Max /Min Cell difference

Hi I need advice,
I have a Victron setup with a BlueNova battery
From time to time I get Over voltage alarms and sometimes a Low battery alarm,
I believe it is related to the cells not being balanced,
When the battery is almost fully charged there are a few cells that reaches max voltage before the other and that results in the over voltage… and then again when the battery is not yet fully discharged ( 20%) then some of the cell already reaches the lowest voltage and the battery switches off to protect itself.
On a normal day the cells are closely matched , but then soon it goes out of sync



Is there any advice on how I can balance the cells?
I did have the complete battery with BlueNova, and they did charge the cells with a very low amperage until they all “bleed at the same time” so, the cells were balanced about a month ago, but then as soon as I start to use Victron ESS : Optimized with BatteryLife = 30%; then the cells start to go out of sync.
What seems strange is , that for a portion of the day they are all very closely matched, but then as they get closer to 80 - 90% then they start to go out of sync… (see screenshot of the graph)

The voltage of lithium cells are very constant throughout the capacity of the cell and only drop/increase at the ends. That is why we need to count the energy flow, as we can’t look at the voltage of a cell to find how full it is.

So while all your cells are between 20% and 80% their voltages will look all the same. You can only see the difference when you need full capacity and the one cell’s voltage will start moving up while the other are still staying lower. (or the other way around when the battery is empty)

Cells don’t age the same, and as they get older (usually in cycle count not just in years) their capacity might change differently as well. So your one cell might reduce from 100Ah to 98Ah and the other one to 95Ah. Which means the 95Ah cell will be full sooner.
You don’t expect to see this in a new battery, but as they get older this is normal.

You have 2 options to mitigate this.

  1. Reduce the charge current as the battery gets close to full. This will give the BMS balancers time to work on high cells. If your BMS does not support this feature, then you can try to add external balancer that can balance larger currents.
  2. The other option is to reduce your charge voltage. This will also give the balancers time to work as you are not trying to reach the high voltage where your cell needs to be protected.
1 Like

Optimized without BatteryLife.

Presume you have set your inverter and MPPT to the exact specs provided by Bluenova?

1 Like

I’m going to try this first, as it is the easiest and fastest… just one click…
How do I check the setup between my MPPT and Blue Nova, as it was originally configured by the Co that sold me the setup
If this doesn’t work… do you know if it is possible to set the battery charging voltage on a Victron setup, as Louis suggested?

image

1 Like

Give it about a week.

Assuming your inverter/MPPT is set correctly, then look under DVCC to control it there manually. You can set the max volts/amps there.

1 Like

I don’t think your battery is unbalanced at the top. It looks perfectly okay to me.

It is normal for one cell to spike up before the others. It is inevitable really, physically one of them must have a slightly smaller capacity than the average, and that one will spike up first. What you need to look at is the voltage of the rest of the cells at the time this happens.

If the rest of your cells are all above 3.4V when this starts to happen, there isn’t that much of an imbalance. And in your case, it actually looks good:

But if you have a battery that is giving overvoltage alarms, then you need to lower the charge voltage, as @Louisvdw already indicated, but also critically, you need the battery to spend some time at higher SOCs so the battery can balance.

Edit: Lower the max charge voltage, but set it high enough that you still get that spiky behaviour. You need the imbalance to show up for the balancer to work.

What I normally do is take n-1 times the average cell voltage, plus the highest cell voltage (where n is the cell number in the battery, 16 in this case).

For example, it looks like the battery aims for 3.45V at the top, and at that time the other cells are at maybe 3.42V.

So 15*3.42V + 3.45V = 54.75V, so try setting it to 54.8V and see how it goes… lower it if needed. Look at your VRM charts every day and make sure you see that one cell rising out. Increase charge voltage limit over time until you can remove it completely.

I have seen almost this exact graph many times. The area you have circled in red is before the battery is full and there is a fairly high charging current. It looks like all the cells’ voltages are climbing nicely together, but in fact quite a bit if that is voltage drop due to the high charging current. Once the battery is nearly full and the current starts decreasing then you see the voltage of the lower cells decreasing, eventually settling at some low voltage when the charging current has tapered all the way to 0 (or close to 0A)

Edit: Here is my battery cell voltage graph from yesterday, doing exactly the same thing.
I need to put a few Ah into Cell 14. (My bank is made up of a bunch of random used cells in various states of health)

Thank you guys for these valuable suggestions.
I now see that my Victron screen differs slightly.
Mine allows me to limit the Current, not the voltage…


and if I switch of the [Limit Charge current] , it doesn’t show the Voltage option.

What version of VenusOS are you running? That feature was added in Venus 2.65.

Oh my how would I read that, and also how do I update the version if I want to ?

Settings->Firmware, right at the top.

Go to online updates to see if there is a new one, and you can update from there as well.

Some devices needs to be taken to 2.73 before they can be upgraded the rest of the way. But I have no idea what you’ve got there. The missing option simply suggests that you must have something old… because I know of no other reason that setting would be missing.

Thanks, am upgrading to Firmware v 2.91 as we speak.
The Tech from Blue Nova says Maximum Voltage 55,8V and Float 55,2V
Once on the new firmware, should I set the value 55.8 or 55.2 or should I set it to 54.8 as you have suggested?

You won’t break anything by starting low and then moving up after a while of it working well until you get where you want to be.
54.8V /16 is 3.425V per cell. I see anything above 3.45V as full, so this is a good place to start
55.2V /16 is 3.45V per cell and bank on the money where you want to be. My serial battery driver aims for this value.
55.8V /16 is 3.4875V and still well below the normal protection voltage of 3.65V

I would start at 55.2V. If alarm still shows up, move it down by 0.1V until it stops. From your chart this might be as low as 54.7V.

The objective is to get it back to the point where you don’t have to override the charge voltage, so watch your VRM chart and once you see the highest cell no longer exceeding 3.45V, increase the charge voltage by 0.1V.

You want to hold the highest cell within the 3.45 to 3.5V window, so the balancer can work at it.

I have upgraded to the latest Firmware version and set the charge voltage to : 55.4v
image

All seems to be much smoother.
I’ve had no alarms since we last spoke.

But I am still baffled by the strange way that the Max/Min cell voltage seems to run in sync and then suddenly they start to diverge…
See in Image after 10 am today


why would that be, and should it bother me, or is this natural and I can ignore it?

Lithium cells always spike up when they get full. Even a few milliamps can drastically push up the voltage. To give you an idea, between 3.45V and 3.6V there is not even 1% of extra energy storage. That means even minute differences in capacity can cause that dramatic rise in voltage that you see.

What usually happens is that the battery will top-balance. In other words it will slowly pull the other cells up to a higher SOC so that the voltages stay closer together at the top, but you will still always see one of them jump out a little. That is also how the BMS knows the battery is full :slight_smile:

For your battery, it does seem like the lowest cell is only at 3.35V at that time. That means there is a slight imbalance that needs to work out. Keep monitoring it, it should get better over time.

Edit: In fact, I would increase the charge voltage a little already. Try 55.5V tomorrow, if it still avoids an alarm, then by next Friday take it up another 100mV.

I got an Overload Alarm again this morning.
It happened just when the Charge Current limit and Discharge Current limits spiked.


So, I have limit the charge Voltage, but currently the Limit Charge current is not on.

Should I switch it on? If yes, what value will be acceptable

That sounds like something unrelated to the battery.