Deep Dive Into Tesla’s Low Cost, Short-Range Model 3 Pack Design


What’s the story here and how much might this new battery pack really save Tesla?

We’ve had lots of news in the last week. Tesla told us during their earnings conference call that they maintained 20% gross margins during the 3rd quarter. On the other hand, it also delayed production on the short range version (again) because the gross margins are not quite there. It appears Tesla is waiting for a new short-range pack design.

We came up with a new design that achieves the same outcome, that’s actually lighter, better, cheaper and we will be introducing that around the end of this year – probably reach volume production on that in Q1 or something. That will make the car lighter, better, and cheaper and achieve a higher range.

What is this new pack design that’s lighter better and cheaper, how much money does it save Tesla, and what might be its impact on gross margin?

Here’s a tweet from CEO Elon Musk about the MR pack:

It’s a long range battery with fewer cells. Non-cell portion of the pack is disproportionately high, but we can get it done now instead of ~February

4:14 PM – 18 Oct 2018

In this last tweet, Musk is saying that the new MR pack will use the same module casings as the long range, only with fewer cells. However, it is our feeling that that case configuration will just be temporary until around February. At that time, the MR pack will switch to shorter module casings. Tesla will then simply depopulate the new MR module casings or just de-rate it with software for the 50-kWh pack.

Aside from the two Musk tweets above, Tesla hasn’t told us much about the new short-range pack. However, we can deduce what it is based on what they have told us and a bit of computer modeling.

Electrek has reported that the mid-range pack is 62 kWh’s usable (ref). That number makes sense based on our computer modeling. Our modeling also predicts the size of the SR pack required to get 220 miles AER = 50 kWh. So, we have three size packs: the LR pack with a total available usable 78.3 kWh (ref), the mid-range pack with a usable 62 kWh plus a reduced size module (in February time frame), and the SR pack with 50 kWh usable in the new shorter mid-range module casing with fewer cells or just a software de-rate.

Tesla has also told us that they will be getting new Grohmann battery assembly automation equipment in the 4th quarter, which will speed production of battery packs and reduce prices. We don’t know exactly how much the new Grohman machines will lower pack cost for Tesla, but we can estimate fairly accurately how much money Tesla will save by reducing the cell count in the two packs. In addition to fewer cells, we also believe Tesla has designed a new mid-range module that is shorter than the LR modules, so they can save on all the connections and module casing etc. as well. So, the move to the new MR module should save slightly more money than just removing cells.

We also are estimating that Tesla will simply de-populate (or de-rate) the new mid-range modules in order to get the 50 kWh SR module, based on what they have done in the past.

Here’s our estimate of what the new pack designs will be and how much they will impact gross margin.

An explanation of the math:

The first step was to calculate the gross margin in $ using average sell price times Tesla’s claimed 20% margin.

Each time Tesla lowers its cost of the pack, it lowers its cost of goods sold. That’s nice, but if you lower the selling price of the car more than you lower your costs you end up losing profit margin. So with each new configuration, we subtract the difference between what Tesla saved on cost and the drop in the average selling price. If Tesla drops the price more than their cost savings, it drops their gross margin by that amount as well.

We calculated the cost savings as follows:

Based on the $115/kWh cost to tesla for cells and another estimated cost of $10/kWh for smaller module casing and connections, we estimate Tesla lowered its cost of goods sold by $2,075 for the MR. It dropped the selling price of the mid-range version $3,000, so it came very close to holding the same margin with the mid-range version.

The short-range version is tougher. Tesla takes out less kWh’s than going to the mid-range version, plus, if our theory on depopulation (same module size with dummy cells) is correct, then they don’t save the extra bit associated with going to a smaller module. Now, add on top of that a bigger jump down in price that leads to a pretty big hit to gross margin reduced from 19% for the mid-range version to only 12% for the short-range version.

However, this 12% margin has doesn’t factor in the new Grohmann assembly machines, as well as whatever else is done to the assembly line to increase production rate. Even with those improvements, it’s still tough for us to imagine Tesla getting back to 19% gross margin with the new Grohmann machines. The numbers are just not there. Even in the extreme case where Tesla reduced the pack cost by an additional 10% with the new Grohmann assembly machines, that would only increase its gross margin from 12% to 13.4% … still way shy of getting back to 19%.

In comparison, Troy Teslike is predicting Tesla will reduce its gross margin from 15% to 5% with the short-range pack (ref). So again, way shy of acceptable.

