-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
I did my tour of duty with S/34 and S/36. At UIUC I did print queue maintenance and toyed with BITNET Relay on their bigger brothers, but no further than that. But now we're headed too far off topic.
rxlawdude
Active Member
Oh, do I remember pricy and hard to find MCA peripherals. Fortunately, the PS/2 reign was rather limited.
snort
Member
functionally, that's just what Musk is proposing. I think he'd prefer help on the buildout of compatible superchargers with other manufacturers names on them; this would be a clear message of support for the technology. but money to help build them would be good, and in effect advertising for Tesla. The politics of this is dicey.
Trouble is, the other manufacturers are trying to compete at the low end. On the scale of a company like Nissan, $15 or 30M to build 100 superchargers is small potatoes, but if it succeeds, there's an implicit commitment to build more at a rate commensurate with their production rate. there is presently about 1 supercharger for every 250 model S's in the US. If they sell 10M cars, which is entirely plausible, they'd have to build 40,000 to maintain that level...that's $6B and is real money. to do that, they'd have to add at least $1K to the cost of every vehicle just for construction. Maintenance and the cost of electricity would add more. If Elon continues to insist that other players front load the cost, I don't see how anybody but Ferrari and Aston Martin and Jaguar and their ilk can step up. I don't think small battery EVs use DCFC at the rate that big battery EVs do but it's still a lot to add to the cost of a cheap car. thus my enthusiasm for adapters and pay as you go.
you are exactly right, my bad. I had a different conversation and a different M on my mind..
I used both 286 and 486 based PS/2s. What they did, they did very well and they did a good job of flying the IBM banner. (unlike the PCjr and some others.)
snort
Member
why is that?
Small battery EVs won't be doing as many long trips on average as the average Tesla, but if the Gigafactory works, there will be a lot more big battery vehicles, not all of them Teslas. I can't think of any reason they would not do the same number of long trips on average. I'm still new to this but I've never yet been alone in a supercharger station. If my small sample is representative (is it?), tripling the demand per station would have me waiting more than half the time. that would reduce the per-vehicle construction cost overhead by that much, but not the power or maintenance bills.
there are about 1100 car per gas station in america. that's a much poorer ratio than we have for SCs and most Teslas use home or destination charging for the majority of fillups. Most of the time when filling up in urban or suburban areas, I had to wait in line; almost never in rural areas. I think I'd be P.O.ed if I had to wait for as many cars at a SC as I almost always did at Costco....
Queuing theory.
The conventional way to plan this sort of thing is to say what service level you aim to provide (expressed as the probability that someone turns up at a location and finds it busy). With customers arriving at random you can then calculate how much traffic a given number of channels (supercharger stalls) can take.
For sites with lots of stalls, double the number of stalls gives you double the capacity. But for small numbers of stalls it doesn't work like that - two stalls gives about 15 times the capacity of a single stall, and 4 stalls gives about 6 times the capacity of 2 stalls (assuming 1% acceptable chance of waiting, and assuming all the stalls are equal - actual superchargers are more complex to calculate but the principle is the same).
So, for sites with not much traffic Tesla has already had to provide quite a lot of superchargers in relation to the actual traffic - and doubling up on what they have now will give them more than double today's capacity, while doubling the sites in already dense areas merely gives double the capacity.
vitaliy
Member
No, your experience is not typical of the entire network. You are in a fairly high density of Teslas.
My experience is also not typical of the entire network, but in my case, I rarely have any company and have never seen more than one other Tesla at a SC except on our trip to California. And that only happened in California.
Superchargers needed to be placed in less Tesla dense areas to allow interstate travel. Many of these have not yet become saturated.
Yes, supercharger density in these areas will also need to increase eventually. But not as much as those areas that are at, or near saturation.
paco3791
TMC OG
Not to mention that Tesla could make their SCer protocol just a little bit smarter and greatly increase the capacity of the over all system. They know where every car is already, they know how many cars are using any particular SCer and what their SOC is already, they could easily throw up a rerouting suggestion to take you to a less congested SCer. Even better, if everyone is more or less using the on board trip routing software the system can optimize the routes so everyone gets a spot when they need one. It's not a problem yet, but if/when it is just building more SCers won't be the only part of the solution.
snort
Member
This is true if there are enough superchargers around that alternates are available. Here in the Pacific Northwet, they're still about 150 miles apart. Skipping one is not an option. I'm looking forward to the day that they're as convenient to get to as gas stations. (and I totally agree: do intercity first, before they build out where a lot of people will use home charging)
Lest there be any question that Tesla knows exactly who is using what supercharger and could trivially change to a pay as you go billing model for some users if they chose to:
--Snortybartfast
- - - Updated - - -
On the subject of queueing theory: increasing the number of berths increases the odds of not having to wait mostly if the number of berths is small, given a constant average demand. I think it's O(ln(n))... the second berth quadruples availability. the third a little more than doubles (9/4). the difference between 7 and 8 is relatively small.
the point is that demand with a small number of vehicles is inherently very bursty. with larger numbers, the flow will be higher and less bursty. it'll be easier to match the number of berths to demand but we'll be running a lot closer to capacity. my hope is that Tesla and whoever else is building SCs at that point thinks that almost always having a berth available is of high enough value to make it happen. The risk that they won't resembles the crowding out fear that Johan expresses.
--Snortybartfast
It was unusual for us to see no other Teslas at the SCs on our trip. And this included OK, KS, WY, and MT. I don't believe any of those states have a particularly high Telsa density.
Similar threads
- Question
