Battery cost / improvements for Model 3

Hogfighter said:

Okay, so we all have heard Elon say a number of times that the cost of the battery will decrease by 30% due purely to the economics of scale / logistics by constructing the gigafactory (reference 3Q 2014 conference call):

Elon Musk: Yes, to be precise about our prediction was that we felt comfortable with at least a 30% improvement in cost or reduction in cost just based on the location and economies of scale. That's without taking any technology improvements into account and we will certainly do technology improvements. If we can't get to 30% even without technology improvements, somebody should shoot us because that would be in complete defiance of economies of scale and obvious cost savings.

Then, in the other corner, we have JB Straubel saying that the chemistry improvements to the battery will give the Model 3 20-30% improvement in energy density over the Model S. This is not the same 30% that Elon is talking about....this is chemistry improvements. Reference around the 16:00 mark.

http://insideevs.com/tesla-cto-jb-straubel-discusses-electric-cars-video/

He has stated this before, in Sept. of 2013. The reference is below the comments:


Tesla Chief Technical Officer JB Straubel says the company's battery costs are half or even a quarter of the price of the industry average, partly because of the company's strategy to use thousands of commodity battery casings rather than the specialized batteries that GM and Nissan use.

"The battery prices in the Model S are substantially lower than what everyone expects today," he said in an interview. Mr. Straubel expects the energy density in Tesla's batteries will increase by more than 20% by the time Tesla's mass-market car comes out in about four years. More energy in batteries should equate to longer driving range for roughly the same price.

http://www.wsj.com/articles/SB10001424127887323981304579079492902482638

My point is this: We are looking at around a 50% cost reduction for the battery packs in 2-4 years. That's absolutely a game-changer, and will create an even bigger moat for the competition to cross.

Click to expand...

The improvements from chemistry are not assured. There is certainly a historical trend of cost/kWh getting better over time. But, just saying here, there might well be a max theoretical density that they start to approach as an asymptote. Or, it will come in non-linear jumps as they find some new additive mixture that increases density by 15% and that is the total improvement over a 2-3 year period. We might be in a multi-year dry spell where no new improvement has been identified or has not been qualified for production.

Consider that we are not yet seeing the much anticipated 100kWh model S. The 70D seems to be more cells not better cells.

There is a lot of time between now and GF output. And, we should be getting the latest version of the battery making machines. The GF will amount to the largest order of battery making equipment in the history of the industry, so I would hope we do have some nice improvement on deck. But it should be identified as a risk. There is no smooth "7% a year" curve to count on forever.

On average 8% improvement per year has been mentioned by a few including Carlos Ghosn, Elon and JB. I remember JB also mentioned that this trend is good for at least another decade.

austinEV said:

there might well be a max theoretical density that they start to approach as an asymptote. Or, it will come in non-linear jumps as they find some new additive mixture that increases density by 15% and that is the total improvement over a 2-3 year period. We might be in a multi-year dry spell where no new improvement has been identified or has not been qualified for production.

Click to expand...

Correct on both counts. While I don't think we are reaching the theoretical limit anytime soon (~1000Wh/kg for lithium-ion with more room to grow given lithium-air and lithium-sulfur vs 265Wh/kg for Model S cells), it's true improvements have been coming in a non-linear fashion. They usually involve chemistry changes over a course of a couple of years that give a big jump at one time rather than steady annual improvements.

Hogfighter said:

JB's quote from the lecture was not, 'we hope to see 20-30% improvement', or 'we'd like to see 20-30% improvement'; it was 'we are seeing 20-30% improvement for the Model 3'. JB is not one to make statements like that without it being all but assured.

And none of us have any clue what the actual battery makeup is of the 70D. Conjecture, yes...fact, no.

Click to expand...

Good, I certainly hope that is the case. I also hope we see some movement on what is actually being sold. Sometime between now and 2017 we should be offered a better battery. The announcement for April 30 is by all accounts a way to sell *lower* density cells. (which is a fine thing to do.)

Hogfighter said:

JB's quote from the lecture was not, 'we hope to see 20-30% improvement', or 'we'd like to see 20-30% improvement'; it was 'we are seeing 20-30% improvement for the Model 3'. JB is not one to make statements like that without it being all but assured.

And none of us have any clue what the actual battery makeup is of the 70D. Conjecture, yes...fact, no.

Click to expand...

I think someone mentioned here recently that they were told at their local SC that the 70D has more cells than S60, rather than higher density ones.
But then they've been known to be completely wrong about things before.

Just to be clear about the math, 30% reduction in the cost per kg combined with a 20% increase in density kWh/kg Leads to a 42% = 1 - 0.7/1.2 decrease in cost per kWh. A 50% reduction would require a 40% increase on density.

