Replace individual battery cells to renew/increase capacity

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ripple4

Well-known member
Joined
Sep 18, 2018
Messages
110
Location
Toledo, Ohio
New leaf owner here. lurked quite a bit. i gathered up pieces for an idea around the web and wanted to float an idea to see if i was missing something, what could go wrong and try to get ideas on how to work around any potential issues. I was looking at 32ah 3.7v pouch cells and found that it may be possible to make a leaf battery for less than Nissan is selling for. Alibaba has got 3.7v 32ah pouches for $27/each which times 192 is $5200 + shipping. assuming no volume discount. these cells advertise 3000 cycles, which is more than 9 years of life.

https://www.alibaba.com/product-detail/3-7v-80156240-32Ah-Lithium-polymer_60741109055.html?spm=a2700.8592591.0.0.23a910a9o1UCd1

After watching some YouTube about dissembling a 4-cell unit the pouch tab design looks to be unique on the Nissan cells, both side to side and in protrusion from the cell. Also the internal black plastic supports for the cells might not be reusable since the internal connection would need to be in different places. But with ABS plastic and 3D printing a new end cap could be made with new bus bars resistance welded to the generic cell tabs and locate the buss bar connection lugs in the correct spots for the leaf pack wiring. or the cell ends could be wired to connection points and then encapsulated in epoxy with mounting holes that are cast in.

Going further, if one was going to make a clean sheet cell format, since the Nissan cells are ~52mm thick then one could stack 6x 8.5mm pouchs into the same thickness. then it would not be space prohibitive to upgrade from the 2p2s cell arrangement to a 3p2s arrangement and still fit in the stock battery case. also one of the biggest complaints i read is battery cooling. i suggest that since cells are slightly thinner without the aluminum cans and various air gaps, it would be possible to install copper heat conduction sheets every second cell layer and tie the ends opposite of the buss bars with a fluid heat transfer block. the heat transfer fluid i would use something like mineral oil or low viscosity transformer oil, in case it ever leaked. and then plumb each of the 48 heat transfer blocks to a liquid/liquid heat exchanger that connects to the leafs gylcol coolant system. coolant pipes run back to the charger on '11 and '12 cars so teeing into those would be the plan.

proposed layout:
e7dk2K


https://ibb.co/e7dk2K


if that was not enough to soak in, they have pouch cells that go up to 50ah in close to the same size format, which would push the stock 22.1kw cars up to 35.5kw in the 192 cell 2p2s. but with 288 cells in a 3p2s arrangement then pack capacity would be 51.8kw, still fitting into the stock battery case and with active temperature management.

Lastly, looking at videos of people who put in G2 cells into old G1 cars, they seem to say that the cars original BMS will not immediately allow the extra capacity, but over time it will learn to use it and give extra miles on the range meter. (http://www.mynissanleaf.com/viewtopic.php?f=30&t=24732&start=10) so in principle with enough time, the courage to open the traction pack, re-engineer what hundreds of very very smart professionals perfected, might save some money and/or give capabilities not previously possible. it maybe possible to create a battery after Nissan stops supporting the car at the end of the day. Do the experts see problems here? what could go wrong with each step and how could it be worked around? i personally dislike newbie post with grand plans, and if i'm not missing anything i was going to get 6x 50ah cells and make one battery unit cell and attempt to test it, like seeing how fast it can charge without overheating with the cooling sheets. what the real capacity would be of the 50ah 3p2s unit cell within the limits of the BMS voltage cut offs. etc.

opening up pack to get to unit cells
https://www.youtube.com/watch?v=0dBjPwJ_Qaw

opeing up unit cells and looking at pouch cells
https://www.youtube.com/watch?v=WcqaAz1vjlo
 
Your proposed layout image/doc isn't displaying for me presumably because I don't have a google account. Do you have a way to upload/display that without being logged into that service?
 
ripple4 said:
also one of the biggest complaints i read is battery cooling. i suggest that since cells are slightly thinner without the aluminum cans and various air gaps, it would be possible to install copper heat conduction sheets every second cell layer and tie the ends opposite of the buss bars with a fluid heat transfer block. the heat transfer fluid i would use something like mineral oil or low viscosity transformer oil, in case it ever leaked. and then plumb each of the 48 heat transfer blocks to a liquid/liquid heat exchanger that connects to the leafs gylcol coolant system. coolant pipes run back to the charger on '11 and '12 cars so teeing into those would be the plan.

