I'm having some issues with my 100Ah LiFePO4 Smart battery (no bluetooth signal, connected SmartSolar charger gave voltage as 4.2V) and I'd like to take a look under the lid. I've seen it's been done before (https://community.victronenergy.com/questions/20875/reconditioning-victron-lfp-batteries-cell-voltages.html), but there seems to be a chance of damaging the electronics as the goop rips off with the lid. Any advice? Would heat help soften the adhesive?
The lid is glued onto the battery. If you have a large flat screwdriver you can put it into the holes under the lid and just crack the battery open. Remove the battery poles first. The electronics sit on the top of the cells, not on the lid.
It is for sure never ever advised by Victron to do that, but if it is your battery, nobody can stop you from doing this either. Apply common sense!!!
But this will be considered as a modification and is therefore moved into the mod space here.
BR
Thank you very much for answering, I certainly didn't expect it after this time. The holes on the side allowed access with multimeter to check the individual cell voltages, so I didn't need to take the lid off after all. 0.5V, 0.5V, 2.0V, 2.0V...
Next question is what could have caused this? I had miniBMS with load and charge disconnect wired in. No DC loads except for miniBMS itself, SmartSolar bluetooth module, SmartLithium bluetooth module.
Most likely, this was caused by self discharge of the battery. If the battery is left alone for a longer time, with little state of charge, this can happen. But to answer your question more concretely, you would have to supply more context what exactly happened in your case and in which timeframe. You could try a rescue attemt to charge the battery cells with very little current until the cell voltages come back to normal, but with that little cell voltages (0.5V) left, the chances that this will work are small. But be very careful and apply supervision all times and if you have doubts, to do this yourself, don't do it. Damaged Lithium batteries are not the best material to play around with...
Even if that is going to work, battery cell chemistry has changed and the battery is not considered safe anymore.
Next question is what could have caused this? I had miniBMS with load and charge disconnect wired in. No DC loads except for miniBMS itself, SmartSolar bluetooth module, SmartLithium bluetooth module.
You have your answer in your question.
Try to individually charge (with a lab DC power supply) each of those two cells which have 0.5V, with very small current (1A for your 100Ah cells) until you get them to the same voltage level with the other two cells (2.0V). Then keep charging all cells individually (or now maybe the entire battery at once) with the same small current, until all cells get to 2.6-2.7V. From there (if you get there) you can charge the battery normally.
Thanks, seb71, for replying. I just managed to go to the remote installation and get the data off the SmartSolar. It's pretty clear that overdischarge due to self-consumption is at fault.
https://docs.google.com/spreadsheets/d/1rD1U8CY_fLX56mNbrWMWepArLXwg1RxDTiTr0P9Gtsc/edit#gid=1322837552
The battery was in cold temperatures, hence no recharge on most of the 30 days. I did the math and arrived at a system self-consumption between 75 mA - 150 mA, which was more than enough to do it in.
As for your suggestion to charge cell-by-cell, would I need to actually detach the cells from the battery pack to do this? Because AFAIK the cells have some connections with resistors across each others terminals for balancing purposes.
Hi @Artur Pata,
I just had a closer look to your LiFePO4 capacity chart.
For lithiums it is generally wrong to conclude from battery voltage to state of charge. This might work ok for lead acid chemistry but definitely not for LiFePO4 or lithium chemistry in general.
I can give you a live example:
My SmartLithiums get charged from a SmartSolar. This morning state of charge was 73.7% and battery voltage at 13.21V. After sunrise charge took place with 440Wh total until sunset.
State of charge went up to 77.4%. In between charging battery voltage went up to 13.46V @peak charge current.
Now, 3 hours later, battery voltage is still 13.32V and SoC 74.1%. Discharge current was always the same at 1.5A.
As you can see now we have two different voltages but more or less the same state of charge. And the more you charge in between, the more difference you'll get between calculated state of charge and measured battery voltage.
Conclusion:
You need a battery monitor (for example Victron SmartShunt, Victron BMV-712) that counts the amp in and out of the battery and then calculate state of charge based on that.
Thanks for the reply! I am aware of the inaccuracy of voltage / SOC relationship for LiFePO4. It works for the big picture though: below 11V very low, above 13V not bad.
I would add a SmartShunt for the improved system.
Ideally, though, I'd spec the system so every Ah need not be counted, rather the data would for diagnosis only. In my climate, that would mean a truly massive solar install or a wind genny, which I'm on the fence about.
It seems that your system has gone through a situation that I am trying to work out how to protect against. When load disconnect is triggered, there are devices in a remote system that will continue to draw power and keep drawing the battery down. Smart Solar controllers! My solar charge controllers are drawing about 0.2A on a 24v Smart Lithium Bank (2 x 12.8 batteries in parallel) at night. If the battery is unable to get charge (say due to cold, or some other issue with the panels) and one is unable to get to the system, how do we isolate everything?
I have been thinking of putting something like a Cyrix-Li-Load on the Smart Solar controller battery connection. This could be triggered to shut down, and this would power the solar charge controllers off and remove that parasitic load. Only to be powered up again when the batteries warm or some other manual action occurs. What are peoples thoughts?
@bathnm That is indeed a question worthy of its own thread entirely!
According to the datasheet, the self-consumption of the SmartSolar controllers should be around 20-35mA. 0.2A for two seems way high. How did you measure this?
The problem with disconnecting the SmartSolar entirely is that it messes up the time measurements. As an example, I removed the faulty battery 2 weeks ago from my system. Yesterday, I repowered the SmartSolar with a couple of AAA batteries in series to get the data, and it showed as if the battery failure had happened the day before.
It is the current draw that is reported by my BMV at night. When I physically disconnect the Solar Charge Controllers it drops to 0A draw, and only thing on being the BMV712. If one is on then the BMV reports 0.1A. Yes I agree it feels high. However 35mA, over a 24 hour period is still 1.75A for two devices. That is almost the reserve capacity of 2 Smart Lithium 100A batteries. I am luck that I have 4 x 200A (2S2P for 24v) so a slightly larger reserve available.
It still dose not get around the issue that during lock down I cant get to my boat. Also during periods when causing we may leave the boat for weeks and not be able to easily get back to her.
Corner case situation, but something that does need thinking about. I have read two posts in teh last few weeks about over discharge of Smart Lithium Batteries due to parasitic loads and failed charging systems.
@Bathnm What's your BMS?
A couple of things could have saved me.
First, more diligent visiting of site. Just a couple of days earlier and I could have caught this. Please learn from my mistake!
Second, remote monitoring so I would know from afar when I need to visit the site. This is something that adds quite a bit to self-consumption, but could allow alarms etc. I kinda wanted this for security anyways.
Third, some sort of dead man switch which would disconnect batteries from the rest of the system at around 25% SOC and require manual action to reconnect. That would still leave them open to overdischarge via self-consumption (Bluetooth module). According to datasheet, they are storable for 1 year at full charge.
@Artur Pata, Agreed the dead man is what I am thinking about. I can remotely monitor, but if thst dead man is needed I cant get to the boat to do anything due to lock down. Also in the future the boat and I may be in different countries and getting back could be a challenge.
Have you thought about how you might add the dead man in. Triggering could be done using the Node Red capabilities that are coming.