I am hopeful that Tesla has a few surprises for us.

About two months ago, after a factory tour, Andrea James said Tesla told her they have recently revised their thinking on how quickly battery costs will drop, and now expect it will be below $100/kWh by 2020. I strongly suspect it was Elon who said it, or he at least gave the nod to sharing this (as Dairmuid O'connell said, paraphrasing, you don't stick around Tesla by getting in front of Elon). In mid 2014, Elon had said he'd be disappointed if they couldn't do this within a decade.

Elon has a habit of repeating talking points. If it was him, I think there's quite a good chance we hear about this again during next Thursday's announcement, or the following week on the earnings call.

Link to article with comments from Andrea James:

http://www.streetinsider.com/Analys...ougherty+Following+Factory+Tour/10348143.html

Hogfighter said:

It's funny that the media blew this commentary off, but jumps on B of A's negative analysis. It's impressive that Tesla's battery cost will fall by (close to) half in the next 4-5 years, without any huge leap in technology.

Click to expand...

Very much agree. Often attempts are made to tar us longs as naive "fan boys" out of touch of reality by detractors of the company and stock implying they are sober authorities. I guess, in a sense, we're being payed for tolerating some buzzaro world events : )

How is it possible that 70D and 85D have the same GM?

I want to explore the possibility that both the 85D and 70D have the same GM. The key differences in the two models is in size of battery pack and higher power. Elsewhere we have discussed the possibility that Tesla has a more dense cell, maybe, but that is not enough to make the GM the same. What is needed is a cheaper battery pack. So let's make a few assumption and see if we can back into reasonable costs.
Assumption 1, before options both models have a 25% GM, on prices $85000 and $75000 for 85D and 70D, respectively.
Assumption 2, The 85D battery pack has cost $225 per kWh vs. $175 for the 70D.

So assumption 1 implies a total cost of $63750 for the 85D and $56250 for the 70D.
Assumption 2 implies pack costs of $19125 for the 85D and $12250 for the 70D.
Thus, the cost excluding packs is $44625 for the 85D and $44000 for the 70D. The difference of $650 could easily be the cost of more powerful motors and perhaps inverters.

So is it plausible that Tesla has some advancement that improves the cost from $225 per kWh to $175? A density improvement of about 17% would not suffice to reduce cost this much. We also know that Tesla has been considering a larger format cell which could save money on casing and manufacturing. A larger format battery could fit 70 kWh into the pack, but perhaps not 85 kWh, which would explain why a more expensive cell might still be needed for the 85D. This is all conjecture, but the 70D is still a mystery.

jhm said:

I want to explore the possibility that both the 85D and 70D have the same GM. The key differences in the two models is in size of battery pack and higher power. Elsewhere we have discussed the possibility that Tesla has a more dense cell, maybe, but that is not enough to make the GM the same. What is needed is a cheaper battery pack. So let's make a few assumption and see if we can back into reasonable costs.
Assumption 1, before options both models have a 25% GM, on prices $85000 and $75000 for 85D and 70D, respectively.
Assumption 2, The 85D battery pack has cost $225 per kWh vs. $175 for the 70D.

So assumption 1 implies a total cost of $63750 for the 85D and $56250 for the 70D.
Assumption 2 implies pack costs of $19125 for the 85D and $12250 for the 70D.
Thus, the cost excluding packs is $44625 for the 85D and $44000 for the 70D. The difference of $650 could easily be the cost of more powerful motors and perhaps inverters.

So is it plausible that Tesla has some advancement that improves the cost from $225 per kWh to $175? A density improvement of about 17% would not suffice to reduce cost this much. We also know that Tesla has been considering a larger format cell which could save money on casing and manufacturing. A larger format battery could fit 70 kWh into the pack, but perhaps not 85 kWh, which would explain why a more expensive cell might still be needed for the 85D. This is all conjecture, but the 70D is still a mystery.

Click to expand...

Excuse me for being out of the loop here - how do you know both models have the same GM?
Also, there is so much speculation around the idea that the model 70D has a different cell, perhaps someone on the forum could check this?

From what I understand the generation of the battery is written somewhere on the car, and current 85D is gen E. So, if this is true any 70D could check if the battery is gen E or something newer. Without this basic information it is just a lot of speculation.

jhm, they reshuffled the pricing of various options. I think it's rather likely that they looked at their data on various option take rates and tried to put together a new pricing scenario which they expected would recover some margin (obviously, some guess work on how new pricing would effect consumer behavior).

it's a complex situation. let's hope on the call a couple of weeks out we get some clarity on their expectations of margins going forward and their level of confidence hitting these targets. as to a change in chemistry, they may not comment on this. if there is a new battery chemistry, it's not hard to see that it would be going to go into the bigger battery soon, so they are unlikely to acknowledge a new battery chemistry until they think it's the right time to let it be known the big battery is either getting bigger or cheaper.

daniel Ox9EFD said:

Excuse me for being out of the loop here - how do you know both models have the same GM?
Also, there is so much speculation around the idea that the model 70D has a different cell, perhaps someone on the forum could check this?

