2017 Investor Roundtable:General Discussion

[neroden]

Tesla semi convoy vs freight train: discuss... Also what are metrics for final miles from rail depot to destination vs direct with semi...

Freight train prices are paying for the railroad tracks. The convoy isn't paying, *at all*, for the road paving. If it were paying for the road, it wouldn't even be *close* -- the convoy would cost more than twice as much.

That's the big picture.

 

[neroden]

Further thoughts on Tesla's EV Truck announcement...from a former fleet owner

Thanks for the link!

I have to just pile on with the "most trucks are doing short distance runs" comments, and the fact that that makes the Tesla truck perfect for most use cases. Although the truck will have 500 mile range, the fact is, just like we usually drive way less than the full range of our Model S, most truck drivers will typically be driving half the range of their semi. But they'll be happy to have the "reserve energy", just like we are.

 

[neroden]

So... given what I've gathered so far and the data collection from VIN sightings, here is my latest estimate for Model 3 production.

I think we are seeing the results of fix #1 which is manual production for the zone #1 and zone #2 areas of pack production that Musk referenced during the ER. I'm guessing we are at 75-100 vehicles/week now and therefore the total build for the month of November is somewhere around 400. I am expecting fix #2 to go in and in December, I'm thinking somewhere between 3,000 and 5,000. So for the year, I'm looking at around 4,000 to 6,000 if fix #2 works.

We probably won't get verification that fix #2 works until some time in December. Likely smart money will figure it out before we do, or it might be the moment fix #2 works, they open the configurator for non-employee orders.

If fix #2 does not work, then we stuck at the much lower manual build rate. Maybe another 400 to 500 in December. That won't be good, even if the Semi and Roadster reveals brings in a slew of deposits. The cash situation would be a significant concern then.

Agreed in all regards, but:

-- I'm personally confident that fix #2 will work: I see no inherent reasons why that stage of the battery pack production line should be particularly difficult to automate (as opposed to, say, wiring harnesses being difficult to automate, which is a more fundamental issue).

-- I am, however, less confident about *when* fix #2 will be implemented. December seems potentially optimistic to me. It might not be up and running until January, even late January.

Do you have a particular reasoning for why you expect fix #2 to be live in December? My speculation is based on vague hunches about lead time for both hardware and software.

 

[oneday]

Freight train prices are paying for the railroad tracks. The convoy isn't paying, *at all*, for the road paving. If it were paying for the road, it wouldn't even be *close* -- the convoy would cost more than twice as much.

That's the big picture.

I don’t think that’s exactly true, at least if I’m understanding you correctly. I’m assuming you’re talking about the fact that electric semi’s won’t be paying gas tax?

That is certainly something that will need to be addressed, eventually... However, I don’t think it is that large of a financial burden even if electric semi’s pay their fair share if you look at the numbers.

Based on some napkin math, federal diesel excise tax is $0.24 a gallon. State diesel tax ranges from $0.08 to $0.50 gallon. With the average being about $0.26 per gallon of diesel. Semi’s get roughly 6 miles per gallon, so $0.50 gallon / 6 Miles per gallon = $0.0834 per mile that diesel trucks pay via diesel excise tax to cover road maintenance.

You add that to the Tesla Semi’s operating cost per mile of $1.26 per mile and you get $1.34 per mile which is still a lot better than industry cost of $1.51 per mile.

Now that doesn’t include all the economical efficiencies of platooning. Add in platooning AND electric semi’s paying their fair share in road taxes and you still could potentially beat rail.

 

[adiggs]

*Cough*

If you'll note, they aren't even trying for the "sleeper cab" / "tandem driver" market. Musk didn't mention it, because that's the one market where the semi doesn't charge fast enough, and it would also require a different cab design. That's the very top high-end of the semi market. However, it's also a very small part of the market, and fundamentally not important.

I think you have a false equivalency here @neroden. The tandem driver market and the sleeper cab market clearly do overlap, but they are not synonymous. One purpose of a sleeper cab of course is it enables the other driver to sleep while the truck is still moving.

