How many kWh can they squeeze into the Model 3...?

[googlepeakoil]

We do need to remember that improvements in nearly all the energy usage factors (e.g. inverters, circuitry, vehicle control, battery management, heating and cooling, lighting) will yield very substantial improvements to effective range. Motors may see some improvement too, probably less significant, but if DC motors can be made to be practical, the improvements even there may be notable.

In other words kilowatt-hours is not the whole story, albeit it is one of the most important parts of the story.

The electricity supply from battery to motor is supposed to be about 90% efficient. So i don't see any even small improvements beyond model 3 - which might improve these things by at most 2-3%. Their 3rd iteration of the motor/battery/electrics. That's different to petrol / diesel engines that are about 30% efficient (lots of waste heat and more noise).

 

[Red Sage]

But why would you think that the exponential growth of your computer's RAM capacity would say anything about the growth of energy density in batteries?

Until now, the battery energy density has been improving exponentially, at about 7% annually.

For someone else's above unsubstantiated guess 8 years into the future, that works out to about 72%, meaning that a current state of the art battery of 90 kWh would improve to about 155 kWh, i.e. far below the guessed range of 275 kWh to 350 kWh, which would require the annual growth rate to improve to between 15% and 18.5%.

Per Elon Musk, most start-up's claims regarding improvements to battery technology are utter BS, so I would be very careful about assuming a vastly improved growth rate.

But in 8 years I will probably have nothing better to do than to follow this forum, so let's see.

It's simple really. I pay very close attention to what JB Straubel says. I also have done very good research into the battery cells that Tesla Motors has used thus far.

The battery cells that are used in the Tesla Model S were officially chosen in late 2010, after being unveiled by Panasonic in late 2009. Those were used in cars that were released in mid 2012. In the roughly 18 months or so after Tesla Motors chose Panasonic's 18650 battery cells, they proposed changes to those cells, including removing the integrated chips that acted as a fuse for each battery. Tesla Motors chose to have an external battery management system instead that controlled all the battery cells in an array to ensure they were monitored fully for charging and discharging. Tesla Motors also chose an active liquid cooling system to maintain a constant optimal temperature range for the various battery modules that formed the battery pack.

JB Straubel has noted there was about a 40% improvement in energy density between the battery cells used in the Tesla Roadster in 2008, and those used in the Model S in 2012. More recently, he has stated that he expects to see a similar 40% improvement in energy density of battery cells used in Model ☰ in 2017 as compared to those used in Model S in 2012. No, this is no Moore's Law... But within the framework of a couple of decades, it is more than enough to make sure electric cars can become very viable much sooner than many expect.

Please take the time to watch some videos online that feature the Chief Technology Officer at Tesla Motors, JB Straubel. He explains that the energy density of battery cells typically doubles every ten years or so. Thus, the amount of battery cells needed for a given capacity drops over time. Which means the weight goes down for that capacity, increasing range. Likewise, for a given weight, there will be more energy stored, also increasing range. He also believes it will be possible to increase charging speeds by a wide margin, and likely within the next five years.

JB Straubel | Energy@Stanford & SLAC 2013, September 13th, 2013

JB Straubel & Ira Ehrenpreis | The Tesla Story, October 8, 2013

2014 Energy Storage Symposium - JB Straubel's Keynote, May 21, 2014

Tesla Model 3 Tesla Motors JB Straubel, October 11, 2015

JB STRAUBEL ON HOW TO SPEED INNOVATION LIKE TESLA - CO-OP THINK, July 15, 2015

I'm sure there was another video I wanted to direct you to, but I'm having trouble finding it just now.

 

[Red Sage]

The current model S is not using anywhere close to the limit. There are at least two other systems zombie 222 and flux capacitor that are using greater than 2000 amps from a smaller battery pack, both can accelerate faster than the P90DL .

Yeah. They also have a much shorter range. Tesla Motors concentrates on providing safety, fun, and range. You can perhaps have a bit more fun for a bit less time, with much shorter range. I'd rather the 250+ miles of range than the barely more than 50 miles variety for a street/highway capable car.

 

[JeffK]

Yeah. They also have a much shorter range. Tesla Motors concentrates on providing safety, fun, and range. You can perhaps have a bit more fun for a bit less time, with much shorter range. I'd rather the 250+ miles of range than the barely more than 50 miles variety for a street/highway capable car.

Exactly! It's just whenever someone says you NEED a larger capacity battery to offer more power they're often misinformed. More power with the same range then sure, you'd need a higher capacity battery.

