53kw CATL Upgrade on 2013 Leaf S

[Simon357]

Cell test results provided by seller.

 

[fotajoye]

Thank you for the information.
In 2011, when I bought my first Leaf, I thought that part of the plan for Nissan would be to have a well planned strategy to upgrade and replace battery packs as they aged...turned out I was wrong.
However, it's good to see that there is energy being expended by at least one battery maker to supply aftermarket replacements...I hope this is a good beginning and a wakeup call for Nissan to address this problem and not make the assumption any longer that the high cost of batteries make the older Leafs Throw Away Cars.
Hopefully with people like yourself and Dala helping to steer the industry and it's users,, we will get by this hugely negative obstacle. $6.5k for a pack is a start.
Please keep us in the loop.

 

[cwerdna]

Can folks here stop using the wrong units?

3) Are you able to quick charge to full 53kW and at what power? would be great a report of charge speed during time, a leafspy snapshot would be enough IMHO (having reference to the temperature too)

Not sure what you're asking on your first part.

Energy and battery pack capacity is measured in kWh.

Charging rate is measured in kW, a unit of power. If you charge at 50 kW * 1 hour --> 50 kWh came out of the "wall". Multiply units and values.

50 kW * 2 hours --> 100 kWh
100 kWh / 2 hours --> 50 kW since h divided by h cancels out.

50 kW * 0.5 hours --> 25 kWh
25 kWh / 0.5 h --> 50 kW

The weight difference seems to be 170km for the 24kw and 350km for the 53kw
a difference of 180km x 2.2# = 396#
I'm assuming that NEDC is the weight of the kit

See above.

Vehicles and charging equipment use the proper units. Skip to 6:50 of https://insideevs.com/news/495913/nissan-leaf-dc-fast-charging-curve/, for example. The charging rate (power) is in the middle, in kW. Energy dispensed is in the upper right, in kWh.

If you look at the video at https://electricrevs.com/2018/07/17...hargepoint-express-250-charge-at-up-to-55-kw/, you can see the kW and kWh values besides amperage and voltage.

 

[Simon357]

Here are some LeafSpy charts from a drive with aggressive acceleration.

 

[ginetto]

great report again @Simon357 , for what I can see from the dynamic load, I I did calc correctly and if they make sense (correct me please if I'm wrong) I would say that max difference in internal resistance of the modules (well module + busbar) is around 0.5mΩ

This is my calc where:
V(u) = Voltage under load
V(s) = Voltage before load
R1 = resistace of module 1
R2 = resistance of module 2
Knowing that V(u) drop is V(s) - I*R and knowind V1(s)-V2(s) around ~20mV and V1(u)-V2(u) = 103mV and I = 196.72A

V1(u)-V2(u) = 103 = V1(s) - I*R1 - (V2(s) - I*R2) =>
103 = V1(s) - V2(s) - I(R1-R2) =>
103 = 20 - I(R1-R2) =>
R1-R2 = (103-20)/196.72 ~= 0.42 mΩ

For sure I did some mistake, please verify my though and check if it is correct. If the above is correct, than the quality of the modules seems good. BTW need to do the same calc at different SOC and temps.

IMHO I would monitorize that cells with most deltaV if this is increasing during time.

 

[Tony 240]

Recently I embarked on a battery upgrade using a kit I purchased from SHENZHEN CHUNENG ENERGY ELECTRONICS CO.

https://chuenergy.en.alibaba.com/
www.chuenergy.com

They have several kits, but this one was for their 53kwh version. It included 8 preassembled battery modules, a mounting plate and all required accessories to make the conversion using the old battery shell and BMS.

It does require modification of the old battery tray before securing the new mounting plate inside. It was a pretty straight forward conversion, and the company walks you thru the entire procedure along the way. I was very impressed with their customer service.

The top shell does sit higher which does require spacers when reinstalling the HV battery.

The heater system from old leaf battery was not able to be reused; and I have absolutely no idea what impact, if any, this will have in the future. I am in southern california, so it is not of concern given our climate.

I performed the first full charge last night and the range has increased to 155 miles, but I am sure that will increase as the computer adjusts to the new capacity. The original battery was 24kwh with 64% SOH remaining (55 mile range).

Thanks for this info.
I would like to see the Dc/ to DC (360 volt to 12 volt )convertor be one that went both ways. So that one could hook up extra 12 volt batteries to the DC to Dc converter and increase the capacity of the traction battery without having to remove it and install a bigger one.

 

[cornbinder89]

Thanks for this info.
I would like to see the Dc/ to DC (360 volt to 12 volt )convertor be one that went both ways. So that one could hook up extra 12 volt batteries to the DC to Dc converter and increase the capacity of the traction battery without having to remove it and install a bigger one.

It wouldn't work for many reasons, but the primary one is you couldn't draw enough useful power at 12 volts to be of any use at 360 volts.