It's other purpose, and what occupies the big hunk of the market is single driver / single truck. On long hauls, when it's break and / or sleep time, the truck can pull off at rest stop / truck stop, crawl in back, and sleep / watch TV (or so I hear), etc.. Carry your motel room with you wherever you go.

I'm thinking that if they're going to be serious about distance hauling, then there's going to be a sleeper cab variant.

 

[Waiting4M3]

*Cough*

If you'll note, they aren't even trying for the "sleeper cab" / "tandem driver" market. Musk didn't mention it, because that's the one market where the semi doesn't charge fast enough, and it would also require a different cab design. That's the very top high-end of the semi market. However, it's also a very small part of the market, and fundamentally not important.

I think one case where sleeper cab is less necessary is when EAP+convoy driving is permitted, then put 2 drivers in 2 trucks, one truck leads, its driver drives, the other truck follows autonomously, and its driver sleeps/relaxes. Every 8-10 hours or so, stop to charge the trucks, and swap places. So instead of 2 person driving 1 truck 24 hours a day, they can drive 2 trucks 24 hours a day.

 

[adiggs]

I don’t think that’s exactly true, at least if I’m understanding you correctly. I’m assuming you’re talking about the fact that electric semi’s won’t be paying gas tax?

That is certainly something that will need to be addressed, eventually... However, I don’t think it is that large of a financial burden even if electric semi’s pay their fair share if you look at the numbers.

Based on some napkin math, federal diesel excise tax is $0.24 a gallon. State diesel tax ranges from $0.08 to $0.50 gallon. With the average being about $0.26 per gallon of diesel. Semi’s get roughly 6 miles per gallon, so $0.50 gallon / 6 Miles per gallon = $0.0834 per mile that diesel trucks pay via diesel excise tax to cover road maintenance.

You add that to the Tesla Semi’s operating cost per mile of $1.26 per mile and you get $1.34 per mile which is still a lot better than industry cost of $1.51 per mile.

Now that doesn’t include all the economical efficiencies of platooning. Add in platooning AND electric semi’s paying their fair share in road taxes and you still could potentially beat rail.

I believe @neroden's point is that even with gas taxes, trucking is being subsidized by tax payers in the form of road maintenance - the gas taxes aren't even vaguely adequate to fully pay for the road damage and repair required for trucks.

By comparison, the trains own their tracks and pay for all their own maintenance.

The Hidden Trucking Industry Subsidy

I couldn't find an article talking about current railroad track maintenance subsidies - there's been a tax credit in previous years that has since expired used to help small / regional railroads pay for track maintenance. My brief research tells me the full maintenance cost is carried by the railroad (and therefore all of us as direct customers of the stuff the railroads are moving). As taxpayers though, we're off the hook and the correct market signal about teh cost of moving stuff by rail is being provided.

 

[NicoV]

Freight train prices are paying for the railroad tracks. The convoy isn't paying, *at all*, for the road paving. If it were paying for the road, it wouldn't even be *close* -- the convoy would cost more than twice as much.

That's the big picture.

In the USA.
Most European countries have some kind of pay per use highway fee.

EDIT: Forgot to mention that in Europe (at least in my country) trains are (very heavily) subsidized.

 

[GoTslaGo]

I recall someone had posted some pictures of the back of the Semi cab (where things hook up to the trailer). His/her comment was that there was "so much" space... implying there was wasted space in the back. Maybe allowing for a future sleeper cab variant?

 

[SwTslaGrl]

^One of the press kit images shows an extra folded seat in the back,

 

[erthquake]

To be absolutely clear about this, the electric bus companies did pick up the baton, which is why I think Tesla is not even trying to make a bus. They're only going to go into the markets nobody else is going into.

Musk originally planned to not invest in electric cars at all -- he saw the EV-1 and first generation RAV4-EV and figured GM and Toyota would do it! It's only *after* they "killed the electric car" that he invested. He's *only* going to invest in sectors where he has *evidence* that the other guys *aren't* going to invest seriously. Incredibly smart, actually.

