diy solar

diy solar

Correct charging profile for LG NMC battery pack

BajaDave

New Member
Joined
May 29, 2022
Messages
19
I’m trying to determine the correct bulk/float voltages for my batteries to set in my inverter.

  • Inverter is growatt 12000T
  • Batteries are an LG NMC Lithium battery. Each battery is 6.5KW, so over the 7 batteries in the pack, 45kw total. These batteries were planned for an off-grid desalinization plant, but some were never used. They are a bit obscure, LG Model #: M48128P6BA. The battery modules built by LG contain 28 pouch style NMC cells (14 series, 2 parallel) and a module-level BMS.
  • I currently have the bulk and float voltages set to 54V, per guidance from my vendor here in Mexico. I don’t have high confidence that these numbers are precise given other issues with this guy.
I just installed Solar Assistant and finally have the data to observe my first charge cycle. Here is a timeline:
  1. Sun came up, and batteries started charging, peaking at 68.9 amps
  2. At 10:43:20 – battery charge current started dropping from 68.9 Amps, tapering down to 9 Amps at 11:09. So a 25 minute gentle curve.
  3. Charge voltage climbed from 53.2 at 10:28:00 to 53.5 at 10:42:40, and 53.6 at 10:43:20.
  4. It remained at 53.5 for the rest of the day until the sun went down (even though inverter is set to 54 float).
  5. Load was relatively constant in this period at only 200w.
I’m trying to understand which of these is true:
  1. Batteries are actually 99% full at 53.6 volts and not 54, and hence batteries are unable to accept more current. Setting should be changed to float voltage of 53, and maybe bulk of 53.5.
  2. Even though inverter is set to 54v for both float and bulk, it’s actually tailing off charging early. This seems unlikely. I thought current was controlled entirely by how much the battery could accept, subject to the max charge current and max voltage setting in the inverter.
I know given the NMC chemistry I need to be a little conservative with charge/discharge cycles. Appreciate any advice on my settings.
1653879716573.png
1653879729627.png
1653879743133.png
1653880010501.png
 
57.6 bulk charging
54,4 float
Thanks. I'm trying to understand your math. Is this correct?
LG has not published a specific charge profile for this pack, so I'm using generic NMC charge/discharge curve values.
Pack voltage (2p14s)Cell voltage
Max volts58.84.2
Min volts423
Nominal volts51.53.67
90% SOC56.424.03
10% SOC47.883.42

Using this approach I get 56.4 volts at 90% SOC. But there's the internal resistance to overcome. Is that how you came up with 57.6?

And for discharge, I'm calculating 47.9V at 10% SOC.

For float, I'm seeing the battery at rest at about 53.5v. How are you getting 54.4 for float? Wouldn't that result in overcharging?

1653926288032.png
 
Because NMC is a more volatile chemistry I only took mine to 4.05 volts per cell. That is consistent with the chart which says 4.03 volts is 90% SOC
 
Because NMC is a more volatile chemistry I only took mine to 4.05 volts per cell. That is consistent with the chart which says 4.03 volts is 90% SOC
And what are you doing for float and min voltage values?
 
And what are you doing for float and min voltage values?
I use a minimum of 3.75 volts because mine are old Nissan Leaf modules that have lost some of their capacity. I only use them for some garden equipment conversions so I do not use Float at all. On my LFP stationary system I use a float value just above resting voltage to get the last bit of solar from my panels into the pack before the sun goes down. That pack is always loaded so it never stays at a high voltage for long. but generally Float is not recommended for Lithium batteries.
 
Thanks. I'm trying to understand your math. Is this correct?
LG has not published a specific charge profile for this pack, so I'm using generic NMC charge/discharge curve values.
Pack voltage (2p14s)Cell voltage
Max volts58.84.2
Min volts423
Nominal volts51.53.67
90% SOC56.424.03
10% SOC47.883.42

Using this approach I get 56.4 volts at 90% SOC. But there's the internal resistance to overcome. Is that how you came up with 57.6?

And for discharge, I'm calculating 47.9V at 10% SOC.

For float, I'm seeing the battery at rest at about 53.5v. How are you getting 54.4 for float? Wouldn't that result in overcharging?

View attachment 96409
It won't overcharge and because lithium doesn't need floting and you need to complete your charging cycle you have to put voltage point as close as possible to the bulk charging without causing an overcharge
 
It won't overcharge and because lithium doesn't need floting and you need to complete your charging cycle you have to put voltage point as close as possible to the bulk charging without causing an overcharge
My cells complete the charge cycle anywhere I tell them to by setting the voltage point to where I want it. The Bulk cycle is a Constant Current cycle that cuts off at a voltage which is often called the Absorb voltage and is a Constant Voltage stage. Usually one setting takes care of that. Sometimes there is a rebulk setting. Do you know or can you set the voltage for when your BMS begins to balance the cells?
 
