Parallel LiFePO4 12V 90Ah pack with 12V 280Ah pack, any issue?

[Bud Martin]

In the November I built LiFePO4 12V 90Ah (Liitikola) pack with 100A BMS, now I also built 12V 280Ah (LISHEN) pack with 100A BMS (same model as used in the 90Ah pack).
So does any one ever try this setup (parallel two different Ah battery packs) and see any problem in long term? My max draw at most will be about 40A, most of the time will be <20A.

 

[wmills_92105]

In the November I built LiFePO4 12V 90Ah (Liitikola) pack with 100A BMS, now I also built 12V 280Ah (LISHEN) pack with 100A BMS (same model as used in the 90Ah pack).
So does any one ever try this setup and see any problem in long term? My max draw at most will be about 40A, most of the time will be <20A.

100A in the bms gives you a lot of room to grow then.

 

[Phil Orton]

In the November I built LiFePO4 12V 90Ah (Liitikola) pack with 100A BMS, now I also built 12V 280Ah (LISHEN) pack with 100A BMS (same model as used in the 90Ah pack).
So does any one ever try this setup (parallel two different Ah battery packs) and see any problem in long term? My max draw at most will be about 40A, most of the time will be <20A.

I am considering paralleling a 160ah with a 280ah (12v each) and adding a bms to each battery also. Have you experienced any problems or know of any reason this won't work? thanks.

 

[Mike Jordan]

Connecting batts of that big of capacity difference will create issues. Typically you want batts in parallel to charge/discharge at the same rate. They cannot do that when they are not the same size. The smaller batt can be damaged as it reaches the bottom of its capacity while the bigger one has plenty to go. Or the smaller batt reaches FSOC and the bigger still needs more. Or the smaller batt may start pulling charge from the larger batt. The smaller batt BMS can shut down the whole system when it reaches its full or low state of charge, effectively giving you a smaller batt, not a bigger one. Lots of scenarios depending on BMS designs and wiring connections

I think there are ways to connect the disparate capacities with proper diodes. But you are probably better off to find separate uses for the differing sizes, and add an equal size batt to the system that needs more capacity

 

[Horsefly]

Forum guru @Steve_S has a system that has two 174Ah banks and two 280Ah banks, all wired in parallel. See here: https://diysolarforum.com/threads/my-diy-off-grid-cabin-setup-in-ontario-canada-24vdc-120vac.1484/

Like @Mike Jordan I initially thought that would cause some problems, but apparently not. Each bank has its own BMS, so if a small bank BMS shuts down, it only isolates the cells in that bank, and the others remain on-line.

My own thought is that hooking them together in parallel will obviously keep their voltages in sync, so as the smaller bank goes down the larger bank will effectively be charging the smaller bank in addition to powering the loads. My guess is that if one were to put a shunt on each bank, you'd see current flowing out of all four, then as the SoC of the smaller bank got to a certain point you would see current going into it, and more current going out of the larger bank. It's like an active balancer but on a pack basis instead of a cell basis.

 

[jpimlott]

I am running a 90 ah and 180 ah packs in parallel with their own BMS. They seam to work perfect together. Each pulls or provides the amount of power is capable of.
My large pack pulls about 2/3 of the currant and the small about 1/3.
If the packs are out of balance then the lower pulls more if charging.

 

[Mike Jordan]

Forum guru @Steve_S has a system that has two 174Ah banks and two 280Ah banks, all wired in parallel. See here: https://diysolarforum.com/threads/my-diy-off-grid-cabin-setup-in-ontario-canada-24vdc-120vac.1484/

Like @Mike Jordan I initially thought that would cause some problems, but apparently not. Each bank has its own BMS, so if a small bank BMS shuts down, it only isolates the cells in that bank, and the others remain on-line.

My own thought is that hooking them together in parallel will obviously keep their voltages in sync, so as the smaller bank goes down the larger bank will effectively be charging the smaller bank in addition to powering the loads. My guess is that if one were to put a shunt on each bank, you'd see current flowing out of all four, then as the SoC of the smaller bank got to a certain point you would see current going into it, and more current going out of the larger bank. It's like an active balancer but on a pack basis instead of a cell basis.

Thats what i was saying. The larger capacity spends time charging the smaller capacity and you don't get full use of the total capacity. And you better have some quality BMSes if they are having to fight each other with LV and HV shut down all of the time.