Plus many (most?) bus lines are run or subsidized by municipalities, and they're a much smaller component of total miles driven compared to trucks. The profit motive to reduce costs is much stronger with trucking. Tesla Semi has a much more lucrative addressable market than a bus does.

 

[RobStark]

Marques Brownlee, tech youtubber with 5.3M subscribers, discussing Roadster 2.0. Ends by saying his lease on his Model S P100D ends late 2019.

And is probably getting a Roadster.

 

[JBRR]

Marques Brownlee, tech youtubber with 5.3M subscribers, discussing Roadster 2.0. Ends by saying his lease on his Model S P100D ends late 2019.

And is probably getting a Roadster.

And did.
Marques Brownlee on Twitter

 

[CorneliusXX]

Batteries are a generic product. I envision a company (companies) with expertise in the field will build gigafactory-equivalents and sell the products to a wide range of auto companies. That makes more sense than each auto company developing the necessary skills and building its own battery factory.

I think we are not too far away from automotive batteries becoming a generic product. At the moment improvements in battery density and performance lead to real world advantages in terms of cost and performance. However once they become sufficiently powerful to not provide significant real improvement to the vehicle then cost will become the main driver. Would anyone really care if you had to recharge after 1k miles instead of 2k miles?

There is obviously other areas where battery performance gains will have more of an effect, e.g. planes.

 

[LoL Rick]

I am wondering if the Roadster that was shown already had the 200kWh pack. Or if this is progress they will realise by 2020. After all, current tech caps the maximum they can stuff into the much larger Model 3 at 75kWh.

I think there is a difference between current production tech and current prototype tech.

 

[DragonWatch]

Interesting read for Bulls, Bears, well not so much.

Tesla's biggest bull sees $60 billion in revenue by early next decade

There is another contribution by Lutz at the bottom of the article touting that there is no secret sauce in Tesla. My take is that Lutz is striving for a political appointment like his Exxon buddy. I have not reviewed the YouTube rant because I would wake up my wife. Poking hot pokers at sleeping bears is not a habit I find rewarding Besides, watching and listening to Lutz would dry up any sweet dream.

 

[Bobfitz1]

I am wondering if the Roadster that was shown already had the 200kWh pack. Or if this is progress they will realise by 2020. After all, current tech caps the maximum they can stuff into the much larger Model 3 at 75kWh.

To output the current needed to power 3 motors pushing the car to 60 in < 2 seconds, repeatedly, strongly suggests they have either 150 or 200 KWh now. Since Roadster II is built on the M3 platform, my guess is they started with an M3 power pack and modded it to fit in double the number of cells. That configuration may be refined and become the production car pack, but using the next iteration of cells under development or possibly the one after that. This is purely speculation, as the history of Li ion cell improvements has been more like 7 - 10 % increase in energy density with each succeeding iteration. Nobody outside the Tesla Li ion cell 'Skunkworks' can be sure what they may have under development that could give a 30% increase by 2019/2020 when Roadster should be ready.

 

[JRP3]

I think there is a difference between current production tech and current prototype tech.

I'm almost positive that Roadster prototype did not have a 200kWh pack, I'd guess maybe they squeezed in half that.

 

[JRP3]

To output the current needed to power 3 motors pushing the car to 60 in < 2 seconds, repeatedly, strongly suggests they have either 150 or 200 KWh now.

Not if they didn't care how long the pack lasted, or if they used higher C rate cells.

 

[NicoV]

Interesting read for Bulls, Bears, well not so much.

Tesla's biggest bull sees $60 billion in revenue by early next decade

There is another contribution by Lutz at the bottom of the article touting that there is no secret sauce in Tesla. My take is that Lutz is striving for a political appointment like his Exxon buddy. I have not reviewed the YouTube rant because I would wake up my wife. Poking hot pokers at sleeping bears is not a habit I find rewarding Besides, watching and listening to Lutz would dry up any sweet dream.

Romit Shah:
“They are going to sell as many as they can make for a very long time” , “stock price well in excess of $500”.

Bob Lutz:
Classic denial.