From what I understand the generation of the battery is written somewhere on the car, and current 85D is gen E. So, if this is true any 70D could check if the battery is gen E or something newer. Without this basic information it is just a lot of speculation.

Click to expand...

It's an assumption as stated, and the result is conjecture as stated. To my knowledge, no one has been able to confirm that the same cell chemistry and format are being used. So the same cell hypothesis is just another conjecture. What that conjecture does not explain is why Tesla would add 10 kWh plus D and Supercharger access and only raise the price by $5000. To do so would compress margins, which does not explain why Tesla at this time would push this product.

We know however that Tesla has a favorable view of using a slight larger format cell, 33% more volume. This may have favorable economics even using the same chemistry. Tesla has not indicated when or how they would deploy this new format. They have reasons for holding their cards close to their chest, and if a gain like this is in the cards, they have reasons to play it. So in all this we'll need to wait for clarification from Tesla.

daniel Ox9EFD said:

Excuse me for being out of the loop here - how do you know both models have the same GM?
Also, there is so much speculation around the idea that the model 70D has a different cell, perhaps someone on the forum could check this?

From what I understand the generation of the battery is written somewhere on the car, and current 85D is gen E. So, if this is true any 70D could check if the battery is gen E or something newer. Without this basic information it is just a lot of speculation.

Click to expand...

I think the battery pack would be new (maybe "F") regardless. Either it uses different cell tech (thats a revision) or more cells of the old type (also a revision). So the packs rev wouldn't tell you anything as far as I can tell.

austinEV said:

I think the battery pack would be new (maybe "F") regardless. Either it uses different cell tech (thats a revision) or more cells of the old type (also a revision). So the packs rev wouldn't tell you anything as far as I can tell.

Click to expand...

Do you know that the 60 and 85 had different revs? I would be hoping they did not since that would indicate the battery rev is for chemistry, not configuration.
Also, as a non Tesla owner (small investor)- is it true that you could see the battery rev, without tearing it apart? I got that from another thread on the site, not sure I understood correctly.
Thanks.

Battery improvements may drive down the costs of existing cells, so Tesla may feel the cost savings are more beneficial than installing the newest cells.

JRP3 said:

Battery improvements may drive down the costs of existing cells, so Tesla may feel the cost savings are more beneficial than installing the newest cells.

Click to expand...

That's right the improvements could be in any dimension, cost, energy density, power density, size, chargeability, safety, etc., so it can be very hard to know for sure when something is improved. What I have tried to do with my specific conjecture is show that it is plausible that the 70D and 85D have the same margin. People have too easily assuming that just because one version costs more than another that it must have a higher margin. But if there are cost differences in the cells sourced by each version, then we cannot be certain that the 85D has a higher margin than the 70D. Of course, if one cell is cheaper than another, you also need some rationale for why not to deploy that to all versions. So here we have two differences: spare volume in the 70D pack and higher performance in the 85D. So a cheaper cell that is physically bigger or fails to deliver the power the 85D needs to do 0-60 in 4.4s would not be suitable for the 85D.

Is it possible that the 85D uses newer cells that are able to deliver more power? When the 85D first came out it could do 0-60 in 5.2s, but newer ones take this down to 4.4. Some of this may be software, but the newer P85D is not quicker than the original P85D. So were does this new burst of power come from in the 85D? Is it possible that there is a new higher performance inverter? Is it plausible that some sort of cell improvement could deliver more power with the same inverter? Or is the performance gain due to higher performance motors? It seems there are four factors to consider: software, battery, inverter and motors. We know that software and motors have been improved. Is this this sufficient to explain how the 85D got quicker, delivering 15% more power? These are not rhetorical questions. I really am not sure how Tesla has pulled this off. If a higher performance inverter is now included in the 85D, then that would add cost to the version and reduce its GM. Another hypothesis is that 85D now uses a new higher performance cell. So the 70D might simply use older, cheaper cells while the 85D and.P85D are the early recipients of more costly higher performance cells. This theory is also consistent with a pack price difference of, say, $225 vs $175 per kWh. So we just don't know which cells are newer, but there are plausible rationales for using cells of different cost in the various versions.