Investor Engineering Discussions

[Doggydogworld]

Could a heat pump based electric oven have a performance edge over an oven based on resistive heating?

Not much. Heat pump COP is a function of temperature differential. It's the reciprocal of Carnot heat engine efficiency -- a heat pump is just a perfect Carnot engine running in reverse.

Engine - use a temperature differential to create work
Heat pump - use work to create a temperature differential

COP = Th / (Th - Tc)

As you can see, COP is highest when temperature differential is low. That's why air source heat pumps do really well pumping heat from 10C (50F) outside air to ~35C or whatever internal duct temp is needed to prevent the air blowing out the vents from feeling cool. That's a 25C differential vs. 55C when outside air is -20C (-4F).

20C room temperature to 205C (roughly 400F) oven temp is a whopping 185C differential. In theory you could still have a 2.5+ COP at those temps with a perfect heat pump, but real world COPs are usually less than half of theoretical. So maybe 1.2-1.5? A 25% efficiency gain vs. a huge cost/complexity penalty for an appliance that's used an hour every couple days sounds like a terrible deal. It's hard to beat a resistance coil for simplicity.

Induction cooktops are cheap.

$65 at Home Depot -

Hamilton Beach Single Burner 10 in. Black Induction Cooktop 34104 - The Home Depot

Expand your cooking options with fast and flexible induction cooking. From entertaining and home renovations to outdoor grilling and small spaces, the Hamilton Beach Induction Cooktop goes wherever you

www.homedepot.com

Putting waste AC heat into hot water heating is a more obvious first step as those two units are frequently in the same room already.

Been done for decades. Called a "desuperheater" for reasons that escape me. Works great in the summer. Kinda useless the rest of the year. Can be worthwhile in climates where you run A/C from April to October, like Phoenix.

 

[jerry33]

$65 at Home Depot -

Hamilton Beach Single Burner 10 in. Black Induction Cooktop 34104 - The Home Depot

Expand your cooking options with fast and flexible induction cooking. From entertaining and home renovations to outdoor grilling and small spaces, the Hamilton Beach Induction Cooktop goes wherever you

www.homedepot.com

A single burner heating appliance is not a cooktop.

 

[mongo]

Been done for decades. Called a "desuperheater" for reasons that escape me. Works great in the summer. Kinda useless the rest of the year. Can be worthwhile in climates where you run A/C from April to October, like Phoenix.

Superheat is the amount above boiling the refrigerant is heated (opposite of subcool). A desuperheater reduces that excess energy by heating water.
Superheat and Subcooling Defined

 

[MC3OZ]

I suspect that this is just a market specific thing, but in Australia Induction cooktops are often 2X the price of a gas cooktop.

Many households who have gas cooktops need to migrate to induction, cost is a factor.

 

[Oil4AsphaultOnly]

A single burner heating appliance is not a cooktop.

And should be compared to a single burner resistive hot plate (which is only $15). But that single burner induction cooktop does illustrate that induction cooktops shouldn't be inherently expensive. They're probably seen as niche, so manufacturers aren't incentivized to reduce their prices to be competitive to gas/electric cooktops.

 

[jerry33]

And should be compared to a single burner resistive hot plate (which is only $15). But that single burner induction cooktop does illustrate that induction cooktops shouldn't be inherently expensive. They're probably seen as niche, so manufacturers aren't incentivized to reduce their prices to be competitive to gas/electric cooktops.

There is stronger induction on actual cooktops. Difference between 50 amps and 15 amps.

 

[Oil4AsphaultOnly]

There is stronger induction on actual cooktops. Difference between 50 amps and 15 amps.

50A is for the entire 4-burner cooktop, so less than 15A for each, but at 240v instead of 120v. So your point about higher power still stands, just not as dramatic as the amps would suggest.

 

[MC3OZ]

I know, we've seen them testing an Air at Freemont

It's just for the comparison I did it doesn't make sense, if we want to figure out how much more efficient a Cybertruck could be than a F150, we compare the Mach-e to the Model Y, not a Lucid with Model S

I've moved the debate to this thread because it is more suitable and I'm going to go into more detail on my thinking. I'm not claiming to be right, just stating a point of view.