 

[Red Sage]

This is just an aside to your argument about the impact of weight, but.... the RAV4 EV 's EPA range estimate wasn't actually tested (just roughly calculated) and it was adjusted down due to the normal vs. extended charge modes. In reality, you get over the official 103 mile EPA number for a normal charge and significantly more with extended charge (like 130 miles and up, when new). Here's a set of charts for highway usage:
Toyota Rav4 EV Forum • View topic - Range Chart

As a RAV4 EV owner, it's a little sad to see that 103 EPA number used widely when it's easy to do 120+ miles at 70 mph even after 25K miles, but that's the only official number we have. Would've been interesting to know what the true EPA tested number would have been. It also would be interesting to know what the CdA (drag x area) came out to. (Keep in mind that the RAV4 EV had one of the lowest drag coefficients for SUV's at 0.30, but I can't find a frontal area figure for it.)

RAV4 EV: CdA = 0.30 * ?? = ?? m2
Model X: CdA = 0.24 * 2.59 m2 = 0.622 m2
Model S: CdA = 0.24 * 2.34 m2 = 0.562 m2

Here's Toyota's video about improving air drag on the RAV4 EV:

Cool. I use EPA numbers because they are 'calculated' the same way for all vehicles. Supposedly. That is, they are based upon 55% Stop-and-Go, and 45% Highway driving, at a variety of different speeds, over the course of five cycles. I think most cars, whether ICE driven or fully electric, do pretty well at a constant speed of 70 MPH or less on the highway. That's why the highway MPG/MPGe is typically higher than the combined MPG/MPGe that the range is calculated from. And, I had seen the video before, but for some reason I thought the coefficient of drag for the Toyota RAV4 EV was 0.31, not 0.30 -- though still the best for any SUV available at the time -- it could be I remembered incorrectly.

 

[Red Sage]

Excuse me, but the efficiency of the different parts you mention are already around 90% or better, so any improvement there will be negligible compared to an annual improvement rate of 7%. You cannot double the capacity of something every 10 years, when it is already at 90% of its theoretical limit.

This is the reason why Elon Musk talks about the importance of the machine that builds the machine:

For a given effort you can only achieve tiny improvements in the components you mention, but get significant improvements in the manufacturing efficiency, which is the key to knocking down the price, for the whole vehicle and for the battery.

It sounds as if you are speaking of two or three different things. Even if the machinery that surrounds the battery cells themselves within a car maintain the same amount of efficiency, from cooling systems, to power electronics, to inverter systems, to drive systems...? Increasing the available capacity of the battery pack will still improve range.

Remember, the battery cells used in the Tesla Model S and were originally announced by P in late 2009, and chosen by Tesla Motors in late 2010. It was another 18 months or so before a modified version of those battery cells appeared in the Model S.

Per JB Straubel, the battery cells used in the Model ☰ will represent about a 40% improvement in energy density over those initially used in the Model S. Even if the improvement in battery technology only allowed a 7% improvement per year, in the five years between 2011 and 2016... That means the volume of battery cell storage that would have held 60 kWh in 2011 would hold 84 kWh today. Oh, wait... That just happens to be a 40% improvement.

Likewise, you could use a volume that would have only held around 43 kWh in 2011 to hold 60 kWh today. And 100 kWh could be stored in the space that would have held roughly 71 kWh in 2011. Or, about 84% of the space that would have held 85 kWh back then.

Improvements made in 'The Machine that Makes The MACHINE' will lower costs of manufacture, and improve efficiency of output, so that it takes less money to build more product. And that is the reason why Elon Musk stresses the advancement of Economies of Scale. A wide variety of constant, incremental improvement will benefit both affordability and profitability in the long run.

 

[Red Sage]

Except JB has said batteries are improving at an aprox 5% a year - while CPU (and ram density) was doubling every 18 months. There was a big jump with the move to lithium - but there's only incremental improvements. I'd bet money that in a decade battery density won't have even doubled. You might see 50% - but remember lithium technology has been around 30 years. My first Sony camcorder had a tiny lightweight lithium battery that could run the digital-8 tape for 2-3 hours. I'd be surprised if Panasonic / Tesla make even a 20% improvement on density with going to their 2170 (they're not calling it 21700). Nissan Leaf + Volt -> Bolt are making huge improvements because their format was very low density compared with Tesla's 18650.
75kwh is a huge amount of electricity - enough to power the electrical appliances in my house for 1-2 weeks. You could run a 100w 60 inch tv for 750 hours! These aren't small amounts of electricity currently.

