestion. Full autonomy changes the scenario even more.

Thoughts?

 

[Saghost]

May I add another wrinkle? We have a patent pending on a device that will allow vehicles to plug themselves into a charge port. Maybe more to the point, unplug themselves. With summon already in place it would be a pretty simple task to set some stalls apart as "no human zones" - you pull into a queue and get out (after touching the "connect" button to confirm). Then the car would pull up to the first available stall and, after charging, move to a parking area where you would pick it up (a pull through spot seems advised). Since it seems clear the 3 will have an even better autonomous driving suite this might make a considerable difference in congestion. Full autonomy changes the scenario even more.

Thoughts?

Tesla clearly has something like this in mind in the next couple years - Elon is promising they could make coast to coast summon work in two years, which inevitably involves multiple automated supercharging sessions. They've demonstrated the Snake, so I'm assuming they'll start rolling some sites like that out sometime next year - maybe one or two stalls converted at an existing site? Maybe all of the "pull in" stalls, since the demo seemed more suited to that geometry than the more typical back in siting?

 

[dgpcolorado]

I think it's reasonable to assume that with the MX and then the M3 the distribution of owners and driving will be different than today. Could you please expand upon the model that should be applied and what it tells us about the necessary expansion of the network?

And the thinking which dismisses the assertion raised by the OP and myself that some rationing system will need to be applied? It's certainly an assertion on which I would like to be horribly wrong. I don't look forward to seeing SC's look like gas station queues of the 70's.

Although it is vigorously disputed by some here, I won't argue against the idea that some sort of rationing system might be more efficient than trying to build enough Supercharger stations to accommodate peak usage in crowded locations for brief periods of time (holiday weekends, for example).

Nevertheless, your suggestion of a 4X increase in cars needing a 4X increase in Superchargers is something of an overestimate. Queuing Theory refers more to the efficient utilization of a limited resource and is a bit technical. But I tried to give an example of why the number of Superchargers does not have to match the increase in cars in a linear fashion (the Superchargers in my state are almost never crowded; although they serve a vital purpose for travel, they simply don't need to be increased). There are ways to model this problem that are well beyond my expertise. But I can give an example with some made up numbers:

Suppose that 90% of the Superchargers currently in place are rarely used (but serve a necessary purpose in allowing long distance travel through remote areas). The vast majority of those locations would still suffice even after a 4X increase in cars because they would still rarely be full. That means that the increase in Superchargers and station locations could be focused on expanding the 10% that are heavily used. To keep those locations from being too crowded they will likely be need to be expanded by considerably more than 4X, since a few locations are already filling up at times with the current fleet of Tesla cars.

So, suppose that there are currently 1800 Superchargers in North America (it is actually a bit more than that)
90% underutilized would be 1620
10% overutilized would be 180

If there was a 4X increase in Tesla cars the 90% underutilized Superchargers would still be adequate and need no increase but suppose an increase of 15X for the fully or overutilized areas. The total after that expansion would be

1620 + (180 x 15) = 1620 + 2700 = 4320 Superchargers after expansion

By contrast, a linear increase in Superchargers would be 1800 x 4 = 7200

Note that despite a drastic increase (15X!) in Superchargers in highly populated areas, the total is nowhere near a 4X increase overall.

Of course this is an oversimplification, but I hope that the general idea is clear.

 

[Boatguy]

Well done! So if the assumptions are all correct, then in the next 30 months Tesla needs to add 4,320 - 1,800 = 2,520 new stalls, or roughly 2.4x what is installed today.

If they were installing linearly over time (a questionable assumption on my part), they installed 1,800 stalls over four years or about 38 stalls / month. Now they will need to install over 80 stalls/month for the next 30 months to meet MX M3 demand based on dgp's model; more than twice as fast.

That's a lot of new stalls and as I said earlier, backhoes don't scale like semiconductors. Some form of rationing seems prudent.

 

[Electroman]

When I get my Model 3, I'd like to know that there will be a free slot at a supercharger when I need one for a long journey.

Here's a modest proposal to make sure superchargers are free for those doing long distance travel.

You can't charge up to more than 2x the distance between the supercharger and your home.

You can pay to charge more than that, but it's no cheaper than charging at home.

I like it. I love it. Let's implement this.

Let the free loading suckers go to hell.

 

[EVie'sDad]

It's not free for those who purchase the use of the SCs, nor is it free for the facility building it, or for Tesla arranging charging rights with local businesses, nor for those paying taxes that go to help support this infant industry.

But I understand what you meant, yes free can confuse economics, but it really isn't free. It's just the perception it is free because no transaction is occurring at the time of recharging at a SC, but you too must acknowledge it is anything but free.

And with less wealth and resources at hand, I see more Model 3 owners taking advantage of the SC's than Model X or even S owners, except as intended for long distance trips, which I take full advantage of. Wanting to travel and seeing the countryside in my MS is one of the joys the Tesla SC network allows me to do in comfort of my BEV. I hope many others get to have the same experience and see firsthand the awe and wonders clean sustainable transportation has to offer whether they be in Model S, X or 3's.

 

[Boatguy]

Sunday morning I stopped at the Petaluma, CA supercharger to give my wife a brief tutorial on using a supercharger to find just one empty slot. We charged for about 10 minutes so she could get the idea, then moved on. I hate to think what it looked like on Sunday afternoon as everyone would have been heading home to the south from points north.

Several proposals have been made to ration superchargers and ensure their availability for long distance drivers. Tesla needs to decide which one to implement because I don't think it's physically possible to build enough new Superchargers between now and when the first, 100,000 M3s hit the road.

 

[Drucifer]

I would go with home plus 15% SOC buffer. Charging to 12 when you live 6 still too close to the edge.