 

[denwood]

I was quite pleasantly suprised to read this post as there is hope for an easy rebuild when my 2018 requires it. I'm still at 88% SOH after 94K kms, so good for now.

How is everything going with the car?

 

[Simon357]

So far, everything is fine physically. The only issue I have is related to the range meter not learning true capacity. For example, the meter will go down 40 miles when I only drove 10, and this is after full charge. Then with about 30-40% of battery remaining, the range meter will not change hardly at all and sometimes enter turtle mode.

This seems to be a somewhat common issue with battery swaps in general. I have been working with the supplier and they are testing out a new canbridge to help solve the issue. It has nothing to do with Dala's canbridge, I guess, which is what I am using now...not that I am knocking his brilliant work.

 

[MikeinPA]

So far, everything is fine physically. The only issue I have is related to the range meter not learning true capacity. For example, the meter will go down 40 miles when I only drove 10, and this is after full charge. Then with about 30-40% of battery remaining, the range meter will not change hardly at all and sometimes enter turtle mode.

This seems to be a somewhat common issue with battery swaps in general. I have been working with the supplier and they are testing out a new canbridge to help solve the issue. It has nothing to do with Dala's canbridge, I guess, which is what I am using now...not that I am knocking his brilliant work.

Please keep providing your observations. Very useful. The Leaf may be comparing the voltage of the new pack against a voltage map of the original cell and getting confused about State of Charge.

 

[Tony 240]

I read that Simon Andre EV's ( https://ve.simonandre.ca/ )in Quebec Canada had some problems like this but they said they had to reprogram the computer because it still thought it had the original battery

 

[mux]

So far, everything is fine physically. The only issue I have is related to the range meter not learning true capacity. For example, the meter will go down 40 miles when I only drove 10, and this is after full charge. Then with about 30-40% of battery remaining, the range meter will not change hardly at all and sometimes enter turtle mode.

This seems to be a somewhat common issue with battery swaps in general. I have been working with the supplier and they are testing out a new canbridge to help solve the issue. It has nothing to do with Dala's canbridge, I guess, which is what I am using now...not that I am knocking his brilliant work.

We've done some of this magic in the somewhat distant past. What essentially needs to be done is a remap of the capacity curve, where the CAN bridge reads the cell voltages from the battery every 10 seconds (the internal refresh interval) and calculates capacity based on that, plus some knowledge of the IR, SOH, temperature. It's pretty tricky without good characteristics of the cells, but you get a serviceable result that doesn't cause the worst of the issues of a BMS swap.

What I am concerned about with these packs, and really anybody putting a 24kWh BMS onto NMC cells, is the super-weak balancing and inherently inaccurate balancing algorithm on that BMS. It's a really outdated BMS that can't really cope with these cells. This leads to some kinda nasty issues in the medium to long term:

- Because the BMS can't do proper coulomb counting, it will eventually wraparound and cause the pack to overcharge to approx. 420V. We've had e-mails from 3 or 4 customers over the years with this issue.
- With a lot of intensive use and/or frequent fast charging, it can't keep up with balancing and you can end up with 1 stray cell sitting 100mV+ below the rest of the pack. This is not even just an issue with modded packs, it happens in 30kWh+ packs as well
- Sometimes you seem to get random errors. I'm 50% of the opinion that this is mostly caused by balancing tabs not being mounted tightly, and 50% some weird bug in the BMS.

Our solution has been to install a CAN-bridge that:
- Just completely fucking ignores capacity, SOC, SOH and recalculates this according to the manufacturer-supplied capacity tables
- Blocks all errors from the battery pack and resynthesizes the most pressing ones
- Hard-limits charging to 404.5V - and/or - limits charge rate based on manufacturer-supplied c-rate tables

And then we instruct the customer, who realistically is a pretty decent electrical engineer to even attempt this, to regularly check cell balance and just keep an eye on things just in case.

 

[Tony 240]

We've done some of this magic in the somewhat distant past. What essentially needs to be done is a remap of the capacity curve, where the CAN bridge reads the cell voltages from the battery every 10 seconds (the internal refresh interval) and calculates capacity based on that, plus some knowledge of the IR, SOH, temperature. It's pretty tricky without good characteristics of the cells, but you get a serviceable result that doesn't cause the worst of the issues of a BMS swap.

What I am concerned about with these packs, and really anybody putting a 24kWh BMS onto NMC cells, is the super-weak balancing and inherently inaccurate balancing algorithm on that BMS. It's a really outdated BMS that can't really cope with these cells. This leads to some kinda nasty issues in the medium to long term:

- Because the BMS can't do proper coulomb counting, it will eventually wraparound and cause the pack to overcharge to approx. 420V. We've had e-mails from 3 or 4 customers over the years with this issue.
- With a lot of intensive use and/or frequent fast charging, it can't keep up with balancing and you can end up with 1 stray cell sitting 100mV+ below the rest of the pack. This is not even just an issue with modded packs, it happens in 30kWh+ packs as well
- Sometimes you seem to get random errors. I'm 50% of the opinion that this is mostly caused by balancing tabs not being mounted tightly, and 50% some weird bug in the BMS.