We are comparing the F150, R1T and Cybertruck in the following areas:-

• Weight
• Pack energy density
• Drag
• Rolling Resistance
• Electrical efficiency.

Pack Energy Density

Tesla battery day claims a 54% range increase but 20% of that is from silicon in the anode which hasn't happened yet. I think silicon in the anode and the full 54% range increase is needed for the Roadster, but initially I would guess the Cybertruck may have a 20-30% pack energy density over the R1T which from memory using 2170s. Hard to compare to the F150 here, but at this stage I would guess 10%-15%

Weight

The F150 is body-on-frame and the R1T us unibody, generally I would expect a unibody vehicle to be lighter, but the R1T is heavier.

For the CT it has a unibody cabin, I think Tesla has an edge in price that may allow them to use more high strength steel IMO the truss-like stainless-steel sides of the CT (sails) and something to the towing strength. The weight comparison comes down to the weight of the CT sails compared to the weight of the F150 frame and side panels. If give the CT and edge here, but it is hard to know how big an edge.

There are weight savings from the heat pump cooling system but they are already in the Model Y, so no additional gains in that area.

I'm speculating that the CT may have a 48V wiring harness, brake-by-wire and steer-by-wire, if true there are significant weight reductions in those areas.

Drivetrain - I'm guessing that the CT might have a 1000 volt architecture and that this would allow smaller electric motors saving some weight.

Drag

The F150 and R1T don't have a slippery shape, the CT should have lower drag particularly with the tonneau cover closed. IMO the difference in this area is greater than the difference between a Model Y and a Mach-e. Ford has at least made some effort to reduce drag for a Mach-e.

Rolling Resistance

We know nothing about this area, but given everything else in this post it is an obvious target to Tesla to tackle.

Electrical Efficiency

Provide the speculation about 48V/100V above is correct Tesla should have lower resistive losses. Other aspects are more efficient, heat pump, FSD, but they are probably covered in the comparison between Model Y and Mach-e.

What it boils down to is, I have a bias against double-stacked packs, IMO a single-stack structural pack is a better architecture, I can't see how the second stack can make a meaningful contribution to structure, electrical connection and venting are much more problematic.

 

[MC3OZ]

For the Roadster, I assumed the prototype was a double stacked 2 x 100 kWh 18650 pack.

Applying the full battery day range increase of 54% we can round that up to a single stack 150 kWh pack.
At 3 miles for kWh that gives a range of 450 miles. With the other efficiencies mentioned for CT above that might be able to be edge that up towards 500 miles.

I've seen the Roaster prototype described as a "tire shredding machine" the combination of a heavy vehicle with lots of power put a lot of stress on many areas, but particularly the tires.

A single stacked 4680 architecture requires silicone in the anode, but the Roaster is a perfect test platform for that battery. Any warranty issues would be limited to a small number of high margin cars,

My original speculation about 46x120 came form a post made on Twitter by "Soylent Brown" get before or after battery day. It was something along the lines of "Tesla are not making the bigger battery". It was unclear if that was permanent or temporary. As I trusted the source, I tried to make sense of his comments. I concluded that a 46x120 cell might be necessary for the Roadster and the Plaid+ Model S and that it could eventually be made at Kato Rid

I now think 46x120 probably isn't happening, because 54% range increase is enough and silicon in the anode can give the desired charge/discharge rates. I also think 46x120 would probably push Roaster deliveries to 2025.

Hard to know if I was on the right track with 46x120, or I just wandered off down the wrong rabbit hole.

But given a choice between double-stacked and a single stack 46x120 pack, I would take 46x120.

IMO Tesla is always happy to trade off a bit of range for lower cost and higher efficiency.

 

[Buckminster]

https://twitter.com/x/status/1696542003968848366

 

[MC3OZ]

More on silicon anodes and the implications for the Roadster.

EXCLUSIVE: Inside the Tesla-funded Jeff Dahn Research Group

Electric Autonomy finds out what makes the Jeff Dahn Research Group so appealing to Tesla in their goal to build to world's best batteries

electricautonomy.ca

“I would say that’s probably the best lithium-ion cell you can make today,” says Metzger, pointing to the thinnest in the line of blue rectangles spread on the table.