I believe that JB Straubel has said that there is an average improvement of 7% to 9% per year. He has admitted that as an average, that means a low year might be around a 5% improvement, while a high year might be closer to 12%, but that the trend still worked out to 7% to 9%. When you consider that between the 2007 unveiling of Tesla Roadster, and 2012 release of Model S, JB observed a 40% improvement in energy density... And that between what he saw in 2012 and what he knows will be in the 2017 Model ☰ he has seen another 40% improvement already... That pretty much exactly works out to 7% per year on average as an improvement in energy density. As a result, the volume of battery cells that used to hold only 53 kWh in 2007 may now hold closer to 104-to-119 kWh in 2017.

 

[Red Sage]

Tesla projection of 8% improvement in energy density per year is too optimistic. Historically battery energy density improvements per year have been frustratingly small. Thats why your computer never has enough juice.

Any comparison with the computer industry shows a complete lack of perspective. We are not talking about the extrodinary improvements that have been achieved in the hardware of computers. Batteries are a totally different animal.

3% per year is the historical rate of improvement. Thus Red Sages projections are much too optimistic and ignore the historical rate completley.

I historically ignore anything that hasn't been directly observed and reported by JB Straubel. He stated way back in September 2013 that he saw no reason why Lithium-Ion battery cells would not continue to improve in energy density, effectively doubling each decade, for the next two decades. And he said that perhaps by then something even better would appear. But for now, they have to use what they have in a practical fashion, to the best of their ability.

That is certainly better than the ridiculous metric of just sitting around and 'waiting' until such time as someone creates a battery cell with 'the same energy density as gasoline'. That concept is utterly ridiculous, because a high performance electric vehicle drivetrain can still be 90% energy efficient, while the wimpiest internal combustion engine might be able to achieve 36% energy efficiency at best. Thus, since 33.6 kWh of energy is stored in a gallon of gasoline, and a Prius can travel 50 miles, using maybe 12.1 kWh of that energy (with the rest wasted on heat, light, and sound) ... That means that an electric vehicle would use 30.2 kWh of a provided 33.6 kWh of energy for motivation, and that's why it would likely travel twice as far, 100 miles, using it.

The fact of the matter is that before electric vehicle have even the energy equivalent of five gallons of gasoline, they will be exceeding the range of pure ICE vehicles in the same market segment. I figure somewhere in the 3-to-4 gallon equivalent range... Perhaps 100 kWh to 135 kWh or so, it will be entirely over for the ICE industry with regard to passenger cars. And at the 170 kWh or higher mark, ICE vehicles will be obsolete for the vast remainder of ground transport applications.

 

[Red Sage]

This is true as long as one only looks at one part at a time, but the whole drive train consists of many ~90% efficient parts, connected in series:
- extraction of energy out of the battery: 10% goes into heating the battery, 90% to the controler
- inverter: 10% goes into heating the inverter, 90% to the motor
- motor: 10% goes into heating the stator and rotor, 90% goes into rotation
- reduction gears and differential: 2% goes into heat, 98% into torque on the output shafts
- half-shaft and wheel bearings: 2% goes into heat, 98% goes "through"
- brake-drag: 2% goes into heat, 98% goes "through"
- tire-drag: 10% goes into heat, 90% goes "through"
- auxiliary systems: 10% goes into heat, 90% goes into useful work

Feel free to replace efficiencies of parts with more accurate numbers.
Above chain of very efficient parts results in 55% overall efficiency. Replace all parts with others that have only half the losses and overall efficiency rises to 75% - one third improvement of range with same battery capacity, say 300 miles instead of just 220.
Efficiency is "system metrics".

Interesting. I don't believe it works that way though. I believe the energy efficiency rating for battery electric vehicles is measured in terms of the work that is done with the energy provided. Thus, the 85%, 90% or 95% efficient statements are for the entire system in concert. The same as the 8%, 12%, 20% and 36% efficient ratings for ICE vehicles are based upon the entire system from beginning to end. That's why a Model S has twice the rated MPGe of a Prius on MPG.

 

[vinnie97]

Like Journey, I hope Elon never stops squeezing.

 

[modamoda]

I didn't quite understand your post, do you mean that largest battery option with the 3 would cover 480 miles with a full charged battery? As much as I'd like that, I don't think that's possible with the timeline that's current in play.

Yeah, that's why I said "radical".
We don't know the efficiency(?km/kWh) of M3 yet, but since M3 is smaller than MS, and will use advance battery layout(so called 21-70), there could be "quantum jump" of range.

it's prediction just for fun.

