2017 Investor Roundtable:General Discussion

[vgrinshpun]

Cu conductivity is ~58% higher than aluminum, so Al cable needs to have 58% larger cross section as Cu cable to achieve the same resistance. If the cabe cross section is round, then the diameter of the cable would be 25% larger.

Given the same resistance, heat generated in the cable would be the same, so insulation thickness won't need to be changed much, my guess.

To clarify, thickness of insulation is driven by voltage level, not heat generated in the cable. The insulation, however, is the limiting factor as far as current carrying capacity is concerned because it has thermal rating that can't be exceeded without an impact on the life expectancy.

 

[jbcarioca]

I'd expect only the higher power cabling to be aluminum. Smaller diameter 12V wiring in aluminum might be subjected to more vibrational stress fatigue and not worth the hassle of extra support and termination methods. Copper is easy.

We also probably should consider that pure aluminum all else being equal, is not only subject to less fatigue resistance than is copper, but also subject to corrosion losses. Presumably a state-of-the-art aluminium cable maker is using an alloy of some sort to reduce those two negatives and probably has some quite refined cladding to seal off much of the negative risks. Fasteners must be pretty special too, else those annual connection cleanings that nobody wants to have. Just as body, chassis and engine components have managed to deliver some excellent benefits including corrosion resistance, so the cables are doing it. The increased size really does not mean much when coupling lower weight and lower cost. I'm curious to know how they deal with vibration resistance.

 

[Drax7]

My view is that the only economically sustainable health care in the USA is "Medicare for all". To afford health care, we need to cut out the middlemen and gain control of the cost of drugs, hospitals and practitioners. Medicare for all would result in the end of "for profit health care". You might consider shorting a weaker health insurer, pharmaceutical or hospital company. The final collapse of these companies is probably several years away. The inability of republicans to end and replace the affordable care act is symptomatic of the unsustainable cost structure of our current system.

Health care will be the next big short. However it's not a highly leveraged industry from
What I can tell, so the bubble won't burst in similar fashion, or be systemic to the point
Of destroying all else.

 

[Dwdnjck]

Speaking of recession. The health care sector has expanded dramatically and now approaches 20% of the economy.
Its in a bubble state of its own with almost no market competition to keep prices in check.
This sector will eventually will experience a recession, as rate increases are not sustainable.

These types of imbalances were responsible for recessions in the past.
Any suggestions on how to capitalize on this scenario.

Don't get sick.

 

[kenliles]

some interesting details re: BMW EV plans

BMW explains its electric vehicle plans in more details and it’s going in the right direction

 

[neroden]

Clearly, it depends. In some notable cases the political preferences shouted out the actual conditions. One I personally saw was the Iranian Revolution. I was in Tehran then, working for an American/Iranian JV. I studied farsi with a pretty smart young man and met many of his colleagues, who universally hated the Shah and his government. Some pretty senior Iranian government officials at the time (I'll name no names, they're still alive) did not even speak Farsi! Not odd, since the elites of the time were educated abroad for the most part. The foreign intelligence community did not see the revolution, which was NOT about Islam, even though fundamentalists ended out taking control.

You seem to be confirming my assessment in that case... that the information was obvious to those on the ground, but somehow filtered out of existence by the "intelligence community".

The long list of intelligence failures, though, can be exaggerated. In my ancient experience, pre-Iran and there too, the information was all there, presented and advocated, but killed due to politics.

I suppose it's a question of definition of "politics": it's not like the CIA didn't have access to the information needed to understand the internal stresses in the USSR (since it was public information), it's just that as a whole they never bothered to look at it because their mindset prevented that.

There's an old "joke", which isn't a joke, that the job of a spy is to speak the local language, live locally, read all the newspapers, and chat with people at coffee shops about them. Then report back what he or she read in the papers and what people were saying about it. Pretty easy job, eh? Nowadays with the newspapers on the Internet and the coffee shop discussions often on public Internet forums *too*, this is easier than ever.

