I84RiS

Do you have access to individual cells voltages? Based on the few examples you have posted you are seeing inconsistencies at the SOC level when the inverter switches back to grid. The only constant in the examples is a large load on the battery bank before the switch back to grid. My guess is that the BMS forces the inverter back to grid because of a weak cell reaching the UV protection point due to large load. So basically a battery issue and nothing to do with inverter settings. You can try and test this by reducing the load to something more manageable for the battery and see if the same happens (obviously you will have to discharge for a longer period). You can also see if the same occurs by removing the communication between the battery and the inverter (and using voltage settings) thereby removing the BMS ability to force the inveter to grid. You might however end up with a battery that has gone into protection mode should it really be due to weak cells.

The batteries generally swell because they are being charged at to high a voltage. The Pylontec batteries seems rather sensitive to this (based on the number of posts floating around on this exact issue which has started to pick up in quantity in the last few months). I am pretty sure the issue is not just impacting Pylontec, my view is that we just seeing this earlier (compared to other batteries) given the use of pouch cells in Pylontec batteries. What voltage are you charging at and what inverter are you using?

Do you have access to cell voltages and BMS data? If you do, post them here. Otherwise not really possible to accurately troubleshoot

What is the 1st slot in the time of use setting? It should start at 00:00 (1st time slot should start at 12 midnight and the final time slot should end at 11:59pm.) Post a picture of the TOU setting taken from the inverter screen.

Was the 1st picture of the battery stats taken at the same time? If you look at the picture you will note that the batteries are at different voltages which is odd for a parallel connection. Does any of the cells ever reach 3.65V during charging which will trigger the OVP? This might explain the different pack voltages in the 1st picture showing battery stats. This might also help explain the different rates of discharge. Looking picture of the battery stats showing uneven discharge it seems like one of the packs is supplying the load while the other is still sitting idle at 100% SOC. This is a common symptom when there is a OVP event since some BMSes (don't know these batteries, not sure if it applies here) will keep the battery at 100% SOC for as long as possible to help with cell balancing. The PACE BMS is one that does this. It is likely that your charging voltage and float voltage is to high. Reduce it as already mentioned is the earlier post.

Yes, this is spot on If the passive dissipative balancer with small resistors built into most of bms we see is not enough to keep up with the balancing of low current cycling often seeb in home PV systems then you likely have unmatched or defective cells. Using active balancers will only mask the problem, it wont fix it. If the cells are not highly matched in capacity then you will always run into either top OV or bottom UV issues no matter how often the cells are balanced. The only solution here is to use the batteries, by example in the 20% to 90% SOC range, and just accept that you have less capacity. Hitting cell OV and UV frequently is not good to the health of the specific cell and will likely lead to premature battery failures once the cycle count gets high (this forum is full of such posts)

Also check any fuses on the DC side (maybe replace them). To rule out issues with the cells you will need access to cell data. During a normal evening where the load is low, to what SOC level does the battery discharge overnight? Does it ever discharge to 0% instantly (as opposed to 6% in the OP) this would hint to a cell issue.

Perhaps you changed the cell and or pack overvoltage level in the BMS settings?. The PACE BMS is one of the most user configurable BMSes in use which is why so many different cell manufacturers use them in their pre configured batteries. Also one of the few BMSes that can limit charge amps. It uses a passive balancer, so like most BMSes that use the same it cannot and will not resolve large deltas in cell voltages.

The 10 year warrantee was on the cells only, if the BMS is really the issue you are out of the warrantee period in any event and the seal being broken becomes irrelevant

It is a PACE BMS.

Have you checked what the State of Health reading on each battery is?. If the SOH is lower on one then the full capacity will no longer be what the original full capacity was when new (by example a 100amp batter with a SOH of 80% will now be a 80amp battery). The 80amp will now be the new reference point for SOC calculation (as well as the LED SOC indicators) and will give the impression that the SOC fall faster compared to those still at 100% SOH during discharge.

The only way to accurately determine what is really happening is to get access to the BMS to see what is happening to the individual cells in each pack. There is no other way to accurately troubleshoot this. It can be that pack 1 is never reaching a point during charging where the BMS resets to 100% and therefore appears to be at a lower SOC when looking at the charging lights(the lights are not an accurate way to determine SOC). It can also be that battery 1 is carrying more of the load as a result of resistance differences and therefore shows a lower SOC. Cell balancing issues might also cause SOC differences. As long as each battery charges to a point where the BMS resets the SOC to 100% regularly (based on BMS readings and not the charging lights) I would not worry about SOC differences as long as the batteries are always at the same pack voltage and the voltage differences between individual cells are acceptable based on where the battery is in the voltage curve. Are you using BMS communication to manage charge and discharge settings? Perhaps physically change where each battery is installed (swop 1 and 3 by example) to rule out any installations errors (incorrect cable lengths, incorrect tightening of lugs etc) and see if the behaviour changes.