I have a set of Valance batts of the same capacity. They are the only factory batts that I know of, (not that others might not exist) that the BMSes communicate. I am not aware of other after market BMSes that communicate either... actually. (not that they might not exist)

I purposely set these batts out of charge balance separately from each other. Then I parallel connect them, and charged to BMS shut down, without the communication leads connected. Then I can connect the communication leads and put a considerable amount more charge back in. I don't remember the number. But if I can remember the forum where I posted that test, I'll post it here

So... maybe that is specific to the Valance BMS. But I suspect the scenario may repeat with desperate capacity systems also

But I would LOVE for someone to run a separate charge/discharge capacity test on their small/large batts. Then parallel connect, and run the charge/discharge capacity test again.

When the separate capacities are added together, is the capacity the same or different than when connected together? And... would that change over several charge/discharge cycles?

 

[Mike Jordan]

I am running a 90 ah and 180 ah packs in parallel with their own BMS. They seam to work perfect together. Each pulls or provides the amount of power is capable of.
My large pack pulls about 2/3 of the currant and the small about 1/3.
If the packs are out of balance then the lower pulls more if charging.

If you have the opportunity, I would love to see the test I described above, in my reply to Horsefly

 

[Horsefly]

If you have the opportunity, I would love to see the test I described above, in my reply to Horsefly

Me too.

Like I said, a shunt on each pack should show if there is current going from larger packs to smaller packs as they get lower. I think it must.

 

[Phil Orton]

Great input from everyone, thanks much. I would also like to see the test outlined above done by someone. If anyone does do it or hear about one being done, please post here.

 

[Steve_S]

Forum Guru ? Ohhh my, been called many things (I have an Ex so no imagination required, eh.) but Forum Guru ? That's New.

2x174 & 2x 280.
YES, the 174's cutoff for LVD before the 280's.
The 280's keep pumping out the juice even with the 174's disconnected.
Once LVD, they do not suck juice out of the 280's, only when they pulse "ON" but without real charge coming they go back to LVD.
I have a shunt on each pack, as well as a "master" shunt for whole system view. I can & do watch them.

I just completed a 3 week THRASH TEST on the packs and really beat on them with Low to High Amp changing & Discharging and with handicapping (packs off and doing only 1 pack or 2 at a time) and various combinations for stress testing to see what exactly they will do and how. I did NOT document that long, tedious and outrageously expensive set of tests... I am offgrid and have to run a Big Genset for forced hard charging and it is a PIG to be polite about it, I dropped an obscene amount of cash into it's tank + 5 oil changes (synthetic only).

BMS !
The devil is in the details... I run with Chargery BMS8T's each with a 300A DCC (Solid State Smart Relay) and that affords me the ability to closely observe & monitor the situation with the packs in the bank. The majority of BMS' when they reach LVD or HVD will cutoff as they should for the condition. They will "pulse" open or ON momentarily to test to see if there is an incoming charge for LVD situation or if there is a draw on an HVD condition and if so, goto ON state and accept the charge or do discharge. Every BMS is slightly different depending on the programming in it and the hardware it uses.

HEADS UP !
When dealing with multiple packs in a bank, things get complicated and quickly. You CANNOT read "End-Amps" or calculate such with multiple packs in place. You CANNOT have any single BMS tell an Inverter to "stop inverting" because it does not know the state of the other battery packs in the bank. The same applies to Charging as well... This also applies to the Solar Charge Controller as well...

If you are contemplating a larger bank with 4+ battery packs, then a different kind of BMS is required, one which has a centralised monitoring & control system that can be the "arbitrator" between devices and the battery assemblies. Alternately, using BMS' which can communicate / interact with a PC or Raspberry Pi which can then be programmed to read / monitor battery packs and issue instructions to Devices like an Inverter/Charger/SCC using CanBus, Modbus or other is another route.

Hope it helps. Good Luck.

 

[Phil Orton]

Forum Guru ? Ohhh my, been called many things (I have an Ex so no imagination required, eh.) but Forum Guru ? That's New.

2x174 & 2x 280.
YES, the 174's cutoff for LVD before the 280's.
The 280's keep pumping out the juice even with the 174's disconnected.
Once LVD, they do not suck juice out of the 280's, only when they pulse "ON" but without real charge coming they go back to LVD.
I have a shunt on each pack, as well as a "master" shunt for whole system view. I can & do watch them.

I just completed a 3 week THRASH TEST on the packs and really beat on them with Low to High Amp changing & Discharging and with handicapping (packs off and doing only 1 pack or 2 at a time) and various combinations for stress testing to see what exactly they will do and how. I did NOT document that long, tedious and outrageously expensive set of tests... I am offgrid and have to run a Big Genset for forced hard charging and it is a PIG to be polite about it, I dropped an obscene amount of cash into it's tank + 5 oil changes (synthetic only).