 

[Incredulocious]

Cool. I use EPA numbers because they are 'calculated' the same way for all vehicles. Supposedly. That is, they are based upon 55% Stop-and-Go, and 45% Highway driving, at a variety of different speeds, over the course of five cycles. I think most cars, whether ICE driven or fully electric, do pretty well at a constant speed of 70 MPH or less on the highway. That's why the highway MPG/MPGe is typically higher than the combined MPG/MPGe that the range is calculated from.

I think you misunderstood me. I'm saying that I recall that the EPA range value for the RAV4 EV was merely estimated/calculated (due to an exemption for low volume vehicles), rather than tested as most other vehicles are. Also, the calculation took the range estimate for a normal, incomplete charge and split the difference with the estimate for a full, extended charge. So the EPA number for the RAV4 isn't even estimated against a full charge. (They did this for the Nissan LEAF as well. Many of us wonder if this is why Nissan then dropped the feature allowing automatic charging to just 80%, thus giving them a higher EPA number.) I don't know how the EPA number for the limited Model S 40 was figured, or whether there's an adjustment to not count a full range charge for Tesla's other vehicles. I think I remember that the EPA doesn't reduce the range estimate on Tesla vehicles since the owner is free to set any limit on charge level, isn't that right?

I used the example of driving the highway just to illustrate what is possible with the vehicle. We RAV4 EV owners also get much more than the EPA estimate for mixed city/highway driving, but of course we aren't able or trying to simulate how the normal EPA test is done. (I find I typically get very close or even slightly better than EPA estimates for city/highway for most vehicles I've driven.)

Lastly, unlike ICE-based engines, it's common for hybrid and electric vehicles to do better in the city test vs. highway.

In summary, the 103 EPA range number is a poor estimate for the RAV4 EV.

 

[Red Sage]

I think you misunderstood me. I'm saying that I recall that the EPA range value for the RAV4 EV was merely estimated/calculated (due to an exemption for low volume vehicles), rather than tested as most other vehicles are. Also, the calculation took the range estimate for a normal, incomplete charge and split the difference with the estimate for a full, extended charge. So the EPA number for the RAV4 isn't even estimated against a full charge. (They did this for the Nissan LEAF as well. Many of us wonder if this is why Nissan then dropped the feature allowing automatic charging to just 80%, thus giving them a higher EPA number.) I don't know how the EPA number for the limited Model S 40 was figured, or whether there's an adjustment to not count a full range charge for Tesla's other vehicles. I think I remember that the EPA doesn't reduce the range estimate on Tesla vehicles since the owner is free to set any limit on charge level, isn't that right?

I used the example of driving the highway just to illustrate what is possible with the vehicle. We RAV4 EV owners also get much more than the EPA estimate for mixed city/highway driving, but of course we aren't able or trying to simulate how the normal EPA test is done. (I find I typically get very close or even slightly better than EPA estimates for city/highway for most vehicles I've driven.)

Lastly, unlike ICE-based engines, it's common for hybrid and electric vehicles to do better in the city test vs. highway.

In summary, the 103 EPA range number is a poor estimate for the RAV4 EV.

Here the thing is... When Tesla Motors first projected range for the Model S, they did so presuming constant speed at 55 MPH. That was sufficient, they thought, because part of the EPA test -- which was only two cycles before -- was conducted at a constant 55 MPH. So the respective ranges were thought to be 320 miles for Model S 85, 230 miles for Model S 60, and 180 miles for Model S 40. That wasn't posted for very long. Very soon after, they lowered those estimates to 300 miles, 225 miles, and 160 miles, respectively. But, the final, official EPA numbers turned out to be 265 miles, 208 miles, and 139 miles instead.

Each of those represents a rather large dropoff from what Tesla Motors/Elon Musk expected to see. I believe they felt slightly betrayed by the EPA switching to a 5-Cycle range test. So yeah, each car is capable of achieving greater range than the EPA numbers, with carefully controlled judicious application of the GO pedal. Thus, I am not at all surprised that the RAV4 EV may achieve better results than the official EPA rating. My point is simply that its official EPA range rating is less than that of the Tesla Model S 40.

