New battery bank cycling

[shadowsteve]

I searched but didn't find what I was looking for but maybe I'm not searching using the right words.

I have two newly built 4s packs using 280Ah cells. They each have a JBD 150A BMS on them.

I top balanced all the cells to 3.65 until the current draw was negligible.

I have one of the packs hooked up and charging via solar & discharging via 3000w inverter with the RV fireplace as a load. The BMS reports cell voltages as 3.18x and pack @ 12.7v. The pack has not been cycled down yet and this is the first draw on it beyond small tests. I can also use the onboard PD4590 to charge the pack once it's discharged.

How far down should I draw the pack? To the 10v BMS cutoff? Should it then be charged back fully to the 3.65v cell cutoff?

There seems to be a few threads with BMS settings but none are exactly the same. Is there a sticky for the settings that I missed?

I found in my initial small tests that the cells would run up and trigger a charging shutoff with 0.1xxV cell differences. Loading the pack would bring them back in line to 0.005V area.

TIA

 

[OffGridInTheCity]

I'm 18650 so I have a wider voltage range to work with so it's easy to just turn my load on/off by voltage to achieve the DOD I want - the BMS is completely separate from this.

However, I have read a bit and can repeat some basic understandings for you and will follow this thread with you.
1) A BMS is more properly a max/min fail-safe, and is not ment to control battery load on/off - but rather to protect the battery in extremes (100% range). The actual load on/off is better handled by a non-BMS method and used to achieve you're personal goals such as 80% DOD. This is not required, just a general principle.
2) LifePo4 is such a flat voltage curve that voltage may not be the 'best way' to control DOD, but rather measuring coulombs in/out - e.g. a shunt and some load control based on this. Again, this is a principle not a requirement.

 

[shadowsteve]

I'm 18650 so I have a wider voltage range to work with so it's easy to just turn my load on/off by voltage to achieve the DOD I want - the BMS is completely separate from this.

However, I have read a bit and can repeat some basic understandings for you and will follow this thread with you.
1) A BMS is more properly a max/min fail-safe, and is not ment to control battery load on/off - but rather to protect the battery in extremes (100% range). The actual load on/off is better handled by a non-BMS method and used to achieve you're personal goals such as 80% DOD. This is not required, just a general principle.
2) LifePo4 is such a flat voltage curve that voltage may not be the 'best way' to control DOD, but rather measuring coulombs in/out - e.g. a shunt and some load control based on this. Again, this is a principle not a requirement.

I have a shunt and it's setup to monitor once the system is done (same setup as from AGM batteries I replaced).

My question is in reference to the initial few cycles of the pack from full charge down to discharged. I'm figuring that the BMS will do the cutoff to establish a full cycle range. This is not for normal usage of the batteries but rather just to establish a few full charge/discharge cycles.

 

[RCinFLA]

It is not uncommon for cell balance to diverge when run at high discharge current or over their total capacity range.

Cell mismatch will cause this. Most folks think same AH capacity and same cell impedance means matched cells. Matching also requires same terminal voltage drop difference for a given discharge current. The greater the discharge (or charge) current the greater the difference in cells terminal voltage drop under load, causing balance to diverge faster. At higher discharge current, cell internal heating differences, just due to their physical position in pack effecting how fast they can dissipate heat, will effect their terminal voltage drop under load.

The amount of terminal voltage drop from cell rested open circuit voltage is called overpotential. It drops terminal voltage under load and rises terminal voltage under charge current. Cells with same AH capacity and same AC impedance can still have some difference in their overpotential values for a given current. This means they have a little more or less cell loss which causes imbalance to accumulate.

Overpotential has a time delay in response to a cell current change. It has an exponential decay to equilibrium level. For an LFP cell this decay takes about a minute to get about 90% of the the way through the exponential decay cycle to the new steady state equilibrium cell terminal voltage.

BMS settings are best selected based on how you typically plan to use the batteries. Peak current level and range of state of charge used are two of the most impact items effecting how you set up BMS. For example, if you typically run higher peak currents you might want low voltage cutoff setting to be a bit lower than if you typically run lower currents. You should also set up inverter so it shuts down for low battery before BMS does. Battery cabling voltage drop has a big impact on this. You should strive to never have a BMS shut down. Greater charge rates means you have to keep cells in tighter top end balance to prevent overvoltage shutdowns of BMS. If your BMS does not balance until a cell reaches 3.4v and you only charge to a lower battery pack voltage you may not get enough balancing time.

 

[shadowsteve]

@RCinFLA I set the cell cutoff to 2.75v and ran the pack down until the BMS disconnected the load. The cells jumped back to 2.9xxV right away and I'm now charging under solar at up to 20A (I'm in Ontario and panels aren't getting best sun). The diff is currently 0.005V and when the load first disconnected they were @ 0.066V diff. I'll let it charge now but it'll likely take a while given the sun situation

 

[RCinFLA]

When you get to low state of charge the overpotential is much greater. From 2.75v under load it will jump up when load is removed. This is the overpotential recovery.

Same applies near full charge, although there is another factor of surface charge buildup voltage when cells are allowed to fully charged taper off to low current. A one ohm power resistor applied across cell for about 30-60 seconds should burn off most of surface charge voltage.

When you make cell open circuit voltage measurements let them sit for one to three minutes after removing load or charge. At low state of charge it can take longer to reach final steady state open circuit voltage. At high state of charge you have to be careful of surface charge capacitance which can take several days to bleed off on its own. Once any surface charge is blead off, a fully charged LFP will have open circuit voltage of about 3.45v.

The amount of overpotential voltage relates to availability of ions within cell to support cell current demand. At low state of charge, free ions are scarce to support discharge current so greater overpotential voltage drop is required to push ions along to support discharge current demand. At high state of charge, free ions to support charging current are scarce so greater overpotential terminal voltage rise is required to push ions along to support charge current demand.