BMS !
The devil is in the details... I run with Chargery BMS8T's each with a 300A DCC (Solid State Smart Relay) and that affords me the ability to closely observe & monitor the situation with the packs in the bank. The majority of BMS' when they reach LVD or HVD will cutoff as they should for the condition. They will "pulse" open or ON momentarily to test to see if there is an incoming charge for LVD situation or if there is a draw on an HVD condition and if so, goto ON state and accept the charge or do discharge. Every BMS is slightly different depending on the programming in it and the hardware it uses.

HEADS UP !
When dealing with multiple packs in a bank, things get complicated and quickly. You CANNOT read "End-Amps" or calculate such with multiple packs in place. You CANNOT have any single BMS tell an Inverter to "stop inverting" because it does not know the state of the other battery packs in the bank. The same applies to Charging as well... This also applies to the Solar Charge Controller as well...

If you are contemplating a larger bank with 4+ battery packs, then a different kind of BMS is required, one which has a centralised monitoring & control system that can be the "arbitrator" between devices and the battery assemblies. Alternately, using BMS' which can communicate / interact with a PC or Raspberry Pi which can then be programmed to read / monitor battery packs and issue instructions to Devices like an Inverter/Charger/SCC using CanBus, Modbus or other is another route.

Hope it helps. Good Luck.

Thanks Steve, I think what you said is that based on your tests and observations that I should be ok with the 160/280 parallel but not a good idea to expand further? Appreciate you sharing your experience.

 

[jpimlott]

A screen shot from my 2 batteries with a 1000 watt load.
Very close to be evenly split.

 

[Steve_S]

The 160's will not last anywhere near as long as the 280's.
Once all batteries are between say 3.100-3.400 per cell, they will all work in tandem quite evenly.
When charging, you will see that the smaller AH packs will take less amps than the bigger battery packs. Proportionately so and that is as it should be.
On Charging... tricky but not too bad...
The 160's will reach HVD first obviously, when this happens, the balance of the charge amperage will shift to the big packs. That will fill them a bit faster at the higher side BUT one will HVD before the other (always happens) at which point, the "Last Man Standing" takes the brunt of the charge potential.

I said this many times before and will repeat it again.
If building a bank with multiple packs, one must ensure that ALL of the BMS' used are capable of being the "Last Man Standing". This means being able to take FULL 100% of the Charge Potential or the full 100% Discharge potential of your system. So if you build your system to handle 300A Max Discharge, the BMS has to handle it, if you can charge at 200A for all packs, any single pack has to be able to take it. Failure to do so and the consequences can be costly.

 

[Ampster]

. The smaller batt can be damaged as it reaches the bottom of its capacity while the bigger one has plenty to go.

The packs in parallel will equalize in voltage so one pack cannot be at lower voltage. They should remain at same percentage of capacity if discharge curves are similar. What will happen is currents will flow between the packs and for packs with different capacities of 90 and 280 Ahrs those currents could be significant depending on load. I do not begin to understand how a 40 Amp load would be shared. Others, with more experience, have already offered opinions. If it were me, I would want to monitor them closely as suggested above.

 

[Mike Jordan]

The packs in parallel will equalize in voltage so one pack cannot be at lower voltage. They should remain at same percentage of capacity if discharge curves are similar. What will happen is currents will flow between the packs and for packs with different capacities of 90 and 280 Ahrs those currents could be significant depending on load. I do not begin to understand ow a 40 Amp load would be shared. Others, with more experience, have alredy offered opinions. If it were me, I would want to monitor them closely as suggested above.

My text you quoted was capacity.... not voltage

 

[Mike Jordan]

A screen shot from my 2 batteries with a 1000 watt load.
Very close to be evenly split.

What BMS are you running?

 

[Bud Martin]

What I see so far is that the 90Ah pack current discharge and charging current is about 1/3 of taltal current, the 280Ah pack current discharge and charging current is about 2/3 of the total current.
I do not have the time to do total capacity test, I will just keep an eye on them at this point.

 

[Ampster]

My text you quoted was capacity.... not voltage

I understand that. Because physics says that the voltage will always be equal and because as others have said the current pulled will be proportional to their capacity I was assuming that the SOC or percentage of capacity will be equal. I think physics prevents one battery from being at 2.5 volts per cell while the other one in parallel would be 3.0 volts. I assumed that is what you meant by bottom of capacity? I can see a scenario under heavy load in which the smaller capacity pack would sag but that would create an eddy current. If you meant something different or if you have an different interpretation of physics I am open to a discussion that might enlighten me. .

 

[jpimlott]

The BMS's are 8S 100 Amp from Overkill solar. He is great out of Florida.
Great warranty great service. He will customize the ends the way you want it.
I got 2 8 AWG cable on both inputs and output that were 2 feet. It save another
connection and hassle.