Inverter Battery losing capacity

R

Respawn

Member
Joined
Dec 9, 2015
Messages
29
Reaction score
17
I installed a Deye 5kw inverter and two 5kw Volta batteries in June of this year, they were set at 85% maximum capacity but now it is October and it stops charging at 72%. On the 15th of September I noticed one of the batteries which have four green lights showing state of charge was only charged up to three while the other was on four with my total charge at 78%, a few days later the other battery dropped down to three lights and my charge was 75%. Since then after a week I lose another 1% which means at this rate I will lose another 36% by the time the batteries are one year old, so they will be dead in other words.
I contacted the installer and they said it is normal for batteries to lose capacity over time which I understand and expect but certainly not at this rate. Seems like they don't want to deal with exchanging them or testing. The most I have ever drawn from the batteries was 10% during stage 6 so it's not like I emptied the things.
Any experts here want to offer some insights?
 
Maybe charge them to 100%. But I recall there were a few Volta issues sorted with a firmware update.
 
The battery is supposed to lose capacity after the specified cycles are depleted, it says 6000 cycles on the specs, the installer cannot just give you that general answer without knowing all the factors, the battery could be faulty. It has a, 5 year warranty for a good reason, insist on getting the battery tested to see if there is nothing wrong.
 
Yeah insist on an investigation, your installer is talking out his rear
 
A lithium battery, if treated properly, will last 12-15 years

I suspect the BMS is faulty. Have installed 62 systems and in that time, had 2 faulty batteries, both with a similar fault to what you described
 
chrisc said:
A lithium battery, if treated properly, will last 12-15 years

I suspect the BMS is faulty. Have installed 62 systems and in that time, had 2 faulty batteries, both with a similar fault to what you described
Click to expand...
That is as general as what the installer is saying, every battery have its specs, it will say, X number of cycles, at this DoD thereafter the capacity will drop to 70% or whatever.

How long it last depends on the use case, treating it properly is not really a factor as that is all relative, for an example someone who discharges it to only 25% is not treating the battery more properly compared to someone who discharges it to its maximum as stipulated on the specs, they are both using the battery properly, the difference is that their use case is different, the one who discharges it deeper will probably get to the maximum cycles earlier, not because they were abusing the battery.
 
just charge it full to 58.4v so the shunt can calibrate before assuming capacity loss
while charging it full do check the difference between lowest and highest cell voltage

catch it at the point where the battery reaches full just before it lets the voltage dips down to float
checking at float voltage is a meaningless exercise to determine bad or unbalanced cells
 
It sounds to me like SOC drift. It is very common when batteries are not regularly charged to 100% SOC. Firstly set your inverter to charge to 100%. If this does not reset the % to 100%, look on the data plate of your battery for the suggested bulk charge voltage and make sure that you get your battery to that voltage, if it does not reset your SOC to 100% I suggest making both your Float and bulk charge voltages to the bulk charge voltage of the battery to ensure you get the SOC reset to 100%. If it still does not work, increment the Bulk and float charge by 0.1V and give it an hour. If still not reset, increment by another 0.1V. (This is because there can also be a discrepancy in the voltage reading of your BMS and what the DEYE inverter aims for. We have had a similar issue on a similar brand inverter being 0.2V off from our battery-provided voltage).
 
yea bms dependant , some only reset to full if a protection voltage is hit ie either cell or pack

if the cells drifted a cell voltage protect resetting to 100%soc may be incorrect
ie it will say the battery is full and 100% capacity available.

lets say the cells drifted by 2ah and thus the other cells are 98% full testing capacity the max you could get is 98% capacity (very unlikely to be that big but it can happen)

a full battery is actually when all cells are full not when the bms says it is full
hence why i said check cell voltages when full if all cells are close to 3.5v then yea it is actually full
 
Agree, the cells will have some drift, especially when keeping your battery from getting to full level for a long time. The BMS should however within a couple of days of reaching full voltage be able to get your cells all back in line again. I recommend at least charging your battery to 3.4V per cell as this will allow in most cases balancing to happen for the cell or cells that tend to go a bit higher. This will be around 54.4V for 16s LiFePO4 batteries and will also help to not push your cell voltages "too high" for extended periods of time. But it will not prevent SOC drift. Then you can, if they are kept at a voltage that assists with balancing as well, push them to the pack full charge voltage to reset the SOC Calibration at the top end.
 