Cooling for the charger and cooling for a battery are rather different.

The charger has a maximum temperature of operation, perhaps over 100C, and as long as it stays cooler than that all is well. Standard automotive component range is -40C to 125C, however there are 150C and 200C components available as well. It is common to have a safety factor between the component ratings and the temperature that the charger stops normal operation because of over temperature.

Battery would like to stay below 35C or so for best life. Normal temperatures that cars with thermal management have as cooling temperature thresholds to are usually in the range of 30C to 40C. Lower than 30C isn't wise due to condensation. The performance types would like NIO EP9 type cooling of the battery, or at least Tesla's cooling. Tesla's cooling will not allow for a full lap of Nürburgring Nordschleife at full speed. A passively cooled battery works out just fine for most of the USA for basic driving needs such as a reasonable commute and around town.

If you wanted to retrofit a battery cooling scheme, you would need to engineer the whole thing.

Nissan has had battery chemistry problems with the 2011-early 2013 models. The 2013 and 2014 are better, and the later cars are better still. Also the BMS isn't an easy problem, both capacity estimation and (in other cars) cell balancing are complex issues that car makers haven't done the best at. And there is potential for improvement of battery life by handling these issues better.

Buying cells off Alibaba doesn't give me warm and fuzzy feelings...

And there are far worse things that can go wrong in Li ion cells than capacity fade. Nissan isn't the worse, unlike all the Tesla fanbouys here might try to tell you. Ford and SSN, GM and out of battery with 50% charge, and the 2012 Zero S.
 
i changed the picture hosting website, but it still does not show up. i provided a link.

What can i learn from the battery cooling information there. keep max battery temp <35C/95F and don't assume that the vehicle liquid cooling loop can carry away the heat? how does the NIO car cool the battery? I've designed low delta-T cooling loops for industrial equipment and it was sort of easy because i knew how much heat was being produced because its stamped on the motor tag. with this battery cooling idea, its not clear how much heat it will need to remove and how quickly. I suppose charging is the scenario to consider and taking a stab at it with a 85% L2 charging efficiency the heat put into the 24kwh pack with a 6.6kw charger would be .5kwh/hour or 2000 BTU/hour and 1/2 that for a 3.3kw charger in a base/early model. On a 92% efficient 50kw DC fast charge the heating rate would be 8kwh/hour 27KBTU/hour. With these numbers either no cooling is required at all, or a 2-ton chiller is. Just using the mass of the lithium in the battery (295lb 192 cell 32ah) to absorb the heat of charging , assuming no heat transfer to the air or frame of the car, would be 36 degrees F for a 24kwh pack taking a 18kwh charge or 50F for the 570lb 288 cell 50ah example taking a 48kwh charge. so maybe designing the most aggressive active cooling system possible up to 30kbtu/hour with a 10F delta T to ambeit air, while still fitting in the factory battery box is the most practical approach. And then whatever is not actively removed will be left over and so heat the battery.



L2 charging efficiency info:
https://www.veic.org/documents/default-source/resources/reports/an-assessment-of-level-1-and-level-2-electric-vehicle-charging-efficiency.pdf

DC charging efficiency info:
http://www.mdpi.com/2032-6653/7/4/570/pdf
 
ripple4 said:
with this battery cooling idea, its not clear how much heat it will need to remove and how quickly. I suppose charging is the scenario to consider and taking a stab at it with a 85% L2 charging efficiency the heat put into the 24kwh pack with a 6.6kw charger would be .5kwh/hour or 2000 BTU/hour and 1/2 that for a 3.3kw charger in a base/early model.

Driving matters as well, perhaps even more than charging. Depends on how you drive and charge.

Power lost in battery is very roughly current squared times series resistance. I^2 * R
Current is power divided by voltage. W / V

Voltage is about 350V to 400V, varies over SOC.
Resistance varies from about 0.1 ohms to about 0.35 ohms, depending on age, direction of current and more.

https://avt.inl.gov/sites/default/files/pdf/fsev/batteryrpt2011NissanLeaf0356.pdf

So with a full power discharge, might be 18kW. Think driving very fast up a very steep hill. Or could get close to that with repeated accelerations and regenerative breaking. (80kW discharge followed by 55kW charge). Round the Nurburgring.