But the information gets lost somewhere between that level and the decision-making level -- whether it's biased informants who are providing selective information, accurate information from the informants lost at the analyst level, lost at the level of the analysts' supervisors, lost at the level of the directors of the agency, or lost at the "political" levels of elected officials. Apparently from what I've read our unintelligence agencies are now seriously short on people who are fluent in various foreign languages, which is kind of ridiculous, since it's really the main thing for which they need specialized expertise. (And actually, if you can't recruit enough people fluent in a particular language... that tells you a very important piece of information, namely that you've developed a bad reputation among speakers of that language. A piece of information which is routinely ignored.)

 

[luvb2b]

ever since i heard elon talk about cash flows increasing as model 3 production increases i've had that thought in my head. i've been trying to formalize the idea into a model but haven't yet. my view has gotten stronger since my last model s went into and out of production in about 27 hours, and then was delivered in about 2 weeks.

the basic idea is that if you have a very high materials vs. labor mix of cost in the model 3 and the suppliers are extending terms to allow payment in 60 days vs. 20 or 30, then the speed of production and delivery will be much faster than the payment cycle. tesla will collect gobs and gobs of cash weeks before it has to pay suppliers. that cash balance/operating cash flow will continue to swell until such time that the production rate levels off for 60 days.

anyone else think about this? once it starts it will be a beautiful thing to watch because it should continue for several quarters as production ramps.

 

[ev-enthusiast]

Looks like #dieselgate might escalate to #petrolgate in Germany (link).
Any news on this topic from other countries?

From the above linked article (via google translate):

- In the current eco test of the ADAC, it becomes clear that a new exhaust gas problem occurs with petrol engines.
- Many modern direct injectors emit large quantities of particulate matter particles - a phenomenon known to date only from diesel cars.
- Thanks to a legal gap, such cars still meet the currently valid exhaust gas standards

The average particle emissions of petrol cars with direct injection technology - about 90 percent of the new petrol vehicles - are on average almost five times the limit valid from September 2017 onwards.

 

[jbcarioca]

ever since i heard elon talk about cash flows increasing as model 3 production increases i've had that thought in my head. i've been trying to formalize the idea into a model but haven't yet. my view has gotten stronger since my last model s went into and out of production in about 27 hours, and then was delivered in about 2 weeks.

the basic idea is that if you have a very high materials vs. labor mix of cost in the model 3 and the suppliers are extending terms to allow payment in 60 days vs. 20 or 30, then the speed of production and delivery will be much faster than the payment cycle. tesla will collect gobs and gobs of cash weeks before it has to pay suppliers. that cash balance/operating cash flow will continue to swell until such time that the production rate levels off for 60 days.

anyone else think about this? once it starts it will be a beautiful thing to watch because it should continue for several quarters as production ramps.

That will be true only if the supply chain is very fast (AKA, JIT), shipping and customs delays are negligible, etc. Tesla already has industry-leading inventory turns, somewhat limited because Tesla recognizes income whe teh end purchaser takes delivery while most competitors do so when vehicles enter shipment to a dealer/distributor.
I do think there is great potential from a very rapid building process for Model 3. There are too many moving parts for me to hazard a guess what the cash flow impact will be.

We may rest assured that storage products will have a pretty fast turn and will be positive contributors in 2018 if not this year. We also may be confident that rafts of new products will be in development so the capex and opex impacts will not diminish at all.

 

[ev-enthusiast]

BMW customer went to court in Germany because his i8 only went 18km on electricity and not 37km as advertised.
BMW taking back the old i8 and delivering a new i3 to the customer while paying for the difference in vehicle price.

BMW i8 Tesla killer, not.

(source: yesterday's German newspaper Bild)

 

[MitchJi]

Absolutely, and very good points.

This has created something else: it's created a new institutional bias, in addition to the hositility to labor -- one which biases against giving advice which says that the military-industrial complex is unnecessary or inappropriate.

This is a massive, and highly problematic, bias in terms of getting good advice. It means that you can't get the (bloody obvious) information that the NSA spying program is damaging our reputation abroad; you can't get the (bloody obvious) information that basing a fleet in Bahrain is simply annoying everyone in the area, and giving the US a reputation for hypocrisy on matters of democracy, while providing no strategic value... because these pieces of information would involve acknowledging that the military-industrial complex can be a *problem* rather than a *solution*. I guess it's the old "if you have a hammer, every problem looks like a nail" problem.