Therefore, we think that Tesla will have to increase the price of the short-range version around $2,500.


Let us know what you think Tesla has up its sleeve with this new “lighter, better and cheaper” battery pack. Maybe they will change to a new cooling scheme (flat plate?). Maybe they have a gen-two 2170 cell design waiting in the wings?

Also, we are proposing that the MR pack is the main configuration and the SR pack is a depopulated (or software de-rated) MR pack. Perhaps you think the SR pack should be smaller modules, not a depopulated MR pack?

For those that are wondering why we haven’t factored in a lower cost electric motor and wiring as we discussed in an earlier article, we have it from a reliable source that the same motor is used in all variants of the Model 3.

This article was a team effort between the author and Keith Ritter,

Thanks for reading our articles.

Categories: Battery Tech, Tesla

Tags: , ,

Leave a Reply

89 Comments on "Deep Dive Into Tesla’s Low Cost, Short-Range Model 3 Pack Design"

newest oldest most voted

Good analysis.

Re short range margin: It depends – are there other overall economies of scale yet to be reach that improve the gross margin on every car?

Then Tesla might be happy with say 10% SR if average is still 20%.

350 km of range is still a good for the SR Model 3.

Not related to the math and production costs but would the current Mid-Range battery pack be better than the post-February version? I.e. better cooling/heating and better regen ability due to using the components from the long range version?

Great Article. Nice work all.

I think that the short range pack will be a unique one properly sized for the 50kWh of cells it contains, as this was always the plan to minimize cost and weight (and thereby maximize range). The MR pack is a recent Limited Edition compromise, and I wouldn’t be surprised to see it go away entirely when the short range comes out.

I agree. Tesla typically has two S/X battery sizes, transition periods with three sizes don’t last long.

I also suspect SR will end up closer to 240 miles combined, with more highway range than the 60 kWh competitors. The few on the reservation list who hold out for true $35k will get a s/w limited car in early 2020.

To get 240 miles, that would be 208 wh/mile. I don’t see how that would be achievable through battery weight reduction.

I think they’ll increase SR battery size a bit.

I agree, they referenced a possible bump in the range of the SR car when it finally comes out next year. Getting it slightly above Bolt and others would be a win, and the range bump could be used to justify the price hike of ~ $2500 to achieve gross margin targets. Entry level at $37,500 isn’t really a big difference vs $35,000 especially when you consider what inflation has done to the dollar since the $35k price was announced. Add a bit more range and it’s a no brainer. Oh sure, there will be people who yell and scream “see, I told you Tesla would never sell the $35k model!” But their screams will be muted by the sales figures as Tesla sells every single $37.5k model it can produce for years to come.

I think there is a couple things going on, Tesla is taking a look at the overall design of the Model 3. Making the standard range pack car a sort of gen 2 car that will deployed across the all Model 3s. I’m pretty sure It has been mentioned, somewhere that Tesla has spoken to Monroe Associates over the phone, and buying their report would have been chump change for Tesla to purchase a report. Tesla keeps making incremental changes to the cars, that lead to cost savings. The additional tweaks could really get the price down. Plus as the volume of the parts goes up the price of the components goes down. When you start talking parts on the level of 500K cars per year you can really throw your weight around to get suppliers to sharpen their pencil. There is also some X-factors we don’t know. The price of the standard interior could be significantly cheaper for Tesla to produce over the premium package. The standard range car could have all of the autopilot hardware removed. The cost of aluminum due to tariffs is also cheaper than steel and could be incorporated into the gen 2 car, thus… Read more »

I agree with much of what you say, but I don’t think they can save much on the interior. Most of the PUP features are very cheap at wholesale. I also don’t think they can delete Autopilot h/w, as they use it for basic safety features like AEB. I doubt AP h/W will cost more than a few hundred by then, anyway.

Agree. PUP deletion (no more power SW, power seats, maybe console, or possibly AP hardware) all total to a margin preserving sum. That, and making “35k” unattractive, or sufficiently “de-contented” can make up for pack costs not falling dollar for dollar.

It’s not like Tesla is going to take its 235 tires down to BMW’s 155’s, or freeze the rear windows shut, but they can still shape demand and offer a lower selling “35k” option.

I don’t understand how you came up with the usable capacity of 50 kWh for SR. We know that the LR has 4416 cells and the SR will have 2976 cells (Tesla source). We also know that the LR has 78 kWh usable (EPA source). A simple calculation results in 52 kWh (almost 53 actually) usable for the SR.

