Gel batteries capacity on 25% after 2.5 years of light use - what am I doing wrong?

[i_papp]

GEL batteries need a long absorbtion period where the charge voltage is held constant at the absorbtion voltage of 14.2 volts maximum. The default absorbtion period for Epever is 2 hours, this is lower than recomended for some GEL batteries under cyclic operation.
The manufactures JYC, recomended charge voltages of 14.4 to 15 volts, these are higher than some other manufactures recomend for GEL, Epever default charge voltage for example is 14.2 volts for GEL .
It's suggested by some that GEL batteries should be charged at 14.2 or 14.1 volts to reduce early failure. (2.35 volts per cell )
Without knowing the typical solar yield compared to daily load its difficult to suggest the batteries were
under charged. For example is there shading on the panels for some part of the day?, is the inverter idle current accounted for? are there often several days of poor solar input, cloud/rain?

Assuming for most of the time the only load is 5 watts at the battery, it's possible at the daily charge at 14.4 volts , the batteries were over charged. Epever don't have variable absorbtion period so even after a slight overnight discharge the battery was subjected to a prolonged charge.

My guess is the charge voltage of 14.4 caused degradation despite the manufactures advice for this value.

Mike

We are in the weekend house only when the weather is nice, and that means sunlight to charge the batteries.
The only thing that could happen is that the fridge is on while we are not there.
The daily consumption of the fridge is up to 350 Wh, so it would take ~4 days to reach 50% DoD.
And if discharge would be prolonged, sooner or later the inverter would switch off. But that never happened.

Since I couldn't explain a deep discharge of the batteries, neither my calculation indicates that this is the case, currently the too-high charging voltage seems the most plausible cause.
It means that setting the boost voltage for cyclic use (or even higher than that) for a battery that is very rarely cycled degraded it. Honestly, I expected the SCC to adapt the charging process to the battery state, but now I realized that it just blindly forced the absorption for a fixed period of time (2 hours) at a too-high voltage. The degradation would have been less if I'd lowered the boost voltage to 14.1V or 14.2V.
Is this a good summary?

That makes me wonder - the lead-acid technology is so fragile that a 0.2V or 0.3V higher voltage can make such a difference?
Istvan.

PS: I hope the LiFePO4 battery will not suffer from this issue (Link)

 

[Mattb4]

...

That makes me wonder - the lead-acid technology is so fragile that a 0.2V or 0.3V higher voltage can make such a difference?
Istvan.

PS: I hope the LiFePO4 battery will not suffer from this issue (Link)

Batteries are chemical devices. They suffer quickly if the parameters for the chemical reaction are not within a certain range. Think of them like a plant instead of a storage container. Overcharging any battery can lead to degradation.

 

[Hank Waconda]

As far as I recall (I don't have the spec anymore) the VOC is around 22V, while the ISC is less than 9A.
Seems that most of the polycrystalline 140W panels have these specs.

This means that in my setup I could:

• connect my three panels into a series, which would generate up to 3x22V = 66V, and up to 8A
• connect the panels to the SCC
• the SCC will take care of converting the input voltage/current from PVs (66V, 8A) into 12V-level output for the battery (14V and current of up to 30A - like a DC/DC converter).
• This configuration would help to extract additional power from PVs in case of a cloudy sky or evening hours.

Is my understanding correct?
Istvan.

Yes, you can look on line at an MPPT charge controller calculator to understand better. You will need you VOC and ISC numbers from your P.V. panels.

 

[mikefitz]

Yes your understanding is correct, although the extra solar energy under low light conditions, whilst useful , is not substantial.
It's only a guess that high volts caused the short service life, it could be over discharge, poor storage and care prior to purchace, or just a substandard battery. Note that the inverter if connected to the battery via the solar controller load outputs, it will have continuous dle (around 10 watts) power drain, even with AC loads disconnected.

