Are the SCs actually able to be self-reliant?

[DEinspanjer]

Was reading comments on Hacker News about the announcement, and one person laid out several facts about potential solar energy. It was a compelling argument, and I'm wondering if there was anything in particular left out or gotten wrong that might change that story.

Well... let's do the math!Incident solar radiation is about 1 kW per meter squar... | Hacker News

Well... let's do the math!
Incident solar radiation is about 1 kW per meter squared. Assume 10% end to end efficiency (including solar -> electric, storage, and actual charging efficiency... this is optimistic, but doable... certainly not the cheapest though). You now need 10 square meters of panel to supply 1 kW. Assume sun is out for 12 hours a day, and for the full 12 hours you get all 1 kW/m^2 incident radiation (this is very very generous). The car has a 60kWh battery. So if we want to find out how much panel it takes for one station to charge one car a day...
60kWh to charge, but 12 hours to charge it up, so you basically need 5kW, which is 50 m^2 of panel. Which honestly isn't -THAT- big. Its a 7 meter square. Now... from my random snooping around, and my own impressions, I would say that a 'large' gas station typically covers something like... 500-600 meters squares (thats 5300 to 6500 square feet). That would be enough to super optimistically charge 10-12 cars a day.
I saw optimistic, because incident solar radiation is not that kind. Can take a looksey here: http://rredc.nrel.gov/solar/old_data/nsrdb/1961-1990/redbook...
If you pick average, annual, and single axis east-west tracking (which I believe is the best bang for the buck), you see on the solar map, the BEST places in the US gets 5-6 kWh/m^2/day, compared to our assumption of 12 kWh/m^2/day.
The feasibility of superchargers being even 90% off the grid is of course completely dependent on the ratio of size of supercharger panel array and rate of cars charging. But given the above, I'm inclined to believe that it is highly impractical for the majority of super charger stations to become actually self reliant, especially in high traffic areas.

 

[deonb]

The math seems fair.

However, also keep in mind that Tesla intends to install 200 SuperCharger locations for $60 million. That means it's funded by 30'000 cars (paying $2000 each for access). I think Tesla will take some profit on it, so let's say the planned rollout is intended to support 75'000 cars (3 years of Model S's).

So on average that makes it 375 cars per SuperCharger location. And it handles 12 cars per day while being self-reliant. Meaning each car can get a charge once per 31 days - so 12 charges per year. Basically a trip from Seattle to San Francisco and back.

It probably works out close to break-even on average, but not by any kind of large margin.

 

[scaesare]

Hmmm, well for starters, I think his assumptions about panel efficiency are off by at least a factor of 2.

I did some quick calcs in THIS POST. While I believe it will be possible to make the SC's reliant on solar energy... I don't think it will be possible to have panels co-located on the site in all instances.

Now some of my assumptions regarding usage of the sites might well be off also... but I still think it will work with arrays tied in to the grid elsewhere. As a matter of fact I once read somewhere that was part of the plan.. but I'll be darned if I can find the reference again...

 

[deonb]

Hah!

Our Solar Panel calculations are off by a factor of two:

26:38 minutes into the SuperCharger announcement:
Elon: "The main thing we need to do is make sure we got a lot of parking spaces available. So what the superchargers are actually capable of doing is routing power to multiple parking bays. Right now it can route power to 2 parking bays, but we're going to be upgrading that, so that it can actually route power to up to 4 parking bays. That's really just to expand the number of parking bays where you can do SuperCharging and make sure there is always an excess number of stations, except it extremely rare circumstances."

So they're doubling the number of parking bays for the same amount of SuperChargers. Interesting... it doesn't increase throughput, but it does help with the charge-and-run problem.

More importantly - it allows Tesla to double the size of the solar canopy. That makes the math a lot simpler.

I'm just not sure if the financials work out. Can you install 13'000 square feet of panels for $150k?

--- updated ---

Oh, yeah, looks like that should be simple.

At 13'000 feet (which covers 40 bays / 10 chargers) you can install 650 Sharp ND-250 QCS panels. Amazon sells those for $223 each. I'm sure if you place an order for 130'000 of them you probably will get a discount, but you also have to put it all up on a structure. So I'll leave it at $223 each. That works out to be around $145k, which is fine.

And those 650 panels will give out 162.5 kW, let's say for 5 hours a day = 812 kWh/day or 5.68 MWH/week. Enough to do a 40kWh charge on 142 cars over a week. (And like Elon said - these show up mostly on Friday's and Sunday's).

That also seems to work out. They're gonna need bigger stationary batteries though...

 

[jeff_adams]

In Tesla's patent filings on grid storage (which is what Elon says they will be doing at SC sites. They already have two grids running now, most likey at both HQs) Tesla talks about a plan to sell power back to local power companies at peak rates. So they can sell less power than they buy but possibly make a profit on the exchange rate. Pretty clever idea

 

[ItsNotAboutTheMoney]

In Tesla's patent filings on grid storage (which is what Elon says they will be doing at SC sites. They already have two grids running now, most likey at both HQs) Tesla talks about a plan to sell power back to local power companies at peak rates. So they can sell less power than they buy but possibly make a profit on the exchange rate. Pretty clever idea

Very important. They'll want to make them self-reliant in terms of energy use, but more importantly they can use arbitrage to make them financially self-sufficient. The bonus is that if they get their battery costs down enough the grid storage value will have utilities clamoring for them to add capacity and being very helpful when it comes to setting up their Superchargers.