However, unless you are driving round the 'ring, anything close these numbers can't be sustained. So you need a model as to how the car is used.

Suppose you L2 charge for a hour once a day, and drive at a reasonable power for a reasonable commute. Then you get to no cooling, unless the outside air temperature is averaging over 30C or so.

Or you could do a road trip model. Driving at x power, DCQC, repeat. In Kansas in the summer, at 45C. Or Nevada, or Death Valley. Or something like that.

Remember that preventing condensation at any time, coolant leakage even in the case of an accident and mechanical stress under all conditions on the cells are all more likely important.

As a road trip example, I recently did a 160 mile trip with 2 DCQC sessions, total time of 5 hours including stop at the midpoint. Battery temperature went from 18 C to 36 C at the end of the trip. Ambient was about 20 C plus minus 2 C the whole trip.
 
Thank you for the suggestion to calculate the heating effect using pack resistance. i looked at several websites and got good values to use for the i^2r heating effect, .12ohm for charging and .14ohm for drawing. also in the scenario of the 3p2s module. the resistance should be reduced proportional to the AH increasing so upgrading from 2p2s with 32ah cells to 3p2s with 40ah cells will reduce the resistance, all else being constant, 47%.

http://media3.ev-tv.me/DOEleaftest.pdf
https://www.energy.gov/sites/prod/files/2015/01/f19/batteryLeaf5045.pdf

chs8uz

https://ibb.co/chs8uz

This chiller product might be an option for cooling that is a stand alone solution. its a phase change cooler that runs on 12VDC (@30 amps) and can provide 3685 btu/hr of cooling with its own liquid plate heat exchange, which would be just the ticket for the oil-based heat exchange fluid i'm suggesting. it has a COP of about 3, which is much more efficient that peltier coolers. it could be mounted between the body and the under body tray, behind a body panel or in the engine compartment somewhere, being 10*7*6 inches. Assuming no heat transfer otherwise and the stock 24kwh pack 3.6Kbtu is enough to quickly reduce temperature while L2 charging at 6.6kw and also hold temperature drawing on the pack at 17.5KW, which is about 65mph on level ground. and in reality with the heat radiation that normally cools the pack it might allow for much higher draws holding the temp steady. Draws/inputs higher than 30kw will heat the battery regardless with this particular cooler. with the lower battery resistance of 3p2s then the cooler can almost counteract a 50kw DCFC.

keeping the 12v battery charged by itself seems to be an issue reading this forum, let along sucking 30 amps out of it all the time, how could that be worked around? aftermarket 380v/12/24/48v DC/DC converter?

https://uploads.strikinglycdn.com/files/5740a203-bf47-41e7-8142-07b1486d64b7/Liquid%20Chiller%20Module.pdf
https://www.rigidhvac.com/store/products/59247-dv1910e-p-12v

1.5kw 380vDc/12vdcDc converter
https://www.mouser.com/ProductDetail/Vicor/BCM384P120T1K5AC1?qs=sGAEpiMZZMvGsmoEFRKS8BBm%252bSaojM%252bA%2fu1lpdx%2f3liDgMeBlXU0ug%3d%3d

I understand that condensation would be bad inside the battery case, but does water vapor diffuse into the battery pack once its sealed? could a large pack of descant be put in the pack and allowed to sit for a few days, then removed just before being sealed up, or left in there indefinitely? alternatively, an ardiuno with a hydrometer sensor could be measuring RH in the battery case, and once it got to 90% or close to condensation it could limit cooling.
 
ripple4 said:
I understand that condensation would be bad inside the battery case, but does water vapor diffuse into the battery pack once its sealed?

The battery case isn't and can't be completely sealed.

As long as no part of the cooling system gets below the dew point temperature there should be no condensation. Dew point record for the world is 35C. US record is 33C.

https://en.wikipedia.org/wiki/Dew_point
 
ripple4 said:
Thank you for the suggestion to calculate the heating effect using pack resistance. i looked at several websites and got good values to use for the i^2r heating effect, .12ohm for charging and .14ohm for drawing. also in the scenario of the 3p2s module. the resistance should be reduced proportional to the AH increasing so upgrading from 2p2s with 32ah cells to 3p2s with 40ah cells will reduce the resistance, all else being constant, 47%.