“Silicon carbide,” reveals Metzger. “If you put silicon into your negative electrode you can increase the energy density a lot. Even in this small volume you can store the same energy as you can in this thick battery — the lithium-iron-phosphate (LFP) battery.”

The silicon battery (Electric Autonomy knows how much silicon is contained in the battery, but agreed not to report it) does not cycle — that is to say charge and discharge — as well as the LFP battery the group developed.

The first questions is whether or not this “Silicon carbide,” is the silicon process described on battery day,

Battery day:-

• Raw Metallurgical Silicon
• Stabilise surface - though elastic ion-conduction polymer coating
• Robust network - Highly elastic binder + electrode design,

Silicon carbide - Wikipedia

en.wikipedia.org

Silicon carbide (SiC), also known as carborundum (/ˌkɑːrbəˈrʌndəm/), is a hard chemical compound containing silicon and carbon. A semiconductor, it occurs in nature as the extremely rare mineral moissanite, but has been mass-produced as a powder and crystal since 1893 for use as an abrasive. Grains of silicon carbide can be bonded together by sintering to form very hard ceramics that are widely used in applications requiring high endurance, such as car brakes, car clutches and ceramic plates in bulletproof vests.

Silicon carbide is expensive but it does seem possible that Tesla could make their own Silicon Carbide which they may needed to do for volume battery manufacture.

However, it doesn't seem that Silicon Carbide is is the silicon process described on battery day.

Drew has also said that all the benefits of battery day might not be realised until 2026, but hunch would be that silicon will be one of the last benefits.

There are implications for the Roadster because perhaps neither silicon path will be fast enough for Roadster deliveries to start in 2024/2025.

However, one form of producing Silicon Carbide seems arguably better than the battery day silicon process.

Pure silicon carbide can be made by the Lely process,[20] in which SiC powder is sublimed into high-temperature species of silicon, carbon, silicon dicarbide (SiC2), and disilicon carbide (Si2C) in an argon gas ambient at 2500 °C and redeposited into flake-like single crystals,[21] sized up to 2 × 2 cm, at a slightly colder substrate. This process yields high-quality single crystals, mostly of 6H-SiC phase (because of high growth temperature).

I assume that Tesla can control this process to grow single crystals of the desired size.

This version does mention polymers, so could perhaps be the battery day process

To form complexly shaped SiC, preceramic polymers can be used as precursors which form the ceramic product through pyrolysis at temperatures in the range 1000–1100 °C.[25] Precursor materials to obtain silicon carbide in such a manner include polycarbosilanes, poly(methylsilyne) and polysilazanes.[26] Silicon carbide materials obtained through the pyrolysis of preceramic polymers are known as polymer derived ceramics or PDCs. Pyrolysis of preceramic polymers is most often conducted under an inert atmosphere at relatively low temperatures. Relative to the CVD process, the pyrolysis method is advantageous because the polymer can be formed into various shapes prior to thermalization into the ceramic

Still this is making the SiC itself into a polymer not adding a polymer coating, unless this polymer is used to coat silicon. (then converted to Silicon Carbide).

What his all means for the Roadster timeline is hard to guess, I am now not sure if silicon anodes are a mandatory requirement.

 

[MC3OZ]

Tesla Lithium Refining Process;

Inputs:

Soda Ash - Na2CO3
Lime - CaO
Spodumene - Lithium - Aluminium Silicates.
Water - H2O

Outputs -

Lithium Hydroxide - LiOH
Limestone - CaCO2 + Silicon - Si (Any residual aluminium also ends up in the mix)

The initial process is a Soda Ash leech - literally Soda Ash and water are added to the spodumene.
Sodium displaces some of the the lithium, and during this during this process, lime is turned into limestone.

We can see why the large water ponds are on the Corpus Christi site, water is needed to this initial step and perhaps some of the later steps.

After the leech the waste mix of sand and limestone exists, and the useful product for the next stage is a mixture of sodium hydroxide and lithium hydroxide in water.?

I originally added chlorine to the process here, but that was a mistake... there is no chlorine listed in the input ingredients The intermediate salts must be sodium hydroxide and lithium hydroxide?

The lithium clay extraction process consumes some sodium from the salt and the output is a mixture of sodium chloride and lithium chloride dissolved in water.