Mike

 

[i_papp]

Yes your understanding is correct, although the extra solar energy under low light conditions, whilst useful , is not substantial.
It's only a guess that high volts caused the short service life, it could be over discharge, poor storage and care prior to purchace, or just a substandard battery. Note that the inverter if connected to the battery via the solar controller load outputs, it will have continuous dle (around 10 watts) power drain, even with AC loads disconnected.

Mike

Regarding the cause of my gel batteries dying prematurely, I don't think it is over-discharge, or poor storage during usage. If it would be due to care prior to purchase - I guess issues would show up from beginning. So the degradation is either due to too high voltage, or (I think you introduced a new perspective with it) a substandard battery. As mentioned, I already consumed 2 sets of the same model (from the same supplier), so this is also possible.
I am quite confident now that I'll go for a LiFePO4 battery. My best pick so far is SOK Marine 206Ah (it has Bluetooth and heating) Data sheet.
If you have better alternatives (value/price), feel free to share (I am in Europe, so there might be some limitations in availability).

As step 0, I'll clean my solar panels for a fresh start
As step 1, I would install a SOK battery and set SCC to LiFePO4 mode (default settings are 14.5V boost voltage, float 13.8V).
Apart from these 2 voltages, there is a handful of voltages in Epever user manual, as well as the battery data sheet.
My questions:

1. I couldn't map one to the other, so I would like to ask you - do I need to adjust some voltages in SCC?
2. Will be my SCC good enough for the LiFePO4 battery, even if it doesn't have some advanced charging algorithms? I wouldn't like to kill a 1k EUR battery with a 0.1k EUR charger :-(

After I manage the battery swap, in step 2 I might change the setup from 3 panels parallel to 3 series connection, for better low light efficiency.
I'll go step by step.

A couple of months back I was hesitant to share my issues with a community, and I tried hard to solve them on my own, by chasing the suppliers/vendors. And I failed in that. Now I feel that I found a great, unbiased community to discuss my problems and work out a solution.
So I just want to express my gratitude for your support and prompt responses.
Istvan.

 

[i_papp]

As step 0, I'll clean my solar panels for a fresh start
As step 1, I would install a SOK battery and set SCC to LiFePO4 mode (default settings are 14.5V boost voltage, float 13.8V).
Apart from these 2 voltages, there is a handful of voltages in Epever user manual, as well as the battery data sheet.
My questions:

1. I couldn't map one to the other, so I would like to ask you - do I need to adjust some voltages in SCC?
2. Will be my SCC good enough for the LiFePO4 battery, even if it doesn't have some advanced charging algorithms? I wouldn't like to kill a 1k EUR battery with a 0.1k EUR charger :-(

After I manage the battery swap, in step 2 I might change the setup from 3 panels parallel to 3 series connection, for better low light efficiency.
I'll go step by step.

Thanks, guys for your patience and advice, I've just ordered SOK 206Ah
If you have some hints on setting up the voltages on my Epever SCC, we have 10 days to figure it out!
Istvan.

 

[i_papp]

Thanks, guys for your patience and advice, I've just ordered SOK 206Ah
If you have some hints on setting up the voltages on my Epever SCC, we have 10 days to figure it out!
Istvan.

The battery has arrived, I have installed it. When it arrived, the SOC was 30%, the cells were balanced, and the calculated capacity was 216Ah.
I set the following for Epever Xtra 3210N:
- boost voltage - 14.2V
- float voltage - 13.6V
The next day I checked, the battery was at 100% SOC. Looks good so far, the SCC switches between boost and float charging states as expected.
I am using the ABC BMS app to monitor the battery status.
I'll keep you informed.

Another improvement I made is that I've installed a cheap 35 EUR soft starter in front of my 500W water pump. Without the soft starter, the initial surge power was almost 1kW, while with the soft starter, it reaches 600W max at the start. The reason I did it was due to voltage drop on start if both the fridge and the pump started in recent moments. It might be that with the new battery it is not even an issue, just wanted to share the experience.