If you'd like to determine the resistance of each battery module used in the Leaf's battery, here's a very good inexpensive device;

https://www.ebay.com/itm/New-Portable-Battery-Internal-Resistance-Meter-Voltage-Tester-Voltmeter-SM8124A/122646399196?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

Resistance values from different modules can be compared and matched when constructing a custom battery system.
From the measured resistance of one of the matched modules and how they're integrated, (XPYS), one can easily calculate
the overall effective series battery resistance. Then one can estimate the battery's internal power loss based on various
loads.
 
Thanks for the suggestion on the resistance checker. I was looking at battery chargers, I want something that charges at the fastest possible rate and I found this 800w one.

https://www.ebay.com/itm/ISDT-T8-1000W-30A-2S-8S-Smart-BattGo-Balance-Battery-Charger-Lipo-LiFe-NiMh-Pb/152796954334?hash=item2393687ede:g:GdYAAOSwiIxaFFaU:sc:USPSFirstClass!43465!US!-1

Also I got final pricing in on the slightly higher capacity 40ah cells and it looks like to build the 3p2s prototype module it will be $250 for 6 cells, and $30 in copper sheets and bars, plus the tester and charger So that's pretty reasonable to start testing on. the 50ah cells don't seem to come in the 8.5mm thickness that I was looking for. with the prototype module I would like to primarily measure the heat generated during discharge/charge. and buy a used leaf battery on eBay and compare the two also using the leaf cell as a taemplate to design the cell holders to match. anything else that I could look at without getting in the car?

The cost for just the 192x 40ah loose cells was $5962 delivered which would be the 2p2s 28kwh version. 288x 40ah cells for the 3p2s would be $8944 delivered, but would supply the 41kwh capacity. and maybe if the thermal system works out it could be looked at as a way to get 2019 performance out of a 2011-2015 for ~1/2 the price. open question: does anyone think a GoFundMe would be successful to build one pack and try it out?
 
Looking further into this i found that there is a cell chemistry called Lithium Titanate in the marketplace. these batteries can have even higher current ratings than LiPo, they also have a much wider temperature operating range, up to 50C, and down to -30C. They also have 10x the cycle life of LiPo, with LTO batteries a pack would last 20-30,000 cycles.

https://www.alibaba.com/product-detail/Victpower-Lithium-Titanate-Batteries-30ah-2_60710440797.html?spm=a2700.7724857.normalList.110.1c775fefVVxGYE


https://www.ebay.com/itm/Lithium-Titanate-LTO-Kokam-55ah-high-performance-Battery-cell-6-pcs/153192262309?hash=item23aaf86aa5:g:Ah8AAOSwgGBbp984

But they have the lower cell voltage, so how can we work around that? The idea to have 6 cells in a module instead of 4 can help with this, running the numbers a 2012 24kwh pack has 96s Lipo max charge volts are 4.0 and the cells is flat at 3.3v that works out to 384 max volts, 360 nominal (@3.75v), and 316v minimum with a total of 65ah. A module made of 6x 55ah LTO pouches in a 3s 2p arrangement would have a max pack volts of 396v (144s*2.75v max charging), nominal of 345v (2.4v nominal) and minimum of 316v and have a capacity of 110ah (36.4KWH and 54-87 years life!!!)

While the 2012 nissan leaf BMS obviously cannot measure two taps inside of each module it could be worked around by putting a high impedance voltage divider across the center LTO cell in each module, that way it would report to the BMS nearly the same “cell voltage” it would expect from a center tap in a 2s LiPo module. But doing that will prevent any balance current coming into the cell. How big of a problem would that be for LTO chemistry?

Also the turtle mode would be activated with some power still left in the LTO pack, but would the BMS have the ability to learn to allow lower voltages before turtle mode kicked in over time? some of these questions can only be answered with experimentation, since its never been tried. sort of like how the BMS learned to allow extra capacity when G2 cells are put into a G1 pack DIY. What about completely fooling the BMS and having an elaborate voltage supply board that scaled the LTO voltages to expected LiPo voltages for any given state of charge at each of the 97 BMS tap locations to allow the LTO pack to go down to its minimum? if the BMS is fooled anyway, then the pack could be made into any arrangement less than the max inverter volts and work with any chemistry! with 100,000+ leafs on the road and near the end of their packs lifes its worth at least talking about it. would this LTO chemistry finally shoot down the main complaint about EVs, that the battery is short lived, expensive and over its life cycle worse for the environment? with a 54 year life, a car would be multi generation, my grandkids could still be cruising around in my leaf in 2072!
 
ripple4 said:
Looking further into this i found that there is a cell chemistry called Lithium Titanate in the marketplace. these batteries can have even higher current ratings than LiPo, they also have a much wider temperature operating range, up to 50C, and down to -30C. They also have 10x the cycle life of LiPo, with LTO batteries a pack would last 20-30,000 cycles.