Mine works exactly as intended (also using AGM V setting) and will use Grid from the relevant set point that has been set in the the Time of Use setting. I would however not recommend using the Low battery setting and the Shutdown setting to achieve what you want, rather use the Time of Use setting to achieve what you want. The Shutdown voltage, once reached, will cause the inverter to shut off completely (all power to the house will be lost) and it will only restart again once the restart voltage is reached. I can only imagine that you have confused the inveter by setting the low battery voltage point lower than the shutdown voltage (I just checked on mine and it does not allow me to enter such a setting) which is why it did not shut down (fortunate for you) as intended.

That looks like a Pace BMS

Might be a setting in the BMS that only allows the battery to charge when say the SOC drops below 95%. Try to simulate the behaviour earlier in the day and gradually drop SOC more to see what happens

The standard setting on a Pace BMS is for the SOC to reset to 100% when 1. Any cell reaches 3.65v; or 2. the combined cell voltage reaches 56v (16 cell battery) and the charge current is less than 2A With Pace BMS tools (PBMS tools) version 2.4 (must be v 2.4) you are able to edit these setting to your own preferred settings (you will need an RS232 to USB cable with the FTDI chip). I have set mine to reset to 100% at a combined cell voltage of 55.5v or 3.6v for any single cell. If the SOC differences are gradually diverging over time it is likely caused by the BMS of one of the batteries not resetting to 100% regularly.

Agree that they should share the current proportionally, but what I have noticed from experimenting with different brand LFP batteries (all 100A) in the same system is that there are differences between them resulting in different discharge and charge rates between the different brands. During charging, one battery might reach 100% SOC while a different brand is still on 95%. In this case the full current is passed to the remaining battery which might (or might not) exceeds the recommended charge current. The same happens during discharge. I used to manage this with SA by tapering down the charge and discharge currents at certain SOC levels. Here is a recent example, yellow and red are different brands to blue and green (so 3 different brands in 1 system). All have PACE BMS. I am reading the data into SA through the RS485 B port on the 1st battery. Dip switches were set slave 1, slave 2, slave 3 and slave 4 (no master). It does not matter is which order I connect the 4 batteries to the busbar, the result is always the same.

How many cells in the 200 amp battery and how many cells in the 100 amp battery? If this is the same you should be able to use them together with a few limitations. What are the recommended charge current of the two different batteries, you need to set the inveter to the lower charge amps of the two. Same with discharge amps. What is the recommended charge voltage of the two different brands?. If you use BMS communication between inveter and 200 amp battery the charge voltage will be based on BMS instructions from this battery so you would need to be comfortable that this is in spec with the charge voltage of the 100 amp battery. You will have some imbalances in the pack especially if one battery reaches full charge before the other that might cause cell balancing issues. Best way to find out if this is the case is to use the system and monitor individual cells closely.

On the Synsynk, to achieve what you want to do, you have to enable Time Of Use, set it to 100% and tick Grid charge. This will keep the battery at 100%. Deye would work the same.

One month shy of 4 years is how long it took for me to get the final commissioning approval letter. It includes this statement Retrospective compliance will be required for existing installations whenever regulations (e.g. the South African Grid Codes, NERSA, OHS Act, etc.) and standards (e.g. SANS, NRS, etc.) are amended.

Depends on what you mean by equalization charge. If you hare referring to charging the batteries at a higher voltage / deliberate overcharging which helps to remove sulfate buildup then the answer is no. A lot of people use "equalization", cell balancing and even battery balancing interchangeable which often caused co fusion. I have read through the entire post, but suspect the OP is referring to battery balancing.

Battery 2 is at 48.5v while the others are are 48.4v which is likely why 2 is discharging into 1 since it has a higher pressure (and perhaps into 3 and 4, althoug not to sure by what you mean that they are on different strings since the word string is an American term for items connected in series). The system seems unbalanced when comparing SOC and Volts between the battery packs. When last ware the batteries charged to full based on cell voltages for the cells in each pack to balance and for the SOC measure on the BMS to reset to 100%?

Press it and hold it in for 3 to 5 seconds and see what happens

Impossible to say without more information. Did you dowoad the logs before sending in the battery? If you did post them here and let's see if someone can assist. The pouch cells used by Pylontech seems to be more sensitive to being charged at a to high voltage (when compared to a prismatic cells). This is just my personal opinion.

Then don't issue it.