Until the new car hits the road, it’s safe to assume the specs of the car(s) are in flux.

Until someone totals one of these and Jack Rickard gets a battery and counts the cells, everything is conjecture.

I completely agree! As a Mazda fan, they employ a gram weight strategy to save weight to make their cars light.
( Know they don’t offer a EV :-/ )

I believe Tesla will do the same on the SR model 3 as long it doesn’t increase cost or compromise structure. Every single cell in the pack adds weight and cost. I’m sure they will work extremely hard to decrease both of these.

That is fine, but in my opinion the official numbers say it should have almost 53 kWh usuable. If you deviate from that, you should support that statement and this article does not.

You can not use a simple extrapolation. The SR will need a bit lower energy per mile of range, because the weight of the battery is less. The same can be seen in S and X, where longer range versions have lower MPGe, though everything besides the battery is essentially the same.

Good point, Olivier. I also think the 62 kWh is too low. This article quotes Electrek at 62 kWh, but they are at 74 kWh usable on the LR, not 78.3 kWh.

I’d use:
SR – 96S31P = 2976 cells = 52.8 kWh usable
MR – 96S38P = 3648 cells = 64.7 kWh usable
LR – 96S46P = 4416 cells = 78.3 kWh usable

SR/MR usable might be a bit less as Tesla may leave more buffer for smaller packs. But cell count drives cost. At $125/kWh I get savings of:

MR: $1700 less than LR (vs. 2075 in article)
SR: $3188 less than LR (vs. 3455 in article)

My gross margins would be almost a point less. The bigger problem for gross margins is EAP take rate, though. People in the $35-45k price range are much less likely to buy than the early adopters in the $50k++ range. Tesla just raised the “penalty” for delaying EAP purchase to try to combat this issue.

I wonder about the margins based on volume. You really only save on a redesign if your volume is equal or greater than the previous. So if the volume of MR is lower than LR you “depopulate” to save money. If the expected volume is greater (as expected for the SR) you would do a redesign of the pack. So the real savings are in whatever is going to be the high volume model. That is the one you do a pack redesign for. If both MR and SR are expected to outsell the LR, only then do two pack redesigns make sense.

“You really only save on a redesign if your volume is equal or greater than the previous.”

That’s a glittering generality, and pretty clearly isn’t always true. Look at the redesigned wiring harness that Tesla is using for the Model 3: Significantly less length of wires results in immediate significant cost reduction, and would even if the volume of production was not that large.

We don’t know how much money the redesign of the Standard Range Model 3 battery pack will save Tesla. The authors of this article don’t, either. Their claim that Tesla will have to increase the price of the SR TM3 by $2500 is nothing but pure guesswork on their part, and I will be very surprised if it turns out to be true.

“For those that are wondering why we haven’t factored in a lower cost electric motor and wiring as we discussed in an earlier article, we have it from a reliable source that the same motor is used in all variants of the Model 3”

Umm, almost. Good that you corrected the issue regarding wiring being the same across variants. But based on released service info on the Tesla service web site, the rear motor assembly is different for the standard & midrange vs. long-range (RWD,AWD,Performance). Two assemblies are detailed for the rear drive. So yes, the motors inside ‘could’ be the same, but there is something different in the assemblies (gearing, etc) which could result potentially different cost.

You’ve ignored the impact of increased production on depreciation. Unlike other manufacturers Tesla uses straight line depreciation. This explains why depreciation expenses have reputedly dropped from 15% to 7%. Raising production from ~4,300 per week to ~7,000 per week raises your margin by about 2.7% so your >12% becomes >14.7% which looks acceptably close to the 15% projected by Elon. That 0.3% amounts to about $100 minus your undetermined Grohman factor.

@Gasbag, I think it’s pretty much an established fact that Tesla has maxed out it’s capacity in Freemont. Assuming even higher production rates without extra CapEx (which may take years) means they need to find even more efficiencies in the current facilities to exploit; it will be very hard for them to do that, to say the least.

On the contrary, Tesla is saying that get to 7000 M3 per week and 9500 cars total at Fremont won’t be hard and will happen soon.

Tesla has been *saying* that for more than 2 years now. In the meantime last quarter they averaged about 4,000/week, after more than an year of ramping up. Then, if insideevs is correct last month they sold 17.7k; assuming production approximates deliveries this is ~3,900 cars/week. So, if they are shooting for 7K, it doesn’t seem they are going in the right direction.

