[propman]

Tesla could realistically hit 500,000 cars next year, almost 10% of total cars sold in America. The electric revolution is here and it’s finally profitable! Budget EVs with solid range are coming next year too with the 2019 leaf!

Convince congress to extend the federal tax rebate (no real increase in spending just keeping what’s already there) by like 2 years which is just maybe $1-2B more and electric will be the way to go for good!

[kllrnohj]

> Tesla could realistically hit 500,000 cars next year, almost 10% of total cars sold in America.

US consumer auto sales per year are in the 16-17 million range. I personally wouldn't consider 3% to be "almost 10%", do you?

But I don't think they realistically can hit 500,000 cars next year, either. They moved 70k cars in Q3. There's no reason to believe they can double that in 6 months or so particularly as the Model S & X sales have been fairly stable. So that means the Model 3 needs to somehow grow to around 400k/year in order to get to 500k cars next year.

They state production of Model 3 was up to 5,300 per week and that the production system had stabilized with gradual monthly improvements. So currently Tesla can only even build 275k Model 3's a year. I don't think doubling that in such a short time frame meshes with Tesla's comments of "gradual monthly improvements."

350,000 cars in 2019 seems much more likely assuming Model 3 demand holds strong.

[propman]

I meant cars, not all autos. Cars are 6.2 million. Musk said he’s trying to get to 6000 a week by the end of August and 10,000 a week in 2019. The demand is clearly there as are the profit margins even sans ev credit. I know he overestimates things a lot but that tells me they are seriously considering ramping up production. As the production and now delivery processes streamline, they can definitely hit 6000 by end of year and realistically hit 8000 by mid 2019.

[swarnie_]

While i'm totally behind this rocket ship you need to take any number Musk give you and add a random number of months between 6 and 18 to it.

[Reason077]

Elon Musk has reiterated recently that Tesla's goal is to reach Model 3 production of 10,000 per week in 2019.

But not all of those sales will be in the US. According to the Q3 update document:

"The mid-sized premium sedan market in Europe is more than twice as big as the same segment in the US. This is why we are excited to bring Model 3 to Europe early next year. "

[kllrnohj]

> But not all of those sales will be in the US. According to the Q3 update document:

Which would make it even less likely Tesla can hit 10% of US passenger car sales as they'll be sacrificing what little production they have to try and break into other markets.

Anyway just hitting 10k/week in 2019 isn't sufficient for the 500k number. They'd need to average 10k week for the entirety of 2019. Very different things.

[FireBeyond]

The number of auto sales in the US is determined by demand, not by how much auto makers can supply (within price adjustment constraints, of course).

[Reason077]

The Model 3 is currently production constrained. Tesla sell as many as they can produce.

[adventured]

> US consumer auto sales per year are in the 16-17 million range. I personally wouldn't consider 3% to be "almost 10%", do you?

The parent comment said "cars." You switched that to consumer auto sales. 6.3 million passenger cars were sold in the US in 2017. With passenger car sales declining and generally losing popularity in the US (dropping from 7.9m in 2014), it's plausible that 500,000 will in fact be "almost" 10% of total cars sold in the US in 2019.

[SilasX]

In common parlance, an SUV counts in "all cars sold", so it's misleading, even if technically correct.

[ProfessorLayton]

I suppose if US passenger car sales, a segment the Model 3 & S belong to, decline by 1/3 in 2019 that statement could ring true.

[burger_moon]

Ford killed off their entire car line up except the mustang so there will be a fairly significant drop from that alone.

[antisthenes]

What's just as amazing is that this whole revolution is done off the back of the Li-Ion 18650/21700 battery cell.

Time will tell how good of a choice it is.

[aphextron]

>What's just as amazing is that this whole revolution is done off the back of the Li-Ion 18650/21700 battery cell.

18650 is the cell form factor, not a specific anode/cathode/electrolyte combo. There have been leaps and bounds in cell chemistry over the last 10 years which have made it possible to build high-end expensive cars, but we still need about double the performance of current tech to achieve true low cost mass adoption. A $45,000 car with 300 miles range is just barely enough to convince well off people to buy a new toy. The first $20,000 EV with 300 miles range will change the world.

[greglindahl]

I recall reading many HN posters commenting that this cylindrical form factor was a terrible choice for cars. Now several other carmakers have adopted it.

[ethbro]

The Edison (E26/7, circa-1900) base isn't the best choice for light bulbs.

But it is good enough.

And now we're plugging high-powered LED bulbs into them. Technology finds a way. Be first!

[gruturo]

It's only a terrible choice if you do not think about cooling the batteries, and having some physical protection against runaway thermal effects.

If you do, suddenly that empty space becomes precious.

The fact that other carmakers started using it is a good sign.

[sliken]

From an engineering perspective it's suboptimal since they have small surface area relative to volume. That's probably why Telsa went with the 2170 instead of the 26650, which has a larger diameter.

Despite this disadvantage the Telsa battery cost, power density, cost per kwh, peak power, and longevity are unmatched. Especially as measure by real people driving real cars on real roads.