There is a reason for that. I believe it is at least twofold. Where the 2013 Toyota RAV4 EV uses a 50 kWh battery pack, of which 41.8 kWh is available for use, the 2013 Tesla Model S 40 uses a 60 kWh battery pack, of which 40 kWh is available for use. The RAV4 EV has a 154 HP motor rating, and the Model S 40 has a 302 HP motor. The RAV4 EV is front wheel drive, the Model S 40 is rear wheel drive. For those thinking in terms of ICE drivetrains, each of those is considered an advantage in favor of the RAV4 EV. More available capacity, lower power to suck away juice, and driving the front wheels all improve the efficiency of ICE -- so that should transfer to EV as well -- right? Wrong. Finally, the sole advantage that most perceived going to Model S 40 was its 0.24 coefficient of drag, as compared to the RAV4 EV having a 0.30 instead. But that advantage was only realized due to the 5-cycle test, as opposed to the previous 2-cycle test.

The fact of the matter is that EPA testing is largely not conducted by the EPA. The tests are performed by the manufacturers themselves, to specifications chosen by the EPA, using calculations that the EPA provides. The EPA will occasionally choose a car at random to double check its actual range and fuel economy rating, just to keep automobile manufacturers honest. That is what happened to Hyundai a few years ago, when it was determined they had flat out lied about the fuel economy for Elantra and Sonata, and had to pay retribution as a result. Because of this 5-cycle testing method, none of the initial Model S variants achieved the range that was originally hoped for. Instead, they got roughly 90% of their expected result. I expect that is similar to RAV4 EV, given your protests, because it uses a Tesla Motors drivetrain. But in any case, my point holds true -- the weaker motor and lower weight of the RAV4 EV does not translate into greater range than the Model S 40 -- despite other 'advantages'.

And this is why I rail against people who continually claim that the base version of the Model ☰ will have a 50 kWh, 45 kWh, or 40 kWh battery pack capacity -- but still achieve a 215 mile EPA rated range. Supposedly because of improved power electronics, improved coefficient of drag, and lower power motors. That is utterly ridiculous, given the evidence. However, it may still achieve a 215 mile range with perhaps a 55 kWh usable capacity, in a rear wheel drive car that has a software limited 75 kWh battery pack.

Here's some napkin math:

208 divided by 5 equals 41.6 ...
41.6 times 6 equals 249.6 ...
249.6 times 0.9 equals 224.64 ...
224.64 times 0.91666~ equals 205.92​

That is a linear progression from the original Model S 60's range to that of a vehicle 80% of the mass using about 55 kWh instead. So, it is probably more than a little bit inaccurate, for various reasons. But it does get you at least 205+ miles of range. I figure various improvements over time in the technology used is what would get that 'extra' ten miles. Maybe.

But what do I know? I'm just some guy on the internet.

 

[Swampgator]

Red,
A new S60 has a 210 mile range and weighs 4600#
The base M3 will weigh 1000 # less. What does your napkin calculator have to say about that?

 

[Red Sage]

Red,
A new S60 has a 210 mile range and weighs 4600#
The base M3 will weigh 1000 # less. What does your napkin calculator have to say about that?

Much the same. A higher capacity battery pack, limited to 60 kWh usable, allows for greater range than an actual 60 kWh battery pack did.

 

[EaglesPDX]

Red, A new S60 has a 210 mile range and weighs 4600# The base M3 will weigh 1000 # less. What does your napkin calculator have to say about that?

215 miles per charge is what Tesla says about that. 258 per charge if T3 follows TS60.

 

[vinnie97]

Probably (another) a wasted attempt but here goes: There's T3 (again, Ahnold?!) and TS60...if you must use acros, might as well incorporate M instead of T since Tesla is understood. M3, MS60. TM3 or TMS60 if you have to reference the understood manufacturer. *puts down nitpicking hat*

 

[EaglesPDX]

M3, MS60. TM3 or TMS60 if you have to reference the understood manufacturer. *puts down nitpicking hat*

Elon is going to announce a jargon abbreviation change at the next reveal so if you want to be cool and get ahead of the curve, start using T3, TS, TX.

 

[SuperOmega]

Interesting read via Electrek about yesterday's media walkthrough at the Gigafactory....

Tesla’s plan for Gigafactory vehicle battery pack rollout and why it matters
...
...
...
With the improvement in drag coefficient, and the loss of a few hundred pounds here and there..... I really think they can get to 300mi on the "max range" Model 3.

Did not see this until today. Thanks.

 

[jbcarioca]

What is the appeal of 48v?

At the simplest, the higher the voltage the lower the required amperage, so higher voltage permits higher loads. That is why long range electricity transmission is at very high voltage. A quick Google on the subject will yield more technical explanations.