I’m running a stack of 5 × US3000C units on a Victron 3ph-ESS.
Until now, the batteries had behaved more or less identically, despite having been purchased at different times. A few days ago, I used the remote update function to bring the firmware on the stack up to the same version.
Everything was still unremarkable afterwards. Today, one battery (the middle one) behaved strangely; see screenshot:
While charging, the cell voltage rises until one cell exceeds 3.5 V. The BMS then appears to switch off, and the voltage of all the cells drops to around 3.35 V.
The other four Pylontechs behaved as usual. The affected unit has never done this before; today was the first time.
Should I be concerned, especially as I’m going away on holiday for two weeks tomorrow morning?
Possible causes? A one-off anomaly or the beginning of battery failure?
Hi Jens, presumably the middle one is stopping the charging because one of its cells is “shooting up”. I had the same problem. One or two initial charging cycles might help here.
For over 3 years, I ran 4 × 3000C blocks with 3 × MP2 3000s, with regular charging at over 130 A, so right at the limit. I tried to “heal” the faulty block as gently as possible, but even with a charging current of 2 A, one cell kept shooting up again. I waited too long.
I would completely discharge the battery straight away and charge the 5 blocks over many hours at around 10 A, then keep them charged for a few hours. If necessary, this can be repeated once or twice.
I had to replace the block and increased the capacity so as not to put too much strain on the batteries.
What do you have for the 3 Multis? I suspect that in your case, too, the capacity was undersized from the outset.
Victron have adjusted their pylontech charging override algorithm in 3.80, in the late stages of beta. Might be worth an update as it helps limit overvoltage conditions.
It never pays to update firmware unless you have a real issue. This is true for inverters, more so for batteries.
You could take a quick look over the casing, just to be on the safe side. If the cells have swollen, you should already be able to see it from the outside of the casing. But that’s really just for peace of mind — I don’t think you have a serious problem there.
You can trigger and monitor the initial charge via VRM. In fact, your holiday would be a good opportunity to do it, as no significant power consumption is expected and you can patiently carry out a few gentle discharge and recharge cycles.
I thought I had to explain it for a layperson, and then he goes and sets up a switch in Node-RED .
That’s a good start for discharging it. Ideally, you’d use an even load—especially when it’s nearly empty, don’t start the instantaneous water heater. Take it down to 10% slowly.
On the remote console, under ESS/Power feed-in, I’d disable DC excess feed-in, then go to DVCC and set the charge current limit to 0 while discharging.
To charge it, set the charge current limit to 10 A and, under ESS, set the mode to Keep batteries charged. Once the battery is full, it’s fine to leave it there for a few hours, depending on the cells’ voltage difference.
Then discharge it evenly again and charge it gently once more. Always leave DC excess feed-in disabled. You can also leave the MPPT switched on and set the charge current limit to 0 when discharging. The battery widget will then show #6, which means that the charge current is set to 0.
So, just to wrap things up for now: after discharging it to 15%, then charging it at 10 A and letting it stand for around 12 hours, the battery seems to be behaving normally again for the time being.
.
and have a lovely holiday.