one more question about Tesla battery

[bobac]

Hi there,

please don't kick me) it is my first installation.
and it isnot finished yet but...

ESS
Victron MultiPlus-II GX 48/5000/70-50
battary 48V (Tesla 2s2p + SIMP BMS)
Victron MPPT and 6s photovoltaic 1.5 kW







thank you Carel
muchasgracias Marc
you help to make dreams come true)


current settings:
BMS
Cell dischare limit 3.2V
ESS
minimum SOC 40%
dynamic cutoff 3.1V
but I have noticedthat it did not get that threshold cuz cells imbalance

for ex.
SOC 45%
but
Voltage: 42.615V
Avg Cell Voltage: 3.551V
Low Cell Voltage: 3.201V
High Cell Voltage: 3.588V
Delta Voltage: 387mV

so dischage stopped.

it seems imbalance caused by load jump






so the question: what is wrong?
wiring?
weak Tesla modules (2014)?
wrong settings?

gentlemen I need your help

 

[ajw22]

current settings:
BMS
Cell dischare limit 3.2V
ESS
minimum SOC 40%
dynamic cutoff 3.1V

Not a Victron/SimpBMS user, but IMHOthe BMS limits should be set to protect the cells from damage (eg. 4.2V ~ 3.0V). Keeping the cells in the lifespan extending range(eg 4.1V~3.2V) should be the job of the charger/inverter.
Also, with older/questionable battery modules,I'd choose a higher inverter cutoff limit of 3.4V to avoid exactly the kind of issues you're facing now.

But the first thing to check: werethe Tesla modules fully top balanced to begin with?

 

[Redpacket]

+ to ajw22's comments, top balance the battery & check the cell(s) that are lowest now is equal full like the rest.
If you have one or more cells in the pack that are low or significantly weaker SoH than the rest, they will drop to low limit first.
The more current you draw, the sooner this will happen.

 

[daromer]

Yeap redo balancing and retest.
And yes what you see is uneven packs in series string. Its rather normal and especially when going below 3.5V where you dont have much left. During load it drops even faster

 

[bobac]

today and before at the end of charge
Delta Voltage was ~20mV

and it is strage but today I have been doing some load tests and it looks much better
now for instance
Delta ~100mV
SOC 41%





almostas desigend... strange

about BMS and Inverter/Charger responsibilities. yes I try to follow
BMS - battary safaty
Inverter/Charger -lifespan

but in case of SIMP + Victron combination it is more smart
BMS manageInverter/Charger via CAN
so BMS knows
Cell overvoltage
Cell undervoltage
and
Cell Charge Limit
Cell Discharge Limit

PlusInverter/Charger knows aboutCutoff it seems to be little bit upper then Cell Undevoltage
If I undestend (I am just a Junior Victron Professional :blush: )

the only difference yesterday I have charge it 60%
but today 75%

and it seems weaker cell become more weaker after 3.5V ... thank you daromer !

 

[Redpacket]

Like we said, you need to bring the battery close to 100% & balance it there.
It worked better from 75% as you are higher up the charge curve.

After balancing near 100% once or twice, maybe operate 95% to 20% SoC for longer life.

 

[CarelHassink]

Roman,
could it be you have a coolant contaminated Tesla module??
does the low voltage cell/cells locate in a single module??
387mV deltaV may point to this
and it should be about ~10mV or better
Perhaps you make a screenshot of the SIMPbms main console,
showing the cell group voltages per module in 3 decimals
this way it's easy to determine which module, and cell group is affected

I have seen this before,
an accidental, but serious mistake was made, when dismantling the Tesla pack,
residual coolant spills over the module, causing corrosion
the corrosion leads to abnormal discharging of the contaminated cell group in the affected module
and cannot be repaired.

The cmu-board in front of the contaminated module is simply not capable of balancing the corroded affected cell-group.
and I guess all your other capacity related issues are connected to this problem.

Carel

 

[ajw22]

I think you're still not properly top balanced.

It looks like you're only charging to about 3.77V/cell. It is very difficult to top balance at those voltages (the almost flat part of the lithium charge curve). A difference of just 20mV indicates a big difference in SoC. So you'd probably have to aim for 10mV or less... inaccuracy of the BMS could start playing tricks here.

The better approach is to charge to say 4.00V (the "spiky" part of the lithium charge curve), where a cell with an even slightly higher SoC will show up a much more noticeable spike in voltage - balancing to 20mV would probably be adequate here.

 

[daromer]

As people say the difference is either above 3.9v or below 3.45v in generall. So start with a full top balance at above 4.0 V and then cycle and you get the true result of the capacity difference.

