Oleksiy
Posts: 82
Joined: Fri Mar 23, 2018 4:30 am
Location: Kyiv, Ukraine

Cell balancing

Thu May 02, 2019 4:54 am

I checked my logs during the extended weekend we had in Ukraine, and it appeared that my last 100% charge took place back in March. And also, I had only five full 100% charges during March as well, with average beginning of day SOC staying at 85% both for March and April. This looked like a perfect opportunity to see some out-of-balance cells.

First thing I tried was to deplete the battery closer to zero (the cell balancing situation should deteriorate at lower SOC), but managed to do this via climate settings only down to 3%. After that the system said "Independent ventilation deactivated", and that was it. I din't feel comfortable driving around lest the car died far away from the charging station.

The result was as follows:

Image

So, I have 3.54-3.55 cell voltage range, cells seem to be in good balance at 3.2% SOC. But other indicators look like a total mess. The upper SOC limit is at 83.8%, the lower - at 8% - both normal and matching 16.5 kWh remaining battery capacity according to Batt.Kapa. What looks suspicious is the cell voltage itself, and the battery voltage as well. Comparing 3.55V cell / 340V battery with the perfectly healthy pack here (albeit at 0% SOC with out of balance cells) https://okedv.dyndns.org/wbb/wcf/index. ... board-png/ (taken from the page with instructions https://okedv.dyndns.org/wbb/blog/index ... -auslesen/), the voltages look way too high for 3% SOC.

I than charged my car overnight, leaving it at 100% SOC for about 6 hours in an attempt to balance the cells properly. The 100% output looks like this:

Image

Upper and lower SOC limits remain the same, real SOC is at 81.7% vs. 100% SOC displayed. But cell and battery voltage (4.09 and 392V seem to match those of the healthy battery at 100% SOC (91% actual SOC) here https://okedv.dyndns.org/wbb/wcf/index. ... board-png/.

Provided that I could trust these readings at all, my conclusions are as follows:
(1) Cell balancing does not necessarily occur at 100% only. Or they are well balanced on their own despite charging to lower than 100% for many weeks in a row.
(2) Lower SOC limit seems to be fixed at 8% for all packs regardless of the degradation (I checked many similar screenshot examples taken from different cars on the web).
(3) Upper SOC limit doesn't seem to represent the real SOC limit since the pack's voltage is the same for both 82% and 91% actual SOC (100% displayed SOC).
(4) Most likely, upper SOC limit is unchanged regardless of the degradation level, but the BMS just limits the discharge capacity of the pack by voltage, hence effectively increasing the bottom SOC limit.

Still, 3.55 cell voltage for 3.2% displayed SOC seems way too high for my case. Taking the healthy battery example under the links above as a base, and provided the voltage drop vs. SOC is linear, and also given the 12% degradation of my pack, the cell voltage should be at 3.17V, not at 3.55V. Or I'm wrong in my understanding / calculations.

Please let me know your thoughts. Any real life examples of your own?

jadnashuanh
Posts: 4483
Joined: Thu May 22, 2014 2:07 pm
Location: Nashua, NH USA

Re: Cell balancing

Thu May 02, 2019 3:19 pm

Discharge of a LiOn battery is not really linear when measuring voltage alone. It drops off like a cliff near the end.
Jim DeBruycker
2011 535i x-drive GT, 2014 i3 BEV

Oleksiy
Posts: 82
Joined: Fri Mar 23, 2018 4:30 am
Location: Kyiv, Ukraine

Re: Cell balancing

Fri May 03, 2019 3:43 am

Thank you, I stand corrected, it does seem to plummet abruptly closer to zero SOC. Unfortunately, I wasn't able to find any voltage vs. SOC curve, but based on the screenshots borrowed from here https://www.goingelectric.de/forum/view ... &start=750, this healthier battery displays only slightly lower cell voltage at 8% displayed SOC (3.53V) vs. 3% of my battery (3.55V).

At 4% the pack's cell voltage is 0.1-0.2V lower than in my case at 3% SOC - 3.38-3.47V. So, it looks like the bottom unusable buffer went up in my case due to degradation.

Image

Image

rcy001
Posts: 2
Joined: Sat Dec 22, 2018 3:46 pm
Location: Netherlands

Re: Cell balancing

Sat May 04, 2019 11:00 am

Since your batt kappa is only 16.5 kWh, you could try to reinitialize the cells with ISTA. Then see if the low SOC voltage changes with higher kappa
2014 Rex
CCS, 428, Andesit
Suite, H-K, Sunr

Oleksiy
Posts: 82
Joined: Fri Mar 23, 2018 4:30 am
Location: Kyiv, Ukraine

Re: Cell balancing

Sat May 04, 2019 10:40 pm

Thanks, I may perform the re-initialization, this is why I spend time to install ISTA+ in the first place. Have been researching the downsides for a while before I pull the trigger.

rcy001
Posts: 2
Joined: Sat Dec 22, 2018 3:46 pm
Location: Netherlands

Re: Cell balancing

Sun May 05, 2019 4:13 am

I found some interesting information on
https://www.newtis.info/tisv2/a/en/i01- ... 1VnXGFKEDA:

The SME regularly wakes up from its rest phases and compares all the cell voltages to one another. Since the adjustment of the cell voltages can only take place via a targeted discharge of the individual battery cells, those with a clearly higher cell voltage, compared to the weakest battery cells, are the ones which are selected. By a request via the Local-CAN to the cell supervision circuits belonging to these battery cells, the discharge is started and performed, until the voltage level has been adjusted. The discharge current flows via an ohmic resistance, which is integrated in the respective cell supervision circuit.
The adjustment of the cell voltages is thus a lossy process, but still useful and necessary for maximising the range and service life. The adjustment of the cell voltages takes place completely automatically during the vehicle standstill.


So when you are not driving or charging, the cell balancing will always happen.... As you've observed!
2014 Rex
CCS, 428, Andesit
Suite, H-K, Sunr

flyboy320
Posts: 13
Joined: Sat Jun 02, 2018 5:38 pm

Re: Cell balancing

Sun May 05, 2019 6:17 am

rcy001 wrote:
Sun May 05, 2019 4:13 am
I found some interesting information on
https://www.newtis.info/tisv2/a/en/i01- ... 1VnXGFKEDA:

The SME regularly wakes up from its rest phases and compares all the cell voltages to one another. Since the adjustment of the cell voltages can only take place via a targeted discharge of the individual battery cells, those with a clearly higher cell voltage, compared to the weakest battery cells, are the ones which are selected. By a request via the Local-CAN to the cell supervision circuits belonging to these battery cells, the discharge is started and performed, until the voltage level has been adjusted. The discharge current flows via an ohmic resistance, which is integrated in the respective cell supervision circuit.
The adjustment of the cell voltages is thus a lossy process, but still useful and necessary for maximising the range and service life. The adjustment of the cell voltages takes place completely automatically during the vehicle standstill.


So when you are not driving or charging, the cell balancing will always happen.... As you've observed!
Thanks for this, very interesting! I have always advocated not following the ABC mythology since keeping these batteries at a high state of charge unnecessarily is harmful in the long run. People always seem to state (especially on FB), that you should ABC in part to keep the cells balanced. I charge to 100% at most once a month, sometimes once only every two months and the battery is usually around 40%-60% charged, and according to ISTA+ my cells are all the same voltage, perfectly balanced (reading taken about 50% SOC).

Return to “Technical BMW i3 Discussions”