The Aftermarket Traction Battery Thread

[mkjayakumar]

No idea how long these aftermarket batteries would last, but I do know one thing:

You'd own the battery outright.

.. and you would know the replacement price too !!

[howardc64]

Currently own a Tesla Model S 85kWhr model. Been eyeing a 2nd electric car for awhile and used Leafs are pretty cheap. But battery degradation is obviously a concern. I'm in engineering and an avid DIYer on auto maintenance so have been learning about EVs.

First, I'd like to explain a little of what I learned about battery degradation. There are major 2 factors (capacity and heat) but most people only talk about heat which is probably the secondary problem.

1. Capacity.

A rechargeable battery life chart shows lifecycle is a lot longer if only cycling with low amount of charge and discharge. Large charge and discharge cycles degrades battery life in any technology including our trusty old 12V lead acid batteries. Here is some info

http://batteryuniversity.com/learn/arti ... _batteries

So a 24kWhr battery used frequently in say 40-50 mile range charge/discharge cycles will loose capacity really fast regardless of thermal control. Using the LiPO4 discharge table from the above link, we can see 80% discharge/charge cycles provides about 1000 cycles before losing 30% of the original capacity. So 50 miles x 1000 cycles = 50k miles. And 4/12 bars loss = 33%. This pretty much matches the data we see from <= 24kWhr battery electric cars.

Cooling is not as much of an issue at this capacity level because high depth of discharge will pretty much guarantee high capacity loss at 50k miles. Other ~24kWhr EVs like Focus Electric also have active battery thermal heating+cooling and all show the same range loss pattern at 50k miles. I think Nissan recognized this and decided to just leave out the expense of active thermal control for a 24kWhr design.

Compared to a Tesla 85kWhr driven 50 miles per day. 20% DOD (50mi/250 range) 2 yields 500,000 miles before 30% degradation. So I hear about high mileage 85kWhr Teslas lose about 10% at 200k miles.

2. Thermal Control

The reason rechargeable battery degrades faster in heat is due to basic chemistry. From high school chem class, we learned every chemical reaction has parasitic reaction. Parasitic reaction basically produces unwanted products and use up some of the reagents. With increased heat (say a bunsen burner in the chem class), reaction (and parasitic reaction) goes faster.

Thermal control also isn't just cool the battery cell. It is about how to cool every corner and interior of the chemical reaction surface. Large cells like Nissan and LG Chem surely have difficulty to cool the center interior.

So what would be an ideal aftermarket LEAF battery look like?

1. Leverage energy density increase to get say 2X capacity for the same space. An aftermarket design using newer battery chemistry should gain the energy density improvements since 2001. Not sure if 2X has been reached but certainly soak up the improvements. This helps with DOD discussed above. If replacement capacity stays at 24kWhr, it will suffer the same fate at 50k miles.

2. Cooling. Tesla battery design is very different from Nissan and LG Chem and I'm guessing there is huge benefit. Teslas uses small cells (little bigger than AAs) and submerge them in coolant for thermal control. Nissan and LG Chem makes big battery cells (size of a shirt box). Bigger cells have better density but smaller cells have better cooling (can't cool the center of the big cells). The key is probably to find the good balance between these 2 factors. Tesla's cell design is evolving from the original Panasonic battery cell to now slightly bigger to provide 2X capacity. Likely a design trade off after gathering statistics on the original design. Cell size can grow a little while maintain thorough cooling to the cell center.

Self contained thermal control would need air scoops on the bottom and internal heat exchangers + pumps + coolant bath) to provide a plug+play solution. Tesla's a battery have 2 coolant ports because they mount the heat exchangers mounted in the front of the car to collect cool air. Leaf have no such coolant port so would need to self source air and internal thermal control components.

Last, I can see Tesla also does 3 other things

1. Limit charging rate when battery is cold or full. Regen is limited under these conditions
2. Limit discharge rate when battery is cold. Can't do rocket take offs.
3. Cooling system goes into overdrive at the Superchargers (basically faster version of Nissan Quick Charge). DC fast chargig is the fastest chemical reaction so it generate the most heat so the cooling system needs to to offer max performance here.

With a few sensors accelerometer and temp sensors, maybe its possible to build in these controls directly into the battery pack.

In summary, following Tesla's world class battery design, a good aftermarket pack would need to

1 Source a good cell design. This is probably the most challenging. Tesla+Panasonic factories probably produce the best cell right now at the right size for density+cooling. They are probably not available for anyone else.

2. Build a good thermal+smart architecture with the good cells. Incorporating active thermal control (heater for cold and cool air for heat). Limit charge/discharge when cold etc..

Probably not going to get this level of replacement battery design but I thought I share with what I learned from Tesla's design.