Bottom line – I don’t believe they are physically able to get more than 4, maximum 5k/week from this facility. It will take additional CapEx and time if they want to do it, but I doubt they want that too, they first need to establish the sustainable demand for the product and 4 or 5K a week may be enough for that.

Gasbag – Fixed manufacturing depreciation is only about 150 million per quarter. The rest of the 503m shown in the earnings release is tooling depreciation (80-85m) which varies with production, depreciation of offices and other non-manufacturing PP&E (30m guess), depreciation of leased solar systems and vehicles (~120m), financing amortization (~80m), impairments (?) and a few really minor things.

150m/quarter is roughly 1800/car (S/X/3) in Q3. Ramping Model 3 volume to 7000/week only decreases that another 300-400 bucks and improves gross margin by less than 1%.

I seriously doubt they will software limit a MR battery pack to a SR capacity, as cell costs in the car are substantial. They did it for a time on the S, but now they are looking at a lower priced car at mass production quantities. A depopulated MR module might make sense.

I agree with you, I’m sure they used as few cells as possible on the MR battery. They probably just hacked of a few cell bundles and altered the code in the BMS. Then filled the rest of the pack with some type of foam. This lowered the voltage and changed the top speed. Increasing the current.

To actually restructure the pack like going from S108P40 to say S108P35 would require a significant change to the layout of the battery cells vs going to a S94P40 configuration. Just hacking off a group of cells requires little to no changes to any equipment. This would change the Voltage from 400v to 350v. Hence the lower top speed and acceleration due to less voltage and higher current for a given amount of torque.

huh ful lsize battery Teslas and most other EVs are 96S

I just threw out a number for the S/P to make the point. It’s easier to change serial config then parallel.

But Tesla never seems to change serial. EPA reports of 350v vs. 400v notwithstanding. I’ve never seen a teardown that was anything besides 96S.

They would not lower the voltage, still 96 cells in series. Fewer cells results in less available amperage.

Absolutely. Tesla only made a software limited battery pack as a marketing strategy; they sold it as a car which could be upgraded to higher capacity… for a fee. It’s not reasonable to suggest that Tesla would simply “eat” the cost of the additional battery cells and connections, with no hope of making up the cost later on — as they did with the newer version of the “Model S60”.

“A depopulated MR module might make sense.”

Yes, I find that to be a much more believable and likely scenario than the one suggested in the article.

I’m leaning towards a custom pack for Standard Range just because of the quantities of vehicles they will be producing. Standard Range will probably be the most popular version with so little leeway in margins.

Zero percent chance of a derated pack. Panasonic has said repeatedly, they can’t make enough batteries. The batteries aren’t available to de-rate. On top of that, Tesla doesn’t have the cash to loan out extra batteries without booking the cash immediately.

De-rating the pack could lead to the possibility of a much longer battery warranty. Degradation becomes a complete-non issue if you’ve got an extra 10kWh to play with…

@John, extended warranty is nice if you are a car buyer but for the manufacturer is just another liability that eventually hits the bottom line. Since Tesla is very concerned with maintaining profit in the near term I very much doubt they will go that route.

Right. Putting more battery cells (and the connecting wires than necessary into every car just to save a bit on the amount Tesla must set aside for warranty servicing, would be “Penny wise and pound foolish”. Not gonna happen.

Cutting down on the number of cells also reduces the weight of the car, which reduces the amount of energy needed to accelerate, which improves energy efficiency and therefore reduces the total number of cells needed in the pack… a positive feedback effect which will help Tesla save money. That’s an opportunity Tesla won’t throw away just to avoid the cost of developing a shorter battery module.

If Tesla ever truly goes the “sports” route, they will select more power dense batteries to go with the lower inertia a lower cell count gives you. But they won’t because their customers don’t appreciate how inertia reduces sports car abilities. Track BEVs create dichotomy in ideal design, and it is understandable the average Joe doesn’t need to hold out for this stuff, or 800v, and would be harmed by carrying extra battery weight around. -“Joe” wants ‘mo energy-dense batteries, instead.

Thus, Tesla nails it.

Agree – the SR will have fewer batteries….

I am leaning toward a new pack structure for all versions of the Model 3. Lighter, more efficient and easier/faster to build with the new Grohman machines. This will increase margins across the board and allow a 35k version.