You would want a completely revamped BMS. Not only because of different cell voltages, but also would need to handle cell balancing, charge tapering and more. Reverse engineer or license from Nissan the communication protocol to the rest of the car.

The Honda Fit has a Lithium Titanate battery. People seem to mostly like it, other than at low temperatures. Cold weather range seems rather worse than other BEVs. A more aggressive battery heater and a larger pack size might make it acceptable in cold weather. However, the energy density isn't as good, so a larger pack size would be more mass and volume, making for an engineering challenge.
 
I made a mistake with my pack voltage calculations. different manufacturers have different cell volt info and I got in the weeds. so I reworked it and the 144s LTO pack is perhaps a very close match to the 96s LiPo pack at important voltage states. The LTO idea I'm talking about is replacing the 48x 7.5v, 65ah 2s2p modules with 48x 7.2v, 3s2p modules of nearly equal physical size. the cells I link to are not the right shape, I was just trying to show the market availability. From a packaging perspective the stock 2012 modules are 303x213x55 which is 3716cc per module and there looks to be space to expand the cell foot print a little, maybe an inch longer. the amp hour capacity of a 6 LTO cells that could fit into the 4100cc maximum envelope for each module would be about 48 amp hour, for a pack power of 33kwh.

the primary idea of this thread was to talk about replacing the cells in the battery with anything else. I still like the new 37ah LiPo plan with liquid cooling. the prototype 2s3p module is going to happen.

You would want a completely revamped BMS. Not only because of different cell voltages, but also would need to handle cell balancing, charge tapering and more. Reverse engineer or license from Nissan the communication protocol to the rest of the car.

As I corrected myself the overall pack volts are perhaps closer than I first though, so would a new BMS really be easier? nearly all market-available BMS' are less than 100a and don't get close to 144s without a scary stack of daughtercards daisy chained together. are we talking about a custom designed BMS? know an EE? also, do other new Nissan cars have the ability to be interfaced with non OEM canbus parts? I was operating on the assumption that Nissan will never help a person or a business interface with older leaf's, because they want to sell new leaf's, or that's the vibe I get reading this forum. while I've never worked with Nissan, my experience in automotive engine control systems is that even the larger tuning houses, AVL, Roush, FEV and the like cannot interface with OEM computer system without inside help. the people who do get stuff to work use open ECUs and toss the OEM computer.

I understand the risk of not using a cell balancing BMS for LiPo cells, as they burst into flames at the slightest provocation. hoverboards right! but even in the stock 2012 2s2p LiPo pack there is not protection for the 2 parallel cells, that's less than ideal for sure and we accept that risk. according to the web sources I've read, LTO cells do not explode and burn when crushed or short circuited and can recover from a 0v discharge. so the risk of a dead cell would be reduced capacity until the faulty cell was replaced and it would show up on LEAFspy as one or two low cells before it got there (the BMS would think its 96s so it would group up to 2 serial cells together in the readout). and the 144s arrangement of 48x 3s2p modules would still allow balance current every three series cells. just brain storming here with the LTO stuff, thanks for the good leads!
 
Here is an update on my idea to renew and expand the capacity of my 2012 Nissan leaf battery. I have a charging and discharging setup ready to go which will be used to measure the heating effect to determine the size of the cooling system I talked about in my first post. For charging I got an ISDT T6 charger, which is capable of up to a 30A charge, although I only have a 20amp power supply for it at this time. This charger, with a large enough PSU, can simulate a 10KW charging rate. On the other side of the equation I have a resistor load bank capable of up to a 45amp discharge, with switches to set draw in 6 amp increments, this can simulate up to 16kw draw. I used 10AWG wires as that was the largest size I could fit into the XT60 hobby plugs, but at 45 amps the wires get pretty warm. I also received 6x LIPO 37ah pouch cells from Longyun Technology Co., Ltd. The price was $40/each shipped which is ~$288/KWH delivered. There is a bulk discount on cells and shipping so the price will scale favorable with a whole pack quantity, but as it stands, this is $6912 for the 27KWH pack size and $11,520 for the proposed 40.5KWH pack size. Does anyone have any idea how to get quality cells for less money?