[edatoakrun]

In summary, following Tesla's world class battery design, a good aftermarket pack would need to

1 Source a good cell design. This is probably the hardest part. Tesla+Panasonic factories probably produce the best right now at the right size for density+cooling. They are probably not available...

Much of your post is well reasoned, but neither Nissan nor any reputable replacement pack manufacturer is ever likely to use current Tesla+Panasonic cells in their packs, due to their flammability and the resulting liability.

IMO, LEAF packs offer superior performance because they do not require ATM.

But even if there were net benefits to ATM, adding it to an aftermarket pack design is highly impractical.

[howardc64]

neither Nissan nor any reputable replacement pack manufacturer is ever likely to use current Tesla+Panasonic cells in their packs, due to their flammability and the resulting liability.

I would agree, unlikely these cells can be sourced.

In terms of flammability. All lithium ion batteries are highly flammable if punctured. Tesla's issue were mainly self leveling suspension cars that auto lowered at highway speeds. Early model's well documented cases where car burned up was due to road debris damaged the front end of the pack at highway speeds. They retrofitted the cars with a titanium shield in the front of the battery pack and issue is pretty much gone. Really not a battery cell design issue as much as low road clearance and physical protection issue. Here is a video of shield vs concrete follow by a video of where and how the shield was installed.

https://www.youtube.com/watch?v=_5OurZx1ihI
https://www.youtube.com/watch?v=AHdxakaGrkw

IMO, LEAF packs offer superior performance because they do not require ATM.

I do think all EV LiON battery packs will be both higher capacity and ATM going forward. Low capacity battery cars depreciate too much to be competitive in the future and ATM will be needed to increase the life cycle of a even higher priced high capacity battery pack.

But even if there were net benefits to ATM, adding it to an aftermarket pack design is highly impractical.

Haha yeah, I agree, quite a bit of engineering involved So I guess best anyone can do is source high cap cells. This would beef up overall range and lifecycle but of course also cost more. So perhaps same 24kWhr capacity at lower end of the cost is all that is practical. All of this is only possible if someone figures out any computer interlinks between the pack and the car.

[glhs272]

So I wonder if an aftermarket pack can be created with the guts of the new 40kwh pack being made for the 2018 Leaf. As I understand it, the pack has the same footprint as the original. Perhaps then if the form factor is close enough, something could be done...

[rmay635703]

Cor will do one sometime mid 2019

[borugee]

kWh battery cost is around $300 for cells itself. Nissan giving replacement for around $6k which is about $5000/24=$250. This is below cost of cells and third party cannot compete, and you should get your battery from Nissan.

[2k1Toaster]

I think it will happen, and if it doesn't I will do it.

I own multiple Toyota/Lexus hybrids as well, and I was hoping for aftermarket cells for the Prius at least since it is so popular. Some of the first that came on the market were crap. Some were OK but didn't play nicely with the OE cells. Now I make a pack that is brand new based of an existing style of kit using my battery factory contacts and they are being installed in cars now. Got them in my Gen-2 Prius and a couple in China that people at the factory own.

I figure when the Leaf's battery degrades enough that I can't drive 20-25 miles on a sunny day, I'll need a replacement. And if there aren't any new cells available, then I'll be making my own. Batteries aren't that hard. Making them physically fit an existing mold is more difficult. But not impossible.

[johnlocke]

Panasonic 2170 cells are rated at 20w capacity. So you would need 1500 for 30KWH capacity. Arrange the cells as 16 cells in parallel by 2 sets in series to match one of Nissan's modules. 48 modules to make up a pack. 1536 cells total. 26.5 in by 40 in. by 2.8 in. thick. Reuse the BMS from the old battery since the cell voltages and currents should be similar. Flood the batteries with mineral oil and circulate it through a radiator for cooling. There should be plenty of room in the under the seat hump for a radiator and pump. You would have to modify the shell for venting for the radiator and fans but it shouldn't be too hard. If you want a bigger battery, just add more 2170 cells to each module. You could actually cram 60KWH into the existing shell but I don't know if the BMS would like it and you would need an external radiator.

As far flammability and liability goes, every 18v cordless device (drills, saws, sanders, etc.) uses the same chemistry and I don't hear any reports of exploding tools or tools catching on fire. The biggest problem with this idea is getting a hold of the 2170 cells in the first place. I understand that Samsung is going to produce a 2170 cell also so that might be an alternative source.

[edatoakrun]

Can someone translate Ukrainian (?) and explain if he has successfully completed a 30-to-24 swap-or not?

Change the battery to the Nissan Leaf. 24 kWh and 30 kWh.

https://www.youtube.com/watch?time_cont ... HLddeIxxig