You are suggesting 47k ASP for Model 3 SR. I question how realistic this is. Looking at how people buy cars in this price group, I don’t see a chance of 35K cars averaging 12K in options. Also after factoring in the tax credit (now to be <4K), the difference becomes even more significant – in short the market for 31K cars is much much larger than the one for 43K cars. If I had to guess I'd say 40-42K ASP is more realistic.

I think in reality we will see a bit of all of this to make ends meet on the Model 3 SR:
1) A small increase in base price (as you suggest ~2,500 sounds reasonable)
2) A sustained effort of Tesla to reduce costs per car – how much that would be is anyone's guess
3) A certain hit in gross margins – but I don't think they'd accept anything lower than 12-15%, it will bring back the company in the red

(edited because of mistake in my calculations)

Perhaps they will press Panasonic more and more on the cell costs once they have demonstrated long-term demand.

Not likely. Panasonic isn’t even making money yet on its Gigafactory production, because of the high capital investment they have had to do.

Where Tesla can save money on parts is by building the Model 3 in higher volume. Ordering a larger number of parts from a supplier can lead to lower unit costs. Ditto for the parts Tesla makes in-house; a higher volume should yield lower unit costs.

Tesla may also be able to save money on labor if it can accelerate the speed of the assembly lines and/or replace some human labor with robots.

In general, I think there are a lot of opportunities for Tesla to shave costs here and there; opportunities which this article is ignoring.

Very true and they said as much on conference call. “We are just now able to focus on cost optimization.. expect that trend to continue in Q4” .. something to that effect.

$2,500 increase of an ATP that’s already projected at a very high at $47,500 for the affordable model? Hmm, that doesn’t strike me as helpful for that 500K/year sales target. I hope Tesla’s gross margin will actually turn out to be way higher at ATPs like that.

I wonder about the future of that SR version anyway. I think 220 miles is just not going to be enough for most customers who will be paying premium for a car with considerable less range than an increasing number of non premium competitors, especially now that they have the MR option. I predict that will be the base version pretty soon.

My bet is the 220 mile range SR becomes the MR low end car (at a higher price) before the start of 2020. Just like the Model S where 40 and 60 we once the low end cars….

40 and 60 were eliminated due to low demand of a premium high end car for rich people or very early adopters who had expendable income. That is completely different than the huge demand for the 35k SR. I am guessing there were about 100k out of the 435k original rez holders that want the SR. 40-60k in the US alone. Some have given in to the LR or LMR but most are still waiting for SR. Once there are a couple hundred thousand in driveways, it will continue to draw the regular folks.

I think it’s pretty clear there will be a new pack design for the SR rather than depopulating the MR, that’s the whole reason they introduced the MR – as Elon said “we can do it now instead of ~Feb.” Based on earlier statements from him, it seems like their new pack design for the SR should be lighter and cheaper, so should have a better GM than your numbers from that. Additionally, if it’s enough lighter they might be able to get away with a few less cells to get the same rated range which would be another chunk of savings. It’ll be very interesting to see a teardown of the SR pack when it eventually comes out.

I agree.

I think the author is just misinterpreting the February quote. To me it reads that the mid-range is a temporary measure until the standard range is ready. Mid-range might even disappear at that point to simplify production.

Wait, didn’t they already state that the SR is a new, less complicated pack design while the MR is the LR pack with less cells?

Thank you!

I find it very strange that the relevant quote actually appears in the article: “We came up with a new design that achieves the same outcome, that’s actually lighter, better, cheaper and we will be introducing that around the end of this year”

Very strange to see that this analysis by George and Keith ignores that, and proceeds on the assumption that the SR pack is just like the LR pack only with shorter modules.

The quote from Elon points pretty strongly to a new architecture, not just making shorter battery modules with fewer cells in them.

In the Swisher interview, Elon said, “large product that will be compelling for utility customers” is soon to be completed.

This could mean something completely unsuited for autos, but I’m wondering if a larger cell format could be among the speculatioin?

He said in the recode interview it had to do with utilities, so the speculation about the MegaPack is probably what he is talking about. A MWh pack for Utility company installs.

Yup. “Large product… for utility customers” points to a large pack size, not a larger cell size.

Geography is also affecting gross margins. Selling cars in the Bay Area is more profitable for Tesla, vs selling in Norway or NJ.

Humorous how some want to micro manage from a keyboard.