My cells complete the charge cycle anywhere I tell them to by setting the voltage point to where I want it. The Bulk cycle is a Constant Current cycle that cuts off at a voltage which is often called the Absorb voltage and is a Constant Voltage stage. Usually one setting takes care of that. Sometimes there is a rebulk setting. Do you know or can you set the voltage for when your BMS begins to balance the cells?
The bms even if it's smart one does not balance effectively you need to add an active balancer usually it start balancing when your battery is full 54,4 or when it's empty
 
The bms even if it's smart one does not balance effectively you need to add an active balancer usually it start balancing when your battery is full 54,4 or when it's empty
I agree because with a large pack the balancing current is so small that it would take a long time. My pack is 42 kWh and the OP's pack is 45 kWh. I top balance my 48 cells and then put them in a 3P16S configuration and there was still some work that the balancer had to do. I used a Heltec 2 Amp active balancer and turned off the balancing on my Orion BMS. The Orion is a robust BMS designed for EVs but EV cells are more matched than the cells I bought so the active balancer is a good call.
 
I use the 5 amp version it's more reliable
I agree because with a large pack the balancing current is so small that it would take a long time. My pack is 42 kWh and the OP's pack is 45 kWh. I top balance my 48 cells and then put them in a 3P16S configuration and there was still some work that the balancer had to do. I used a Heltec 2 Amp active balancer and turned off the balancing on my Orion BMS. The Orion is a robust BMS designed for EVs but EV cells are more matched than the cells I bought so the active balancer is a good cal
 
OK, I adjuated the settings in the growatt inverter. Bulk charge 56.4, Float at 54.4. Did this at 10:30 AM, and a few minutes later the Growatt MPP controllers shut down with the batteries around 53.8. Ian at Watts247 thought the battery BMS might’ve shut down the packs and caused the inverter to turn off solar charging. This is assuming that the max voltage is not actually 58.8 as calculated, but some lower value.

So after rebooting the inverter, I switched to utility charging with the new settings, and this time the battery voltage climbed to 54.4 at around 35 A, whereupon charging again stopped completely and current went down to around 1A. Voltage is currently being maintained at 54.2 and 1A.

Strange that both times charging stopped well short of the hypothetical 90% level of the NMC pack, or the value set in the inverter for bulk charging voltage or float voltage.

I guess I need to poke around and figure out the oddities of the Growatt charging profile, and why it might stop charging. There’s no sign of the battery voltage dropping to zero at any point today.
 
OK, I adjuated the settings in the growatt inverter. Bulk charge 56.4, Float at 54.4. Did this at 10:30 AM, and a few minutes later the Growatt MPP controllers shut down with the batteries around 53.8. Ian at Watts247 thought the battery BMS might’ve shut down the packs and caused the inverter to turn off solar charging. This is assuming that the max voltage is not actually 58.8 as calculated, but some lower value.

So after rebooting the inverter, I switched to utility charging with the new settings, and this time the battery voltage climbed to 54.4 at around 35 A, whereupon charging again stopped completely and current went down to around 1A. Voltage is currently being maintained at 54.2 and 1A.

Strange that both times charging stopped well short of the hypothetical 90% level of the NMC pack, or the value set in the inverter for bulk charging voltage or float voltage.

I guess I need to poke around and figure out the oddities of the Growatt charging profile, and why it might stop charging. There’s no sign of the battery voltage dropping to zero at any point today.
57.6 will work
 
BajaDave, not sure if your ever solved your charge voltage issue, but here some info.
I have 4 x NMC Hubble Lithium 5.5 kWh batteries, in, I believe, the same 14 cell S arrangement as yours. The instructions from the supplier is to set the charge voltage to 53.8V and float to 53.6V, and "battery empty voltage" to 44.0V. My battery BMS also enforces those limits to the inverter, exactly as shown in your graph in post 1. It seems therefore that, although NMC's voltage can vary widely from fully empty to fully full, manufacturers want to keep the cycles within the easy part of the charge/discharge curve. I believe this is (i) to extend life, and (ii) to keep the charge rate (Amps) easily controllable at/near full charge, as the faster you charge at the curve edge, the more it eats into battery life. (Current control can of course also be done by the charge mosfets of the BMS, but I reckon they want a belt and suspenders.)

That's my interpretation. So for comparison with LiFePO4 batteries:

LiFePO4 batteries have a much flatter charge/discharge curve for the "easy" part of the charge curve, but the cell voltage goes up very steeply near full, so a "natural" charge current limit is enforced, as the delta between cell voltage and charge voltage is drastically closed near full SOC. You also need to get LiFEPO4 cells to about 3.45V to even be able to balance cells, as at lower voltage, you are back in the flat curve part, so there are no significant V difference between cells, making balancing impossible. So LiFePO4 batteries can easily tolerate, and likely should get, 56V+ charge voltages. This is not the case with NMC which has a more gradual cell voltage curve all the way.
 
Back
Top