Upgrading LiFePO4 280ah 48v to 560ah 2 batteries in parallel or double cells in parallel?

dswiggum

New Member
Intending to double my LiFePO4 280ah to 560ah. Debating whether to do 2nd battery in parallel or double cells in one battery. Chatting with Kevin Zhang at Daly BMS in Shenzhen he seems to think it is better to double cells and use one BMS rather than making a second battery and paralleling BMS. My thinking is to match each new cell with an existing cell (after balancing as close as I can with my limited equipment) connecting 2 in parallel and then 16 in series and using one BMS. Thoughts? Suggestions?
 

Steve_S

Offgrid Cabineer, N.E. Ontario, Canada
Unless you have Perfectly Matched & Batched cells do not put them in parallel within a pack.
By building 2 Packs in Parallel, that adds Fault Tolerance because one can keep going if the other cuts off for any reason. Secondly, they will SPLIT Load & Charge fairly evenly thereby reducing stresses on the packs, so instead of having one pack push 50A to the Inverter, the two will send 25A each.

Each Pack of 16 Cells, gets a BMS, a 250A/300A Fuse connected in Parallel with Equal Length Battery Cables to a Common Bus is the best way to go. See this Document from Victron starting Page 17. https://www.victronenergy.com/upload/documents/Wiring-Unlimited-EN.pdf
 

Ampster

Renewable Energy Hobbyist
I already have an expensive Orion BMS so economically it would not make sense for me to buy another BMS or two, just to get redundancy. The pack I am configuring in the next few months is 3P16S. I do not plan on doing anything other than top balancing all 48 cells.

There are good arguments for either arrangement. Much depends on the use case and the user's preference. I do not think the balancing current of the Orion would make a dent in three parallel cells totaling 840 Ahrs so I will turn off the balancing on the Orion and probably use a Heltec active balancer. I can activate or turn off the Heltec balancing as needed.
 

DLTooley

Solar Enthusiast
My personal opinion: every cell deserves to be monitored on its own.
This is critical with sub-par cells, but less so with quality product. I am not qualified to quantify that relative benefit. I’d presume it might get more important at the latter part of a cell’s lifecycle.
 

upnorthandpersonal

Administrator
This is critical with sub-par cells, but less so with quality product. I am not qualified to quantify that relative benefit. I’d presume it might get more important at the latter part of a cell’s lifecycle.

Something can always go wrong with a cell - any cell - ask Tesla, or GM, or LG, or Hyundai, or Panasonic. If that one cell happens to form an internal short, for whatever reason, it will take the other cells it's in parallel with it.
 

wiseacre

Solar Enthusiast
Secondly, they will SPLIT Load & Charge fairly evenly thereby reducing stresses on the packs, so instead of having one pack push 50A to the Inverter, the two will send 25A each.
This is a good point about paralleled batteries that beginners like I didn't realize at first thought.
If I didn't binge read everything on the forum I would have been surprised when my soon to be built battery shut down because I tried a 2400w load on a 24v/100A BMS. 🤨

At first, I thought one big capacity (AH) was all I needed.
Then I realized a single battery wasn't the cat's pajamas after all.
Redundancy, load sharing and individual cell monitoring are
 

time2roll

Photon Sorcerer
I would tend to not touch the existing battery and add the addition as a parallel battery. This keeps all functioning during implementation and any future service needed. If you continue to expand I would then add cells to the existing battery(s).
 

upnorthandpersonal

Administrator
I would tend to not touch the existing battery and add the addition as a parallel battery. This keeps all functioning during implementation and any future service needed. If you continue to expand I would then add cells to the existing battery(s).
I currently have 2 x 16 cell batteries connected in parallel. I'll add another 2 x 16 this year (well, probably early next), and will couple them in parallel to the existing ones - so 4 batteries in parallel in total. I like the topology, I can swap out entire banks, I can check individual cells in a pack without taking power down, each cell is monitored so I can find potential weak/dying ones, etc.
 

dswiggum

New Member
Thanks for all the input. Leaning toward 2 battery packs in parallel. I recall Will Prowse stating in a video back when that 2 in parallel would work, but no more than 2. Wondering why the Daly rep was so insistent on the 2p 16s configuration. Have become somewhat disappointed with Daly equipment, too many problems, so probably not an issue, but curious as to why he is on the 2P16S bandwagon. Another consideration is that top balancing bms requires charging over 80% SOC to initiate the balance. I am only charging up to 80% and discharging to 20% for maximum battery life. So I have been changing my charging profile roughly once a month for a couple of days to enable top charging, then dropping back to an 80% charge cutoff. Manually checking balance with multimeter after an hours rest for the battery. Balance has been on the mark so far. The ability to isolate one battery at a time to do this without interrupting service is enticing. Leaning toward Heltec active balancing if I can find a Heltec that will ship to the Philippines.
 

MrM1

I'm Here, But I'm Not All There
I just brought online two separate 8s packs each with its own Overkill Solar 100a 24v BMS and paralleled at busbars 2 days ago.

Everything every one has said here is so so true about 2 separte parallel packs. So many benefits did not even occur to me until after I set the packs up and deployed them
As mentioned:
- individual cell monitoring with 2 BMS units
- redundancy with 2 separate 24v battery packs and BMS units
- Overkill solar and the Xiaoxiang App give me ultimate monitoring, control and total BMS tweakability
- The ability to separately turn off Charge and/or Discharge for either pack individually thru the App for service and "just in case" with out having to shut down the entire system
- etc. but you get the idea.

Love Love Love the setup. 2 parallel packs and 2 OverKill Solar BMS units is a really great combo.
 

Shale MacGregor

Solar Enthusiast
I just brought online two separate 8s packs each with its own Overkill Solar 100a 24v BMS and paralleled at busbars 2 days ago.

Everything every one has said here is so so true about 2 separte parallel packs. So many benefits did not even occur to me until after I set the packs up and deployed them
As mentioned:
- individual cell monitoring with 2 BMS units
- redundancy with 2 separate 24v battery packs and BMS units
- Overkill solar and the Xiaoxiang App give me ultimate monitoring, control and total BMS tweakability
- The ability to separately turn off Charge and/or Discharge for either pack individually thru the App for service and "just in case" with out having to shut down the entire system
- etc. but you get the idea.

Love Love Love the setup. 2 parallel packs and 2 OverKill Solar BMS units is a really great combo.
My question for you is, if you had the option of doing 24 16s packs in parallel, or 12 2P16S packs in parallel would you still be choosing to do all packs unstacked?
 

MrM1

I'm Here, But I'm Not All There
My question for you is, if you had the option of doing 24 16s packs in parallel, or 12 2P16S packs in parallel would you still be choosing to do all packs unstacked?
So you have 384 cells? I would tend to think 24 packs in parallel is a bit much. I think I would limit them to no more than 4
 

Shale MacGregor

Solar Enthusiast
So you have 384 cells? I would tend to think 24 packs in parallel is a bit much. I think I would limit them to no more than 4
Yup, 384 cells, each 16S can put out up to 110 amps, so I am trying to limit size of wire and weight of modular unit, so thinking 2P16S-12P to limit wire run thickness
 

MrM1

I'm Here, But I'm Not All There
Yup, 384 cells, each 16S can put out up to 110 amps, so I am trying to limit size of wire and weight of modular unit, so thinking 2P16S-12P to limit wire run thickness
That might work. I would have to defer to those smarter than me here.
 
Top