Somehow Trump has managed to buy in fully to the military-industrial-complex hype (every problem looks like a nail) *while* alienating the actual agencies involved, which is *quite an astonishing accomplishment* (in the bad sense) -- I would not have thought it a likely possibility.

The worst, most obvious current example is our "war on terrorism". It seems to be predicated on the theory that there are a limited number of terrorists, and like a conventional war, if you kill them all you win.

The truth is that we've given millions of people excellent reasons to hate us, and a small percentage of those people are so angry that they resort to violence. Mostly due to our selfishly motivated policies, like overthrowing the Iranian democracy. The best strategy to deal with that would clearly be to adopt policies that were just and fair, that are designed for the benefit of the people involved rather than benefiting large corporations or our geopolitical ambitions like obtaining oil in the middle east.

Instead we are murdering innocent people with drones and special forces attacks. We're employing a strategy that resembles putting out a fire by adding gas to it. The main strategic difference between obumble and trump is that trump is determined to pour on a lot more gasoline, and waste a lot more money in the process.

 

[racer26]

We're employing a strategy that resembles putting out a fire by adding gas to it. The main strategic difference between obumble and trump is that trump is determined to pour on a lot more gasoline, and waste a lot more money in the process.

I feel like Trump's strategy employs FOOF, rather than gasoline.

http://blogs.sciencemag.org/pipeline/archives/2010/02/23/things_i_wont_work_with_dioxygen_difluoride

 

[techmaven]

In other news: Tesla battery cell breakdown shows what is inside and difference with Panasonic’s regular cells

This article and video confirms that as many of us have said for a long time, the 18650s in a Model S or X are not the same thing that Panasonic sells as NCR18650B, even if they were originally derived from that design.

No, actually it doesn't. The commentator in the video makes a few mistakes, but most people won't care. The Electrek article uses the chart from the Panasonic research paper that I've referred to on occasion:

Development of High Power and Long Life Lithium Secondary Batteries

The primary gain is not from a different cell, but from reducing the exposed SoC during cycles. That same cell that lasts 300-500 cycles at 100% Depth of Discharge (DoD) has dramatically longer cycle life when reducing the DoD.

The original Model S cells in the 60 and 85 kWh battery packs are apparently NCR18650BE's. It isn't clear at all that they are different in any appreciable way, although the basics of binning and possible electrolyte tuning may be at play. In other words, Tesla may get the best output from Panasonic while lower bins of the same cell from the same line sold through, say, cheap Chinese outlets may not have the same performance.

Note, it isn't a bad thing that Tesla uses what is essentially a retail cell. What matters is quality, quantity, value, and performance. Other automakers have been unwilling to design packs that integrate thousands of small and relatively volatile cells and so they miss out on quantity, value, and performance. Tesla leveraged the overcapacity of cylindrical cell production to achieve the lowest price/kWh for any production automotive application.

 

[erthquake]

ever since i heard elon talk about cash flows increasing as model 3 production increases i've had that thought in my head. i've been trying to formalize the idea into a model but haven't yet. my view has gotten stronger since my last model s went into and out of production in about 27 hours, and then was delivered in about 2 weeks.

the basic idea is that if you have a very high materials vs. labor mix of cost in the model 3 and the suppliers are extending terms to allow payment in 60 days vs. 20 or 30, then the speed of production and delivery will be much faster than the payment cycle. tesla will collect gobs and gobs of cash weeks before it has to pay suppliers. that cash balance/operating cash flow will continue to swell until such time that the production rate levels off for 60 days.

anyone else think about this? once it starts it will be a beautiful thing to watch because it should continue for several quarters as production ramps.

Besides quality control, this is another positive of deliveries starting on the west coast, and progressively moving farther out from the factory. At the beginning, the time between production and realization of revenue (upon delivery) will be very short. And this won't get longer until production has ramped up significantly.