My plan is to test a stock Nissan module I bought off of eBay from a 2012 leaf with a 12 amp charge rate (roughly simulating a 3.3kw charger) and then a 30 amp discharge rate (roughly simulating a 12KW, 55mph cruise speed) and measure the temperature rise of the battery in an 1” thick foam insulated box from 4.13v/cell full charge to a 3.43v/cell low charge which I believe is about the very low battery warning voltage (VLBW.) That will tell me the available power and the thermal properties of the seasoned stock battery that I have. I don’t know how I could simulate a 50kw Chademo quick charge and whatever KW a 75+mph, 6% grade uphill draw from the battery would be, or even if that Is necessary to measure the heating effect in those conditions being that I am well insulating the battery. Anyway, then I’m going to make the 3p2s cell with the Chinese cells I bought and repeat that same test. This should tell me if the larger AH size reduces the heating due to the lower cell resistance. Then I’m going to add the heat removal system (comprised of a thin copper sheet between every other cell connected to liquid cooling block at the bottom/non-connection side of the cell) and test it again. Assuming that I don’t start a fire, give up, but get promising data I’ll design and make 3D printed endcaps and any other parts to hold the module in the stock arrangement. Then get copper bus bars to locate the connections in the stock location. Anyone have any suggestions about other things to look at during the heating test or other data to collect? Any suggestions?

Test setup:
https://ibb.co/c8NnjjR



Charger screen and cell balance wires in the side. The 3 cell balance wire can be left in the charger even though the resistor load bank is used for discharge, so if the battery is out of balance I can stop the discharge and also will tell me when the battery is at 3.43v/cell.

https://ibb.co/swfSWzG

 
ripple4 said:
... keeping the 12v battery charged by itself seems to be an issue reading this forum, let along sucking 30 amps out of it all the time, how could that be worked around? aftermarket 380v/12/24/48v DC/DC converter?
...
I'd be thinking along the lines of an auxiliary battery that would charge off the car's DC/DC converter, but would be isolated from the car's electrical system when the DC/DC converter was offline. These kinds of systems are common in the RV world where people want to run their camper without worrying about being able to start their truck.
 
davewill said:
I'd be thinking along the lines of an auxiliary battery that would charge off the car's DC/DC converter, but would be isolated from the car's electrical system when the DC/DC converter was offline. These kinds of systems are common in the RV world where people want to run their camper without worrying about being able to start their truck.

Thank you for sharing this suggestion, I dd not yet know about this! looks like the product you are talking about is called an "echo charger" in the RV market. The refrigeration pump i was looking at would draw 30 amps, and two liquid pumps would draw 10amp each. so i would be looking at a 12v/50 amp charger. the use of this charger would be to facilitate the stand-alone, battery powered refrigeration system with two cooling pumps and the primarily time it will be used would be when the car is DC fast charging. So the need would be to have an extra 12v power source that could provide 50 amp for 1 hour. then when the car is restarted the stock DC/DC could recharge that supplemental power source using this echo charger and it should be transparent to the stock 12vdc power system.

Another approach would be to charge the supplemental 12vdc system with a 120v external power source. The question is what is the use of the car, if its a work commuter with 120v plugs at both ends of the journey then the utility-->12vdc charge is looking pretty good. maybe someone in Arizona needs that. however, if the design is for a road trip situation where one jumps from DCQC to DCQC then the DC/DC "echo" charger is looking good. this project is to design a battery system that will correct the short comings of the nissan no-TMS battery pack with proprietary cells. so i think i'm leaning towards the DC/DC version. that way when electrify america has DCQC across the county it would be possible to run across the whole county in a older leaf, in the summer with 120 mile stops between DCQC.

here is a DC/DC charger that can from a 12VDC source at a 50 amp rate into different lead or lithium cells.

https://www.donrowe.com/KISAE-DMT1250-Abso-50A-DC-DC-Battery-Charger-p/dmt1250.htm
 
ripple4 said:
Another approach would be to charge the supplemental 12vdc system with a 120v external power source. ...

https://www.donrowe.com/KISAE-DMT1250-Abso-50A-DC-DC-Battery-Charger-p/dmt1250.htm

That device you linked has two inputs, one for a solar panel. Perhaps that second input could take a wall-powered DC supply instead, giving you both solutions in one device. If that won't work, then it wouldn't be too hard to put in a relay that switched to a wall DC input when connected.
 