BESS said:
Agree, the cells will have some drift, especially when keeping your battery from getting to full level for a long time. The BMS should however within a couple of days of reaching full voltage be able to get your cells all back in line again. I recommend at least charging your battery to 3.4V per cell as this will allow in most cases balancing to happen for the cell or cells that tend to go a bit higher. This will be around 54.4V for 16s LiFePO4 batteries and will also help to not push your cell voltages "too high" for extended periods of time. But it will not prevent SOC drift. Then you can, if they are kept at a voltage that assists with balancing as well, push them to the pack full charge voltage to reset the SOC Calibration at the top end.
Click to expand...
i have a shoto and the balancing on the specific model only kicks in at 3.5v
which i find odd most bms's start lower 3.4-3.45 like you said

yea and also having a lower charge voltage like you said will stretch out the CV phase meaning more balancing time if the amps is at least over 1A, some only register as charging over 1A (my shoto does this and only charge balances so a slow charge below 1A means no balancing)
 
I would not be too worried if I were you, Just make sure to get your battery to a 100% SOC at least 1 time per week and this will aid the balancing a lot! Because you don't want balancing to drift too far over time as then the time it takes for the BMS to do balancing increases a lot. Depending on how the software of the BMS is written, running into protection on a cell voltage may or may not negatively influence the amount of time available for balancing on each charge. Rhus getting it to 100% at least once per week will just be beneficial in the long run :)
 
As long as the drift is addressed from time to time

The problem of allowing it to drift can cause abuse to certain cells, negatively affecting pack capacity

ie some bms's have ridiculous protection values per cell

Allowing it to drift subjects certain cells to this abuse

ie the shoto i have only stops charging if a cell hits at 2.9v and stops discharging if a cell hits 2.2v at the bottom

So if you have drift it abuses cells at both ends or just one dependant on usage patterns

And even if it has nice values ie 3.65v at the top and 2.5v at the bottom, you essentially degrade certain cells quicker as LFP does thousands cycle less if you cycle close to empty

So example some cells Has drifted over years of 20-80 usage

And have 15Ah lower soc than others

So even though the pack capacity is shown to only drop as low as 20% some cells might cycle down to 5% SOC

The 20-80 notion in theory is the most idiotic imo

That bullshit is propagated with very little insight to how these bms's function

Sure you can do 20-80 and extend the life

but then you can't just leave it on that setting from time to time you have to charge full for shunt calibration and balancing

And know how to identify imbalanced cells and actually check for drift address it and then use 20-80 for a while
(imbalanced cells will look balanced and only show its imbalance at the very bottom and top)

Rinse repeat

20-80 done wrong will do more damage to pack capacity than 20-100 usage imo
 
Last edited:
Why on earth would you limit it to 85%?

That’s the problem right there, it can’t calibrate itself.
 
You also want to check the warranty on the Voltas. Most exclude stanby operation and are limited to daily cycling usage, not familiar with the Volta warranty, but worth checking.

Remains very odd not to charge to 100% at every occasion
 
My Volta Stage 1 did that a week or so ago, still waiting to hear back from them on how to get the firmware update done.

image.png
 
Rouxenator said:
My Volta Stage 1 did that a week or so ago, still waiting to hear back from them on how to get the firmware update done.

View attachment 1609388
Click to expand...
Yea most likely the shunt was out of touch with reality

It happens , they recalibrate to 0/100 when they hit those SOC

The bms shunt can't count losses, so it kinda goes like this

You have a full battery you use 50% and lose 0.5% to losses balancing heat whatever

Bms thinks battery is 50% actual 49.5%

You charge 40% bms thinks 90% but you had losses again so actual is now 89% not 90%

Rinse repeat many cycles while not touching 0/100 battery thinks it is at 50% SOC while actually at 15%

So you use 15% you and bms think you are fine poof low voltage shutdown

All calibrated again and the song and dance repeats itself

If you hit 100% at least every few days , it can correct the displayed soc to be correct

ie when it hits full voltage either pack/cell (bms dependant) the bms will mark it as 100% soc and start counting from there again
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X