The 60 day terms are going to be a huge difference from the S/X ramps.

 

[MitchJi]

The original Model S cells in the 60 and 85 kWh battery packs are apparently NCR18650BE's. It isn't clear at all that they are different in any appreciable way, although the basics of binning and possible electrolyte tuning may be at play. In other words, Tesla may get the best output from Panasonic while lower bins of the same cell from the same line sold through, say, cheap Chinese outlets may not have the same performance.

It's hard for me to believe that Panasonic's production quality would vary that much, and that it would be cost effective to do that the testing needed to sort their cells based on cell quality, particularly testing cell longevity.

 

[racer26]

No, actually it doesn't. The commentator in the video makes a few mistakes, but most people won't care. The Electrek article uses the chart from the Panasonic research paper that I've referred to on occasion:

Development of High Power and Long Life Lithium Secondary Batteries

The primary gain is not from a different cell, but from reducing the exposed SoC during cycles. That same cell that lasts 300-500 cycles at 100% Depth of Discharge (DoD) has dramatically longer cycle life when reducing the DoD.

The original Model S cells in the 60 and 85 kWh battery packs are apparently NCR18650BE's. It isn't clear at all that they are different in any appreciable way, although the basics of binning and possible electrolyte tuning may be at play. In other words, Tesla may get the best output from Panasonic while lower bins of the same cell from the same line sold through, say, cheap Chinese outlets may not have the same performance.

Note, it isn't a bad thing that Tesla uses what is essentially a retail cell. What matters is quality, quantity, value, and performance. Other automakers have been unwilling to design packs that integrate thousands of small and relatively volatile cells and so they miss out on quantity, value, and performance. Tesla leveraged the overcapacity of cylindrical cell production to achieve the lowest price/kWh for any production automotive application.

I suspect you may be correct for the early era of Model S, that the cells were essentially NCR18650Bs.

I believe that the cells currently being put into MS/MX are not quite the same thing. I think there was something about increased density or something that allowed the newer, higher capacity packs with the same number of cells that were in the original packs.

 

[geneclean55]

What did MX finish 4Q16 at?

US customers received MX cars with VINS in the 31XXX (and maybe 32XXX) range at the end of last quarter. VINS in the 33XXX (and maybe 32XXX) range I believe were assigned to overseas customers last quarter with delivery occurring this quarter.

 

[GregRF]

It doesn't need to be 60% thicker, see my computation above, it only needs to be 25% thicker. or 58% more volume (cross section), assuming a round cross section.

With thicker AC cables they may need to be up-sized due to the skin effect. (AC induced magnetic field causes the current to migrate toward the outer skin of the cable, so the current carrying capability of the center is underutilized)

 

[techmaven]

It's hard for me to believe that Panasonic's production quality would vary that much, and that it would be cost effective to do that the testing needed to sort their cells based on cell quality, particularly testing cell longevity.

It is a normal part of any manufacturing process. The cells are aged and cycle tested. At that point, they can allocate them to bins. Clearly, Tesla takes a huge portion of their output. Also, it is possible that the difference between the NCR18650BE and the NCR18650B may have blurred over time... ie. more cells are now basically NCR18650B's coming off the line. Unfortunately, this level of detail is all wrapped under trade secrets.

I believe that the cells currently being put into MS/MX are not quite the same thing. I think there was something about increased density or something that allowed the newer, higher capacity packs with the same number of cells that were in the original packs.

The closest match to the newer cells is the NCR18650GA. Unfortunately, it would be very difficult to know exactly without having inside knowledge of Panasonic's production processes. Of course, Tesla would know.

And clearly, with the Gigafactory production for future automotive applications, Tesla/Panasonic can optimize for whatever makes sense given their technologies, so much of this is moot going forward.

 

[erthquake]

With thicker AC cables they may need to be up-sized due to the skin effect. (AC induced magnetic field causes the current to migrate toward the outer skin of the cable, so the current carrying capability of the center is underutilized)

Someone correct me if I'm wrong, but I believe the inverter is part of the motor assembly. So these long cables that connect the battery to the motors and charger are DC.