Mr Straubel wasn't kidding when he said that the first question they ask of cell manufscturers is how they're going to lower costs.

 

[CapitalistOppressor]

Hah!

Our Solar Panel calculations are off by a factor of two:

26:38 minutes into the SuperCharger announcement:
Elon: "The main thing we need to do is make sure we got a lot of parking spaces available. So what the superchargers are actually capable of doing is routing power to multiple parking bays. Right now it can route power to 2 parking bays, but we're going to be upgrading that, so that it can actually route power to up to 4 parking bays. That's really just to expand the number of parking bays where you can do SuperCharging and make sure there is always an excess number of stations, except it extremely rare circumstances."

So they're doubling the number of parking bays for the same amount of SuperChargers. Interesting... it doesn't increase throughput, but it does help with the charge-and-run problem.

More importantly - it allows Tesla to double the size of the solar canopy. That makes the math a lot simpler.

I'm just not sure if the financials work out. Can you install 13'000 square feet of panels for $150k?

--- updated ---

Oh, yeah, looks like that should be simple.

At 13'000 feet (which covers 40 bays / 10 chargers) you can install 650 Sharp ND-250 QCS panels. Amazon sells those for $223 each. I'm sure if you place an order for 130'000 of them you probably will get a discount, but you also have to put it all up on a structure. So I'll leave it at $223 each. That works out to be around $145k, which is fine.

And those 650 panels will give out 162.5 kW, let's say for 5 hours a day = 812 kWh/day or 5.68 MWH/week. Enough to do a 40kWh charge on 142 cars over a week. (And like Elon said - these show up mostly on Friday's and Sunday's).

That also seems to work out. They're gonna need bigger stationary batteries though...

The large numbers of parking stalls is reduced if they also implement battery swaps. Regardless, I imagine Tesla will just paper solar cells over every available surface in sight until they have enough solar capacity to make it all work. And I am morally certain that Tesla intends to continue to scale up their CES capacity to ensure they are able to support charging requirements during peak usage, while also being able to make the company gobs of money in the CES role.

 

[deonb]

The large numbers of parking stalls is reduced if they also implement battery swaps. Regardless, I imagine Tesla will just paper solar cells over every available surface in sight until they have enough solar capacity to make it all work. And I am morally certain that Tesla intends to continue to scale up their CES capacity to ensure they are able to support charging requirements during peak usage, while also being able to make the company gobs of money in the CES role.

Yeah, this is it. Elon pretty much came out and said so in so many words. From Brianman's transcript:

"Plan is for almost all SC stations to have solar power... solar carport over all the parking bays. Most of the time those bays are mostly empty. They fill up on a Friday afternoon/evening and on a Sunday. We try to size the SC bays -- connectors and hardware -- for peak usage. Intent is to generate more energy from solar power than the cars use doing the course of the year. Net positive throughout the year. Some SCs are very far north. For those the plan is to buy sustainable power - hydro or geothermal."

It seems like that they're not just going for a aggregate net positive, but an individual net positive. And the moment one runs negative, it means it's undersized and needs to get expanded. (I wonder if it means by nature SuperCharger stations in Seattle will be twice as 'oversized' as those in Nevada).

To me it seems like it will be sized so that there is almost always an open parking bay available. Not only to have good customer satisfaction, but also just to size the system in such a way that there are enough solar panels.

Yet another piece of useless math:

Let's say Tesla adds 1 300W solar panel per car in the system. Over a year's period, it means that it generates 547 kWh of electricity per car, enough for > 1657 miles traveled on Super-Charges at 330wh/mi. However, those trips will be charged at homes + stopovers as well, so add let's say 25% (1 stop in 4 is at home/hotel). That leaves 2071 supercharger miles/year. Let's just say 2000.

And since > 100 mile trips (SC trips) is about 15% of all miles traveled, this is sufficient for cars getting over 13'333 miles per year - a bit more than average.

For 20'000 vehicles/year, this means we need to expand the SuperCharger network by 20k panels per year, or about 100 SuperCharger stations/year with 200 panels each. Hey, what do you know.

The math all works out, it just calls for a heck of a lot of SuperChargers - or at least parking bays. I'm all for that.

And with a fleet of 5 million vehicles (500k over 10 years), it calls for 5 million panels, or about 25'000 SuperCharger stations (worldwide). It would cost $7.5b to build, but 10 million vehicles would have also paid $20b in SuperCharger access fees upfront.

So get ready for a future where you'll see a SuperCharger almost everywhere, and that you almost never have to wait for. I'm sure at some point Tesla is going to start putting in shared SuperChargers (just to be able to get more rent-free parking bays in more places), where most of the parking bays are only reserved for Teslas at certain hours of the week - the rest of the time it's an ordinary parking bay with a cool carport.