Won’t they also save on the non glass roof version in the SR

Perhaps a bit, but the “steel roof” option only replaces about 25% of the glass roof with steel. I dunno how much money that will actually save, especially since that means they’ll need two varieties of Model 3 windshields in production instead of just one, so there will be a additional costs for that.

Non premium audio…
Non premium seats….
Maybe no LED lighting….

Audio and seats, yes. LED lighting is far more efficient than alternatives, already in place and it would take significant engineering changes to move to a different style, and I doubt any cost savings given thicker wire, bigger sockets etc. so no, LED lighting will remain.

No rear seat heating, maybe no front passenger seat heating?

No power seats, mirrors, or steering wheel..

Long ago (year or two ago?), I predicted Tesla 3 SR will be 50 kWh while others were speculating 55+, because math showed 220 miles + cg 0.22 to be about 50. Looks like George comes to same from other angle.

They can increase the SR price by $2.5K, but that’s not the only options. They can terminate SR or take a hit in margin and make up for it with volume. But note the average SR price in this article is $9K under $56K, which is $47K. That’s why I “haha!” every time there’s mention of $35K Tesla 3; that will be very token number of cars. I shall dub thee, “Token”

comment image?cid=3640f6095bdc8b5e415a534f67ddefe9

I think they will get to a 35K version as Elon has promised several times. But they have to get to a very safe Gross Margin and strip out the Premium package and glass roof. I can foresee a SR version with mandatory “plus” package for 2 or 3K.

The issue is that the $35k figure is now pointless and is a token sell price. $35k was the starting price for the ICE competition BACK THEN. All have raised their prices due to inflation since then, but Tesla now has undercut the competition in price. Even 60kwh EV’s are starting close to $40k. There is no point in selling a $35k Model 3 at this point, especially in volume.

They offered the S 45 for a short time. A couple months after it was available they raised price to equivalent dollars. Shortly thereafter they cancelled it.

What? No part of that is true.

There never was a Model S45, and the Model S40 was never “available” except as an early pre-order, which was cancelled before production. I don’t recall that they raised the price of the S40, either, altho Tesla tried to persuade all S40 order holders to upgrade to an S60; in some cases offering special deals.

$35k Model 3 is part of the whole goal of Tesla. I think it will happen. Demand is too high for it, thus LMR to tempt some people to spend more and get a car now. LMR is like the software limited S 60. A transition vehicle that works with current production setup and supplier sourcing. In the case of Model 3 it also helps with the lower supply of cells from Panasonic/Gigafactory. Like others, I believe the SR will first be offered with Premium interior, then the supplier sourcing for non-premium interior will start showing up and they can start the teething on that install. True $35k may not be available until EOY 2019, which is sad, but it is what it is. I never thought Europe would get a Model 3 before my day 1 California $35k reservation, but that is how it is playing out.

I think for the SR it must be factured in the profit of the 5000$ of the Premium pack that is standard in the other models. The pack does include very basic options that you cannot renounce on such a car, fitted as std on half cost cars, who would buy it without? Perhaps no more than 5% of buyers. Old trick. So the SR is in fact a 40,000$ car.

Why did you omit analysis of the software-crippled pack when you consider that a probable alternative? Perhaps because the cost saving then is zero?? Reducing gross margin in this model to about 8.5%.

As far as the statement goes that the $35,000 Short range model 3 will never happen, and will be priced higher, I’d think that would hurt Elon’s pride, as this is the most direct promise he’s ever made. Being late with a vehicle is one thing… Fibbing about the price is another.

His pride survived missing the $89k Roadster price by 20 grand and never really producing the promised $57,400 Model S 40.

It’s true the Model 3 is more about price than those others, though. I guarantee they’ll sell some 35k cars. But then they’ll come up with an “improved” version that has “the features everyone wants” for $38-39k.

I think the one point you may be missing from the SR margin is the interior or what changes when there is no Premium option. We don’t know what Tesla will include in the 35K car as far as fit or finish. It seems clear to me that once a lesser interior is chosen the margin improves. We can assume the sound system will be much less and the quality of the seats should be less. Will they save on going to a metal roof? Pair these saving to a smaller battery pack, cheaper battery production from the new machines (and time as battery cost just keep going down) and the improvement Tesla has made with the production line over time and I cant see what the SR Model 3 can’t have a healthy 15% margin. Beside how many of us have purchased the base options of a car in the past? I never have. My first car was a Civic and I paid for many upgrade. This should also help the margins on average.

