Adding a second LiFePO4 battery with different capacity cells. -> Will it work?

[HDB]

Hello everyone,

I've been helping my parents with their off grid system in Spain.

If you want full background, check out these 2 links, the first one is quite Victron specific, the second link is basically the same as what I'm about to ask.

First thread:

Why is this system not fully charging the batteries from AC PV? - Victron Community

community.victronenergy.com

Second thread:

Add a second LiFePO4 battery bank - is my schematic ok? - Victron Community

community.victronenergy.com

So, my question:

My parents have 6 US2000 Pylontech LiFePO4 batteries, 15 cell each (which is rather odd). All are connected in parallel, this is the schematic :



I want to add a DIY LiFePO4.

Thanks to the Victron feedback (check second link above), I learned that the US2000's are 15 cell instead of the "normal" 16 cell.

Therefore: I want to buy either 15cells or 30 (15s or 15s2p) LiFePO4 200AH.

So basically I would go for:

15 (or 30) of these cells: link (common sense would advise 30 cells, but the bank account might advice differently)
And a Daly BMS (because YouTube is full of great Daly reviews, must be good), this for example. I'm not sure if I'd need 2 BMS's or 1 if I go for 30 cells (15s2p), but I'll figure that out later.

Anyway, my schematic would look like this:



Note: I would use 15 cells, not 16 as mentioned on the schematic

I would just parallel them to the existing batteries since there are vacant connection holes on the bars:



The real question is:

• Would this work? Just paralleling the new battery onto the 2 bars (plus and minus). I would definitely say yes.
• Would it harm the Pylontech batteries? I think not, because all 6 pylontech batteries are protected by their own BMS. And Victron limits the current to 140A anyway (setting in Victron which I could modify if need be).
• Would it harm the DIY pack? I think not, since it's capacity is bigger, and each cell is rated arount 200A continuous discharge current, 140A should be fine (as configured in victron). Also Daly would protect it if I take say a 150A BMS.
• I hear talk about it not being healthy to combine smaller & bigger battery packs, but I disagree: Two interconnected watertanks (same height) will fill evenly no matter there volume. Therefore, following the water analogy, both batteries will be full at the same time (provided that each cell in both packs equals 3.2V, resistance is the same, etc.).
• Can i mess anything up? I really thought it over, watched "all" off Will's videos (and others), and think all will be fine.

But you guys are the experts...

Let me know, thanks.

Regards,
Hendrik

 

[Luthj]

Would this work?

Yes.

Would it harm the Pylontech batteries?

No. More importantly you need to make sure your system is fused/breaker properly for the configuration.

Would it harm the DIY pack?

No, not unless your existing charge/discharge limits are outside the new batteries limits.

I here talk about it not being healthy to combine smaller & bigger batterie packs

Water analogies are rough at best, and gloss over some important chemistry. That being said most of the concerns have to deal with High C rate packs, or lead batteries. LFP is quite forgiving, and doesn't suffer from partial state of charge degradation.

Will's videos are good for beginners, and some intermediate DIY stuff. However you have a complete off grid system which adds additional complexities and risks. If you are unsure of wire gauges, fusing, crimping, etc, please consult an expert, or ask here. It is possible to start a fire. For example breakers, fuses, and wire terminations need to be contained within an electrical rated box, or at least kept a safe distance from flammable materials.

Is your parents system built to the local electrical code? It probably isn't based on the boxes, bars, and mounting in that photo. You might consult the relevant code to see how the system stacks up.

 

[HDB]

Yes.


No. More importantly you need to make sure your system is fused/breaker properly for the configuration.


No, not unless your existing charge/discharge limits are outside the new batteries limits.


Water analogies are rough at best, and gloss over some important chemistry. That being said most of the concerns have to deal with High C rate packs, or lead batteries. LFP is quite forgiving, and doesn't suffer from partial state of charge degradation.

Will's videos are good for beginners, and some intermediate DIY stuff. However you have a complete off grid system which adds additional complexities and risks. If you are unsure of wire gauges, fusing, crimping, etc, please consult an expert, or ask here. It is possible to start a fire. For example breakers, fuses, and wire terminations need to be contained within an electrical rated box, or at least kept a safe distance from flammable materials.

Is your parents system built to the local electrical code? It probably isn't based on the boxes, bars, and mounting in that photo. You might consult the relevant code to see how the system stacks up.

Thanks for you reply.

Thickness of wire, breakers/fuses, etc. are no problem. I'm more than familiar with all of that.

Local code: not an issue.

Just sharing my plans and "using" you guys to check if I don't make any huge conceptual mistakes.

 

[Luthj]

Your diagram is a bit too small to read. Either post a higher resolution one, or break it into pieces.

 

[HDB]

Your diagram is a bit too small to read. Either post a higher resolution one, or break it into pieces.

Check out the attachment below or click here. If this doesn't suffice, then I'll vectorize the image.

 

[HDB]

And a Daly BMS (because YouTube is full of great Daly reviews, must be good), this for example. I'm not sure if I'd need 2 BMS's or 1 if I go for 30 cells (15s2p), but I'll figure that out later.

In case that I opt for 30 cells:

1. Would it be best to use one BMS and parallel them by 2 cells each;
2. Or would the better method be to treat them as 2 packs of 15s's and thus use 2 separate BMS's?

Both have benefits i imagine: more/less cable, double the price, more elegant,... But to trick a BMS into thinking that it is managing 15 cells, but actualy managing 30cells (in case 1) seems wrong. Not 2 cells are 100% identical (even though you pair them as good as you can) and I would guess that option 1 would fail faster than option 2.

What does reality/experience say?