In a dry environment like Nevada it is important the capture the water for later reuse, so the salt mixture is probably dried out to just produce the mixed salts as an output product.

These mixed salts can be shipped to Corpus Christi for further refinement, if the salts need to be dissolved in water for further processing, that can be done at Corpus Christi.

Somehow the chlorine needs to be extracted from this chlorine salts produced by the clay process, I was initially wrong here, ,due to my incorrect assumption on chlorine being present in the spodumene process.

 

[Olle]

For the Roadster, I assumed the prototype was a double stacked 2 x 100 kWh 18650 pack.

Applying the full battery day range increase of 54% we can round that up to a single stack 150 kWh pack.
At 3 miles for kWh that gives a range of 450 miles. With the other efficiencies mentioned for CT above that might be able to be edge that up towards 500 miles.

I've seen the Roaster prototype described as a "tire shredding machine" the combination of a heavy vehicle with lots of power put a lot of stress on many areas, but particularly the tires.

A single stacked 4680 architecture requires silicone in the anode, but the Roaster is a perfect test platform for that battery. Any warranty issues would be limited to a small number of high margin cars,

My original speculation about 46x120 came form a post made on Twitter by "Soylent Brown" get before or after battery day. It was something along the lines of "Tesla are not making the bigger battery". It was unclear if that was permanent or temporary. As I trusted the source, I tried to make sense of his comments. I concluded that a 46x120 cell might be necessary for the Roadster and the Plaid+ Model S and that it could eventually be made at Kato Rid

I now think 46x120 probably isn't happening, because 54% range increase is enough and silicon in the anode can give the desired charge/discharge rates. I also think 46x120 would probably push Roaster deliveries to 2025.

Hard to know if I was on the right track with 46x120, or I just wandered off down the wrong rabbit hole.

But given a choice between double-stacked and a single stack 46x120 pack, I would take 46x120.

IMO Tesla is always happy to trade off a bit of range for lower cost and higher efficiency.

An alternative to double stacking or making 46*120 cells is to install the modules standing with the cells horizontal, like in the original Roadster. While I realize that this isn't compatible with Tesla's mass market packs, it might not matter in case the roadster already needs its own packs due to extreme requirements for structural strength/mass ratio and more exclusive materials. A side benefit of standing modules is that you can have carveouts for the passengers so that the roofline can be lower for less aero drag, see picture. Driver position adjustments could be accomplished with adjustable pedals and steering wheel.
Could also carve out some from the bottom of the pack toward the rear where the ground effect tunnels begin.

structural 4680 pack with standing modules

 

[MC3OZ]

Further info on Lithium Refining, melting points that may be able to be used to separate compounds,

Lithium Hydroxide - 462C, 864F
Sodium Hydroxide - 318C, 604F
Sodium Carbonate - 852C
Aluminium Hydroxide - 300C

For clays:-
Lithium Chloride - 605C-614C
Sodium Chloride - 801C

Steel melting point - 1371C-1540C (Can possibly have steel pipes?)

From Master Plan 3, heat pumps can only provide heat up to 200C.

If Tesla develops a heat storage battery Corpus Christi might be a good test site.

 

[petit_bateau]

Further info on Lithium Refining, melting points that may be able to be used to separate compounds,

Lithium Hydroxide - 462C, 864F
Sodium Hydroxide - 318C, 604F
Sodium Carbonate - 852C
Aluminium Hydroxide - 300C

For clays:-
Lithium Chloride - 605C-614C
Sodium Chloride - 801C

Steel melting point - 1371C-1540C (Can possibly have steel pipes?)

From Master Plan 3, heat pumps can only provide heat up to 200C.

If Tesla develops a heat storage battery Corpus Christi might be a good test site.

Those materials are conductive, so the obvious way to achieve those temperatures is with an electric arc furnace. These are available as an off-the-shelf item in a wide variety of sizes with a long track record in high volume manufacture (aka "mini mill"), i.e. no need to reinvent the wheel.

Electric arc furnace - Wikipedia

en.wikipedia.org

 

[MC3OZ]

The 25K car could be a 'stretched' version of the Robotaxi.