Does anybody know where the 115/kWh comes from? I can’t find any source online for that, only a comment from Elon in June that the Grohmann machines should brink the price down to under 100kWh by year’s end.

It could be 80, it could be less. Based on repeated promises about Semi pricing, I think Tesla can see a path to cell cost below 70 that arises from going from 2k cars/week last year to likely 9k this year

I think you’ve conflated two different statements. Elon said that Tesla hopes to get the cell price down to under $100/kWh by the end of the year, but they don’t hope to get the pack level price down that far until, as I recall, 2020.

It is the price Audi said it will pay for its batteries. Now inflated by 10% by LG.

Many people pay for base all the time. Some don’t like leather/pleather, some just can’t make the payment on a higher optioned vehicle. Maybe because Tesla Brand attracts higher end customers you feel most people in that bracket buy the upgrades, but a lot of people who want a good, affordable EV will want the base version, especially if they think they will put it on the Tesla Mobility network. Of course to do that, they will have to buy some upgrades, but saving $5k on Premium Interior may help with getting FSD

True “base” models are pretty rare. You see 20 Camry LEs for every CE. But the only costs 5-10% more. With discounts it’s often less than CE MSRP.

Tesla’s pricing strategy is nothing like that.

I think you’ve misunderstood elons tweet. “Get it done now instead of February” means “get the mid range model done now instead of waiting to be able to get the standard range model done in February.”

I think mid range is just a model they sell now before they get the standard range model with a new standard range model pack out on the streats.

I also think you’ve missed the fact that the premium interior costs Tesla money. When they start selling the standard range model for 35 k it will be without premium.

I highly doubt the SR will just be a software limited MR pack or an MR pack with fewer cells.

It’s clear to me from the Elon quote that the SR is a new pack design with fewer cells and a different architecture, different from the LR or MR designs.

Yes. SR will use the new pack design. LR will, too. MR will go away.

Also missing from the analysis is that as volumes go up the per unit cost to Tesla goes down. Which is why Elon has said they can’t make the base version until 3-6 months after they get to a steady 5k/wk and climbing.

So, while we may see the SR pack in the first part of next year, it will likely still have the premium interior. And maybe SR+AWD+PUP.

True base version ($36k) won’t be seen, IMO, until, mid 2019.

Tesla has clearly stated that the mid range pack is a temporary, depopulated Long Range pack while Short range pack is an as yet not mass produced smaller and newer pack design which saves money at pack level of the order of ~20 dollars per kWh. The rise in volumes would have a greater impact on the gross margins by decreasing the fixed cost component. Losses from decreased ASP would be more than offset by higher volume. Tesla is waiting for a combination of higher efficiency, Lowe cost pack and higher volume to make the base model profitable. My estimate is that they have to make at least 7000 Model 3 per week to make it profitable.

It has been stated that the Grohmann lines will cut cost and time to a third of present:

“Grohmann Engineering will help module production become three times faster, and three times cheaper, according to Viecha. Their new system will be sent to the Gigafactory by the end of Q3 or beginning of Q4. The Grohmann machine will be in Zones 1, 2, 3, and Tesla will be receiving three machines. The process was designed to alleviate the previous bottleneck in module production which delayed Model 3 production significantly. The machine is already built, and points to the advantage Tesla will have in building future Gigafactories. They have learned many painful lessons, but have a solid blueprint for porting the factory across the world.”
Martin Viecha – Tesla head of investor relations.

There are some important factors which will validate SR model 3.

The modules of batteries saved can be used as powerwall which is highly profitable, selling at $437/kwh. Those margin are at 200%+.
The second factor is they don’t actually need to make that much money on low ends models. Let me explain.
SR model 3s are a mission statement, advertising slogan. If more people drives Tesla it boost their market shares as well as a living advertisement everywhere. These cars are a beast and anyone drove it knows.

Tesla’s AP system is also a piggy bank for them. If the 35k models are upgrading with 8-12k with software those margins are instantly at 20%+

So Tesla will definitely release SR model 3s as it will benefit them greatly.

I dunno why but I sense they are working on some higher charge densities for their battery packs… 5% or 10% for the same weight will make alot of difference only by tweaking chemistries, then software a bit. I think they are on their third iteration since startup….maybe 4 is on the horizon. Competition and proprietary tehno would preclude a lot of talk about this.. it would just quietly appear and rags like this one would quickly pick up on the changes and improvements…