So sure in theory batteries with more surface area would have greater package efficiencies. So far nobody has managed to.

[akira2501]

> So sure in theory batteries with more surface area would have greater package efficiencies.

There's way more to a battery than just density; however, one of the things that needs to be managed is the heat generated during charging. There might be a case to be made for slightly suboptimal packaging strategies if it makes cooling all those cells that much easier.

[rconti]

Yeah, this really strikes me as a forest for the trees kind of situation. Sometimes it pays to engineer things to death, other times you need to step back and see the whole picture; how important is packaging for efficiency vs cooling vs manufacturing speed, etc.

Automakers putting battery packs in the trunk of their cars are complaining about battery size.

Is anyone complaining the floor on a Tesla is too thick and compromises interior space? Of course not.

The packaging probably DOES add a not-insignificant amount to the total weight of the pack vs an ideal battery of some custom design. But again, you have to weigh (ha!) all of the factors.

[Gibbon1]

I read an article that made the case that lithium of batteries are sensitive to the direction of the heat gradient. Tried looking, can't find it. But the upshot was you want the gradient parallel to the electrode layers not perpendicular. Which means you want to remove the heat from the ends of the cell.

[ecpottinger]

I can't find it either, but I remember reading it.

However, http://jes.ecsdl.org/content/163/9/A1846.full points out some of the details.

[mavhc]

Don't most other EVs use flat pouch cell batteries?

[greglindahl]

I think GM switched from pouch cells to cylindrical cells between the Volt and Bolt. The Leaf still uses a flat pouch. In the US, the Bolt and Leaf are the only other BEVs with significant sales.

[simonebrunozzi]

Right! Lots more details here (and good comparison of the two): https://www.dnkpower.com/teslas-mass-production-21700-batter...

[marricks]

Is that comment anything but FUD?

They are able to make the cheapest batteries for cars and have seemed to be pretty successful...

[gamblor956]

Panasonic makes the battery cells...Tesla just assembles them together and integrates the cooling system.

[dannyw]

The comment is a precise and accurate definition of the technology, I don’t see why you jump to “FUD”? Not everyone is trying to influence the stock price.

[hikari-boulders]

I don't understand how "just maybe $1-2B more" is "no real increase in spending".

[bsder]

> Tesla could realistically hit 500,000 cars next year, almost 10% of total cars sold in America.

Is that really feasible? That's roughly BMW and Mercedes combined in the US.

It's really easy to forget on the West Coast of the US that BWM, Mercedes and the like are actually rare in the rest of the country.

[xienze]

What? Define “rare”. If you’re in any reasonably-sized metro I’d be quite shocked if a day goes by where the average person doesn’t see a BMW or Mercedes.

[smogcutter]

If the electric revolution requires a federal subsidy to compete then it has not, in fact, arrived.

[ZeroGravitas]

By that logic the gasoline engine has not yet arrived either.

[morsma]

Does gasoline engines receive subsidies? I just bought a brand new ICE vehicle, and I'd sure like to know when my check arrives...

[bjelkeman-again]

Not the cars, but the fuel. “America spends over $20bn per year on fossil fuel subsidies. Abolish them” https://www.theguardian.com/environment/climate-consensus-97...

[cobookman]

US also spends money on electricity subsidies.

https://en.wikipedia.org/wiki/Energy_subsidies#United_States

"""

Renewable energy: $7.3 billion (45 percent)

Energy efficiency: $4.8 billion (29 percent)

Fossil fuels: $3.2 billion (20 percent)

Nuclear energy: $1.1 billion (7 percent)

"""

[ecpottinger]

Good, get rid of them ALL! Then see which is truly better. Musk has clearly said he would prefer if all energy subsidies were canceled.

[nightski]

If EV wins and they can no longer collect all that juicy tax revenue from gas - how are they going to afford to maintain infrastructure? They collect far more than 20bn in revenue from fuel taxes.

[zie]

They will just move the tax to the Tesla Superchargers and the other electric charging stations. Alternatively they could tax it at the DMV level, like they already do to some extent. i.e. instead of a $100 registration fee every year, it will be a $500 registration fee.

I'm going to guess it will be a mix of both, but the govt(s) will be slow to do either, until they realize, too late, that they didn't meet their income goals on fuel taxes last year and suddenly have a shortfall in the budget :)

[philipkglass]

Check odometer readings as a requirement of renewing the vehicle's registration. Assess a fee as a function of vehicle weight per axle and distance traveled. Sure, at the state level there will be some missed edge cases (what if I am registered in Washington but do most of my driving in Idaho?), but it should be close enough. It's basically fair and way less invasive/expensive than GPS-based tracking for road use.

[romwell]

Good news, the check has already arrived!

Perhaps not for you, but it sure did[1]

[1]https://en.wikipedia.org/wiki/General_Motors_Chapter_11_reor...

[kazinator]

Not to mention the pneumatic tire on asphalt road.