I completed some charge/discharge testing and so far I found that the heating of the battery from modest discharge is best modeled with the “inefficiency method” of calculating heating. This is where the power provided to the load, times some inefficiency number is the heating of the battery, and 10% looks right. So a 10kw discharge Load rate means the battery is heated at a 1kw rate for example. An earlier poster suggested the cell resistance multiplied by current ^2 model, which is low by a factor of 10 when compared to the real world data, so for a given load rate the temperature rise will be 10x what that model predicts it will be.

Here are the details of the test data, I discharged the 2012 used module at a roughly 30 amp rate, which took 70 minutes and dissipated 274watt/hours, between a 4.13volt charging max volts and a 3.43 discharging min volts. I guess this cell is about 40% degraded. In that time the cell heated up 15 degrees F in the insulated box. That data tells me the heating rate per discharge/charge rate and I can just proportionally increase the rate up, to know what a DCQC will do.

Doing some fancy thermodynamics math that measured temp rise means the module had a 4,000 BTU/hour heating rate at the 30 amp/12KW discharge, and that would indicate to me a 140amp/50kw DQQC would heat at a 17,000 BTU/hour rate. This shows that the refrigeration system I previously identified with a 3685 btu/hour cooling rate will be too small by itself and will need an air/liquid pre-cooler to keep temps under 95F on summer days. Anyone have a correction to that math?


https://ibb.co/37fpv8p

Sorry if this is hard to follow, I want other people’s input so I don’t miss something, to illustrate this new layout I am now thinking that the coolant loop will have the 48 cooling blocks on each module plumbed all in parallel to ½” diameter pipe fittings exiting the back of the pack shell with sealed stainless bulkheads and disconnects. At the outlet of the pack the warmed oil heat ransfer fluid will first meet a large air/liquid intercooler with a ducted fan, then the refrigeration chiller then a dual element 300/600w heating element, then a reservoir, then the pump. The plan here is all this stuff will fit under the trunk in that open space above the underbody panel. I show temperature measurement points to more tightly control the loads the system will need to power. For instance if the intercooler is cooling the battery well enough by itself the chiller would not need to turn on, but if the temp is too low instead of too hot, then both the intercooler fan and chiller will not activate but instead a heating element will. Any thoughts?

https://ibb.co/YcrC5v0
 
Keep experimenting! I know some people take a different approach at increasing range, just adding another battery pack in parallel with the old one, like Muxsan's solution.

What about 18650's? I have scavenged 10kWh worth of batteries for <400$. Ofcourse they are already used, but should be stable enough.
 
I had looked at the 18650 format battery first, and some of its larger cousins. its obviously a great idea because tesla uses that size i think. only situation i ran into is that there are so many counterfeits that i just did not want to go there. Even ones marked LG and PANASONIC can be re-wrapped with the right color sleeve and date code. websites i found with reputable listings had them at $4-$6/each+ for 2.5ah cells. so if my target improved capacity is 120ah I would need 4600 of them, yikes. the idea was to beat the Nissan price of $7500, and pouch cells are cheaper on a per KWH basis. but i'm open to new info so if you have a source for inexpensive cells please pass it along, i'll build a prototype module out of them at least.

as far as used supply, there are a lot of people doing the DIY powerwall with the good quality laptop cells, i don't think the supply is there for even one car let alone tens or hundreds of car going that route.

update on the testing. I smoked the 20amp power supply so i bought a new 30amp power supply and have began testing the new pouch cells. at the 30 amp rate with one cells its (equivalent to 60 amps in a 2p arrangement) about 1/2 the power rate of the 50kw DCQC. not surprisingly I found that its quite easy to heat a cell charging at 30amps to well over 100F, so the insulated box might have to go in favor of some kind of ambient cooling to compare to the active cooling. here are some pics of the active cooling setup:


https://ibb.co/cL84ZH2


the dictonary is to add weight hopefully gently pressing the two cells into the .035" copper plate.

https://ibb.co/zZFd8cn
 
Back
Top