Quoting my own post here:- Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

I can see a lot of good reasons why the Robotaxi would be a 2 seater:-

• A high percentage of trips would be for 1-2 people.
• Larger groups could travel in multiple Robotaxi and could communicate via video conference.
• Customers will always have the option of booking another Tesla model.
• The 2 seat configuration allows a narrower frontal area which means less drag.
• if a Robotaxi uses a 10% smaller battery and is 15% cheaper than the 25K cat, that is a win.

In summary, a compact 2-seater is likely to be the cheapest car Tesla can ever make, which uses the least amount of raw materials and more importantly battery cells.

The compact size probably helps with other issues, like parking, cleaning or travelling in Boring tunnels.

 

[Krash]

…A high percentage of trips would be for 1-2 people.

• Larger groups could travel in multiple Robotaxi and could communicate via video conference.
• Customers will always have the option of booking another Tesla model.
• The 2 seat configuration allows a narrower frontal area which means less drag.
• if a Robotaxi uses a 10% smaller battery and is 15% cheaper than the 25K cat, that is a win.

In summary, a compact 2-seater is likely to be the cheapest car Tesla can ever make, which uses the least amount of raw materials and more importantly battery cells.

The compact size probably helps with other issues, like parking, cleaning or travelling in Boring tunnels.

I would disagree on the 2 seater. Frontal area only matters for high speed wind resistance. Lower speed rolling resistance Will be the same. 10% battery reduction doesn’t give you a 2 seater. Currently a robotaxi wouldn’t have bucket seats, so a robotaxi would go from a six seater (two benches) to a four seater with a 25% reduction (not quite 33%). In battery width.

And since you don’t care as much for wind resistance, if you really want to park it, you don’t care about the width, you care about the length, so to maximize the luggage and minimize the length, you’re getting rid of either the trunk or likely the hood, to maximize the other if crash testing allows.

I wonder how many taxi rides are one adult with two kids? No way one adult splits up a group with two kids.

For me, the only for sure thing is not length or width changes, but rather that the car will induction charge. People or Optimus are not going to plug in those cars like today’s model.

 

[Buckminster]

The benefit of 2 seats versus 4 is in weight rather than frontal area. Mass counts for taxis where much of their miles are in the city at 10mph in traffic. Also, simpler and cheaper to make accelerates the transition.

I think it will be 2 seats because the $25k 4/5 seat vehicle will also be a very capable robotaxi. If Tesla don't go down the 2 seat path then they open themselves up to competition down the line.

 

[MC3OZ]

The benefit of 2 seats versus 4 is in weight rather than frontal area. Mass counts for taxis where much of their miles are in the city at 10mph in traffic. Also, simpler and cheaper to make accelerates the transition.

I think it will be 2 seats because the $25k 4/5 seat vehicle will also be a very capable robotaxi. If Tesla don't go down the 2 seat path then they open themselves up to competition down the line.

Yes, the main advantage of 2 seats is a cheaper vehicle and less energy required for those trips involving 1-2 passengers. For larger groups and other Tesla vehicle, or multiple vehicles can be used.

I'm not sure a 6-seater Robotaxi will be that much cheaper to produce than a Model Y or will use much less energy when driving around with 1-2 passengers. if that is what Tesla are going to do, just make a Model Y without a steering wheel.

Lower energy usage allows more battery chemistries to be viable, probably including Sodium.

Multiple 2-seaters could platoon in a group, parents would still not trust their children riding in a separate vehicle. The Model-X is a good Mini-van for large groups who must travel together.

There is a case for a 6-10 seater van, I am not sure where that sits on the list of priorities.

 

[MC3OZ]

Jeff Dahn sodium battery research...

Radware Bot Manager Captcha

Based on the results so far, electrolyte additive screening was done in cells cycled between 2.00 and 3.80 V at a rate of C/5 to ensure that the results would not be impacted by Na plating.

Addition of 1 wt% PES resulted in very stable capacity retention of 97% after 450 cycles at 40 °C, completely suppressed all gas generation and lowered the voltage polarization significantly. Future work will focus on electrolyte formulations that enable charging to 4.0 V as well as room temperature cycling at practical C-rates without Na plating.

Seems that fast charging of sodium-ion batteries needs more R&D.