 

[bobac]

Got it
I'll try to charge 95% few times and share result (with SIMP BMS main console screenshot)

Low and High cell are in the same modul. thay are siblings.
Yesterday I have charged it up to 80% and there was no imbalance alarm during the day.
Yes, sure Carel, I remember your advices and have pumped out all coolant first from whole battery and then per module.
But it might be a corrosion problen cuz there was a problam with that battery. And I remember when I have opened it there were some coolant inside ( but not much. I going to refactor first installation in a few days and look at electrodes.

I thought here... What could be a cause as well... Battery was disassembled ~ in May, SOC was ~60% but I start to use modules last week.. I heard about battery should be woken up for full performance

 

[bobac]

Hola Caballeros,

Merry Christmas and Happy New Year.
May the spirit of Christmas infuse your life and that of your family members with hope, positivity, and joy.


But closer to reality ))
Here is update of mi installation v2

it was possible to charge 8 modules up to 4.16v with delta ~13mV
but other fore only up to 4.0 with delta ~25mV

so today
battery is: 12 tesla modules 2s6p

and what we have





end of discharge



start of charge




please comment.
and I have a few questions

1 what do you think about this jump. it is next day and uppere charge threshold but...

 

[OffGridInTheCity]

Back at ya - Merry Christmas to all.

If I'm understanding the above - you're showing that above 3.7v... charging up to the 3.9v range the max/min difference is only 40mv (3.81-3.77=0.04mv) but as you discharge down to 3.6v'ish things go crazy? (3.62/3.21 = 0.41v).

The only time I've seem something like this is when I mixed Sony (green) G5/G7 cells with *huge* IR (>250) with regular cells and even so it wasn't this wide a swing - maybe 1/2 that. So IR somehow? Possible a really weak module. In any case, can you isolate the bad module(s) and remove or replace them?

I'll be interested to see comments.

 

[CarelHassink]

Hi Roman, I still wonder if:
a) you have failing cells in some modules due to Glycol coolant contamination, and the resulting corrosion??
b) inside your dismantled Tesla pack a mixed group of modules from different donor-cars, and hence a different SOH / due to life-cycle??

These huge differences in delta-V: 13mV to 25mV in modules from the same donor-pack, makes me wonder...
best Carel

 

[bobac]

arh, yes thank you.
there is two places (in my case) when on the charge-discharge cycle it is possible to see "strange" delta voltages behavior

1 charging (last two diagrams): just jump from 3.87-3.71 = 0.16 to 3.81-3.77=0.04mv
it must be a result of BMS starts balancing

2 discharging: 3.62-3.21=0.41
it must be some module/s is weak



1 all modules are from the same box) but I'm sure that battery was treated several times by different "masters". so it is possible modules were mixed

2 yes there is some corrosion

3 thank you, got it, try to replace weakest one and report

here is what we have found inside box among other things )

 

[CarelHassink]

Roman,

1) balancing start voltage, can be set at any voltage in menu, wise is above 3.8V/cell or even higher
(when I see good packs DeltaV is about 4mV)
but balancing will take a long time given the size of your pack + capacity burning off excess energy
through these tiny smd resistors.. to have them balanced at all.
In a corroded pack that'll never happen

2) make the wise choice: eliminate the failing modules,
and see how the "good" ones perform,
if all works well,
bite the bullet & ditch the faulty modules

best of success Carel Hassink

 

[bobac]

Hi there
fresh update and few more details and questions

1 thanks to everybody and @CarelHassink it seems I managed to find out two weak modules and after eliminating them it look like this



2 Current configuration:
Victron: ESS with Scheduled charge 95% (every day from 23:00 to 7:00); SOCmin 25%; NO grid feed-in
Simp BMS: Vmax 4.1 Vmin 3.0
so you can see


3 This setup has enough capacity for daily consumption and I decide to narrow voltage operational span in sake of battery lifetime

As far as I know in standard charge process should be two phases CC and CV
In my case it seems there is only one - CC

Q.
How is it good or not for lithium battery to be charged CC only up to 4.0V?

 

[CarelHassink]

OK good news.
and tell me how many modules in the storage battery 4?? or 6??
The bigger the better, and no worries about degradation too much
do realize, draw on your modules is way way lower as in a Tesla.
4.0V max is already very very conservative approach

we use SIMP/Victron/Tesla between 3<->4.15Vdc

with regards, Carel

 

[bobac]

12 modules

household where the ESS installed is powed by electrucity only
so in winter daily awarage consumption is about 50kWh/day
and as a result battery has one cycle per day that's why so conservative settings

so there is nothing to worry about absence of CV charge phase?
(it reduces effective capacity only)