question

heggink avatar image
heggink asked

Multiplus II Lifepo3 settings for REPT 280AH

I really need some help with the correct Inverter and ESS Settings for my new config.

I have the following:

  • 3x MP-II 5000 in a 3 phase config
  • 3 strings with 16 REPT 280AH LIFEPO3 each with a JK BMS, total 840AH. REPT specs seem extremely similar to CATL and EVE.
  • max discharge load 70 amps per inverter

The system has just been brought up since 24 hours. Initially, doing some testing with lower loads were all fine. As soon as I increase discharge load then then the voltage of the string drops quite a bit after the battery gets below 50% SOC causing battery low warnings. Voltage ultimately ends up being 45V where the BMSs (not connected to the ESS) and the smart shunt still think the SOC is 42%. So that is one challenge I have. I made some changes to the settings but I don't think I want to go too low (e.g. 42V cut off) under load. If anyone has any hints as to how to arrive at the best settings and why there seems to be a discrepancy between the SOC and voltage then please shout.

Secondly, I just tried the charge cycle which seems to work similarly: the moment the load increases, voltage goes up quite fast causing the MP-IIs to reduce power quite quickly. From 45V to 52V took 30 minutes where the SOC went from 42 to just over 47. Any advice on charge settings here would also be appreciated.

In summary, the batteries seem to vary quite substantially in voltage under high loads causing the system to struggle with load/unload parameters.

Thoughts? Many thanks in advance.

ESS
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

2 Answers
seb71 avatar image
seb71 answered ·

LiFePO4 (not LiFePO3).


Did you fully charge all cells individually before making the battery strings?

What is your Absorption voltage?


How are the strings put in parallel? Using busbars, where each string has its own wires?

What is the cross section (gauge) of the battery wires and inverter wires? How long are the battery wires and inverter wires?


You should stop the discharge at 48V. Going lower than 3V cell voltage you risk damaging the cells, especially if the cells are not well matched or equally charged.


210A draw from a 840Ah LiFePO4 battery bank should not cause such a voltage drop.

Maybe test each battery string individually with loads up to 100A (on DC side) to see if all behave the same.

11 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

heggink avatar image heggink commented ·
I did not fully charge so in the process of doing that now. I did check all the voltages of the batteries when I got them (all identical).


Changed the absorption to a better level indeed.

Strings are connected with busbars, each string indeed having its own connection using 35mm2 cable to support the 70 amps. Battery and inverter wires are between 40 and 70cm.

I made sure that I can switch off individual batteries so will follow your advice to test the strings individually if the problem persists. Many thanks for your advice.

0 Likes 0 ·
seb71 avatar image seb71 heggink commented ·

I did check all the voltages of the batteries when I got them (all identical).

And what was that voltage?



Changed the absorption to a better level indeed.

What was the Absorption voltage and what you changed it to?

Values are important to know.

1 Like 1 ·
heggink avatar image heggink seb71 commented ·
If I recall correctly, all 48 were exactly 3.26V.

The Absorption voltage was set at 52V so that explains why it cut out so early. It's now been charging for a couple of hours at a reduced wattage (just 30 amps per string) and has been crawling up to 53.8V. I set it to 56.5V so will see how that goes. I would expect the BMSs to provide the minimum protection (over/under voltage) but was somewhat surprised by your 3V since that seems rather high.

0 Likes 0 ·
seb71 avatar image seb71 heggink commented ·

Absorption at 52V is way too low for 16 cells. That's the main cause of your issue.


As a precaution you could first set Absorption to 55.2V and check if the cells are still close in voltage when you reach that battery voltage.

After a few charge/partial discharge cycles ( For now do not discharge under say 50.4V), if all OK (all cells at the same level), you can increase Absorption voltage to 56V.

Repeat.

If all OK, try 56.8V.

You should stop here.

Anyway, the absolute maximum, for perfect cells would be 57.6V (corresponding to 3.6V cell voltage).


-------

I would expect the BMSs to provide the minimum protection (over/under voltage) but was somewhat surprised by your 3V since that seems rather high.

Take a look at LiFePO4 charge and discharge curves (cell voltage over time during charging or during discharging).

They have a slight slope most of the curve, except the ends, where the voltage changes very quickly.


You don't want to discharge to the point where the BMS disconnects the battery or the inverter shuts down (assuming the cut-off voltages are set correctly).

Technically you could discharge down to 2.5V (LiFePO4 cell voltage), but there is not much energy left between 3.0V and 2.5V and more importantly, from about 3.0V downward the voltage drops very quickly as I said, which is dangerous because you could damage the cells.

When using a BMS the cells are "top-balanced" (equal voltage at the top).

But invariably there will be manufacturing differences between cells and if you would discharge the battery that low, some worse cells will get there sooner.

The BMS should protect the cells in that case (by stopping the discharge), but it is much better to not get to the point when the BMS has to stop the discharge because a cell reached the minimum voltage set in the BMS.


-------

Float voltage - set it to 53.6V.

0 Likes 0 ·
derrick thomas avatar image derrick thomas heggink commented ·
Voltage is a horrible indicator of SOC for LifePo4 batteries. Whenever building a diy battery it is imperative that the batteries be balanced before placing them in series. You can have cells with a fairly large difference in SOC and have the same resting voltage. The last bank I built was with balanced grade a eve cells which were a matched set and they still required some balancing before building the pack. If you don't have a power supply to do a balance with before assembly, you can sometimes get by with using an active balancer but it might take a while. Until you are certain that the banks are properly balanced you are waisting your time trying to find solutions to your problem.
1 Like 1 ·
heggink avatar image heggink derrick thomas commented ·
That's really good to know (as a noob). So if I stick to acceptable parameters but limit the power draw from/to inverter/batteries then they should balance out after a while, correct? Effectively, what you say is that the BMS initial readings will be misleading until all batteries are properly balanced, correct? Any thoughts on "take a while"? I suppose I can monitor the BMS cell differences.
0 Likes 0 ·
derrick thomas avatar image derrick thomas heggink commented ·
How long it will take to balance depends on how far out of balance they are and how much balancing the BMS can provide. Just continue charging at a low rate so as not to push the highest cell to the cutoff voltage while still providing enough current to keep the cells in the balance voltage range. I believe you can set the balance parameters in the JK. Anything below about 3.45 really doesn't do much. The higher the voltage gets toward the "full" voltage the greater the imbalance will be.
1 Like 1 ·
heggink avatar image heggink derrick thomas commented ·
That was my thinking indeed. I already configured everything (via node red) such that it slows down charging quite a bit closer to full charge so that should help as well. Many thanks. Much appreciated.
0 Likes 0 ·
heggink avatar image heggink derrick thomas commented ·
At the moment, the biggest imbalance across the 3 strings seems to be 0.02V between highest and lowest cell. Not too bad it seems. Cells are at 3.35V ATM.
0 Likes 0 ·
derrick thomas avatar image derrick thomas heggink commented ·
You're not going to see much cell imbalance at that cell voltage. Higher deviation will start to show somewhere around 3.45 as the cells start to come out of the flat part of the curve.
0 Likes 0 ·
heggink avatar image heggink derrick thomas commented ·
Thanks again for all the input. Having done a couple of cycles, everything has balanbced out really nicely and your input has been excellent to take a breather and see everything stabelise. Never realised how these cells behave in real life but it's all clear now.

Thanks again!!

0 Likes 0 ·
heggink avatar image
heggink answered ·

Answering the second question myself, I adapted the float and absoption settings on the charge tab which made it a lot better.

2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.