Our solution has been to install a CAN-bridge that:
- Just completely fucking ignores capacity, SOC, SOH and recalculates this according to the manufacturer-supplied capacity tables
- Blocks all errors from the battery pack and resynthesizes the most pressing ones
- Hard-limits charging to 404.5V - and/or - limits charge rate based on manufacturer-supplied c-rate tables

And then we instruct the customer, who realistically is a pretty decent electrical engineer to even attempt this, to regularly check cell balance and just keep an eye on things just in case.

So this my be a dumb question but have you tried to charge the car only on level one slow charger for an extended length of time, so that the battery cells do not overheat.

 

[Simon357]

So this my be a dumb question but have you tried to charge the car only on level one slow charger for an extended length of time, so that the battery cells do not overheat.

I can’t speak for Mux, but in my situation, I have the slow 3.6 charger without fast charge. So I don’t foresee this ever being a problem for me. That being said, I was cognizant that the BMS would have to work longer to maintain balance due to higher capacity cells. So far a month in, and they are holding steady at 10mV deviation maximum.

I am disappointed that this company didn’t mention that my current issues could pop up, and it seems they should have known this as they say they have done many of these swaps personally and send out kits like what I purchased.

Hopefully they come thru soon or I will take it up with Alibaba to rectify the situation. It certainly is not at the state they promised. For the most part, everything seems high quality, but they are missing the little piece at the end which makes all work mostly seamless.

 

[mux]

So this my be a dumb question but have you tried to charge the car only on level one slow charger for an extended length of time, so that the battery cells do not overheat.

It's not an overheating issue, it's an issue of the BMS not being able to keep up with the amount of charge mismatch in the cells. That happens both when fast charging and just using the car a lot. So you can still run into the same issue if you, say, drive the car a lot during the day and charge overnight. Exactly when this can happen depends on the amount of mismatch in the cells and the number of cycles you make.

As a calculation example: say the charge mismatch is 1% and the BMS can balance up to 20mA continuously. The 53kWh pack effectively has 140ish Ah cells, so 1% imbalance would represent 1.4Ah. This means that the BMS can keep up with this imbalance from empty to full if on average you cycle the battery once every 1.4/0.02=70 hours.

Things get a bit more complicated than this (e.g. the BMS doesn't exactly work like this, it tries to equalize charge above a certain voltage threshold, so keeping the battery at a certain SOC will eventually equalize all the cells, then if you charge the pack to full, you'll get a mismatch that needs to be balanced again) but in general this time constant is indicative of how 'well-matched' a BMS is. A BMS equalization time of 70 hours is veeeeery long, about 24 hours is considered acceptable, less for high-performance packs. But you only actually get into trouble with this kind of underpowered BMS if you cycle the battery fast or frequently.

 

[Simon357]

And certainly as the cells age, their internal resistance will increase, some more than others, thereby makes the BMS's job more difficult. Reminds me of the early days of creating solar storage systems with B grade cells and having to use an extra active balancer to help out the BMS. I have one such system still going strong after 5 years, but it would not survive without the extra balancer; although new BMS's have high amp balancers now. Little off topic.

Anyways, looks like they are going to send me that new can bridge 3.0 which is preconfigured for the pack size, cell chemistry, temps, and discharge curve, etc. Does nothing to help the balancing, but I will just have to keep an eye on it.

 

[denwood]

Sound like they need to hire @Dala to fix this! In any case, please keep reporting your progress here if you can I have my fingers crossed for a well though out solution using newer tech cells.

@Simon357 , are you using Dala's canbridge solution for your CATL setup? I've had zero experience with any swaps, just looking for clarification on the setup in as simple terms as you can It would have made sense for the pack suppler to provide a can bridge with the correct parameters for this pack from the start...

 

[Tony 240]

I cannot find a thread on this subject I am bringing up, and I am not sure how to start another thread. Here is my question, why does my 2012 Leaf move as soon as I put it in "gear"? Should it not be in complete stop mode until the accelerator is pushed? Can I adjust the accelerator to 0 speed?

 

[Drew21]

Do you have your foot on the brake when you move the "gear" selector? Does your Leaf begin moving with your foot on the brake or are you talking about no brake and no "go" pedal application?

Is the Leaf moving from engine power or is it rolling?

 

[FinnishCoffee]

I cannot find a thread on this subject I am bringing up, and I am not sure how to start another thread. Here is my question, why does my 2012 Leaf move as soon as I put it in "gear"? Should it not be in complete stop mode until the accelerator is pushed? Can I adjust the accelerator to 0 speed?

It's made to mimic a normal automatic transmission car. It might be possible to remove with some advanced canbus trickery but it seems nobody has done it successfully yet.