diy solar

diy solar

LG Chem 2KWH 48Volt Power Wall for off grid house

paulmurdoch

New Member
Joined
Dec 17, 2020
Messages
36
Hello,
Ive posted before explaining how I want to make a power wall using x28 LG chem 48V in x4 14S parallel configuration (56KW) for my offgrid house, each of the 4 will have a Daly 14S BMS. These are the same LG chem batteries, which battery hook up are selling once again. My new question is my offgrid house is a long ways a way, and the current lead acid batteries are dead (This is a new house I just purchased)

My plan is travel to the house, then during a 4 day window, configure the lithium batteries and use the existing inverter and charge controller from the lead acid batteries system. The charge controllers x2 are the Xantrex W60 MPPT which seem to have a robust configuration system, where equalization can be disabled, Bulk & Absorption V can be set and float can be disabled. Is it practical to use this charge controller for these type of Lithium Ion batteries? If so what would be the recommended settings?

My next question is a tad more difficult. I have no idea of the voltage of these batteries and I need to safely balance them without any power being available. My plan is to open each one up and remove the internal BMU, parallel the x2 7 cells to make a 2S from each LG battery. I believe putting them in parallel at this point, will be ok as the BMU is still balancing these cells. My problem is I want to then series 7 of these units to form the 14S, and install a 200amp daly BMS. My problem is how can I balance these large batteries before placing them in series without power to charge? I thought of the obvious, but I'm unsure about linking these in parallel to self balance. I also thought about using the charge controller to charge each cell independently. This might be the best way to go, but I'm not sure if I have enough time.

I would appreciate any advice.

Thanks!
 
Linking any chemistry in parallel and just letting it sit does very little to actually transfer charge. Even after a week or so, they'll still be 5-10% off, which is kinda a mile in this context.

IMHO, abandon current plan. Construct and configure/balance the battery while you're not under such a time crunch while you have power available.

Had I taken my own advice, I would have been done with my battery a year ago.
 
Thanks for the quick reply!
I have to try to get power to the house, so abandoning my plans is not really an option. I'm thinking now of attaching each 48V LG chem unconfigured battery to the solar charge controller and charging each battery in turn using the internal BMU to balance the batteries, while monitoring closely. Then when I have 7 fully charged batteries, configuring them as a 14S, attach the BMS, power the house and charge all other batteries individually.

Do you think using the charge controller above could accomplish this? If so, what type of settings would be recommended for Bulk, Absorption and float?
 
Confirm:


Clearly states you need your own BMS. I wouldn't attempt charging without a BMS installed. When the user is acting as BMS, that's when stuff goes sideways.

I recommend you build and configure your battery before arrival, or at least do it to the point that it's more a matter of simple assembly and BMS leads. While paralleling may only get you to 5-10% variation in SoC, that 5-10% may be better than nothing, and they should stabilize pretty quickly once current is applied.

I'm having trouble visualizing how you intend to configure the final pack given the constraints, but there's going to be inherent imbalance by using 28 different packs. For that reason, I would attempt to build each cell group by interleaving cells from all packs.

Assuming 392 (28*14) cells. Each parallel cell group would be constructed from 1 cell from each pack, but since you want to have 4 separate 7P14S packs, that complicates things. If any given battery is constructed by interleaving all cells from 7 packs, at least each component in the series string will be the same capacity.

battery 1 is constructed by interleaving all cells from packs 1-7.
battery 2 is constructed by interleaving all cells from packs 8-14.
battery 3 is constructed by interleaving all cells from packs 15-21.
battery 4 is constructed by interleaving all cells from packs 22-28.

While each of the 4 batteries might have difference capacities, at least each cell group within each battery will have the same capacity.

Am I making sense?
 
Yes thank you! The plan is to take each 48V stock battery containing x2 24V 7S2P wired is series, reconfigure each stock battery to a x2 3.4V 7S 2p wired in parallel. Then series x7 batteries with a 200AMP BMS to give a 48V 14S (14KW) battery. Then parallel the 4 batteries to give a 56KW battery.
Below is a schematic.
1615400679766.png
 
Q-Dog Thanks! it looks as though this is the safest way to go! If I do manage to get a generator, can any one recommend a good charger unit to charge up these bad boys?
 
Q-Dog Thanks! it looks as though this is the safest way to go! If I do manage to get a generator, can any one recommend a good charger unit to charge up these bad boys?

I think I had gotten there in my head, but needed it sketched out. Definitely interleave them as I described.

Hopefully your inverter includes a charger function. One would hope most off-grid installations do. Do you know what inverter you have?
 
The system has two Xantrex charge controllers (Xantrex W60) and the inverter is a Xantrex XW 6048 series hybrid inverter/charger
 
You should be able to hook a generator up to the AC input of the XW 6048, and it will charge the batteries separately or in conjunction with the solar charge controllers. Consult the manual for the generator requirements (120VAC or 120/240VAC split phase input, etc.).
 
Thank you for your help!
Do you think the schematic is the correct way? Inside each stock battery I will break the series connection from the x2 24V 7S2P and parallel x2 7 cells to form x2 3.4V 7S 2p. Then series them to give me a 6.8V 14S 2p for each stock battery.
Then series x7 of these to make a 48V 14S (14KW) battery with a 200amp BMS. Finally parallel all 4 batteries to give the 48V 56KW battery?
 
Last edited:
Let me take what I've read and feed it back to you to see if I have it.

Series means voltages add, chaining (+) to (-). Parallel means + to + and - to - leaving voltage constant but increasing capacity.

Stock: Inside each box is 14S "cells" ("cells" are actually a module of 2 cells in parallel), two stacks of 7 "cells".

Modified: Inside each box will be two stacks of 7 "cells". Each stack will be 7P "cells," and the two stacks will be series connected making a 7P2S pack, 7.4V nominal.

7 modified boxes will be put in series to make a 7P14S 48V battery with an installed BMS.

You will have 4X of the described batteries all in parallel.

I believe that is the best way. I also believe you should "interleave" the "cells" as mentioned in my prior post.

Take stock packs 1-7, break them down.

Each modified pack will have two stacks of 7 with a single "cell" from each of the 7 packs.

While you can't be certain that your 4X finished 7P14S batteries have the same capacity, you will have a high degree of confidence that EACH 7P14S battery's "cell" groups are nearly identical.
 
Sorry it's rather wordy, you have it correct except for a couple of points below. Please see my comments below in red text!
Let me take what I've read and feed it back to you to see if I have it.

Series means voltages add, chaining (+) to (-). Parallel means + to + and - to - leaving voltage constant but increasing capacity. yes

Stock: Inside each box is 14S "cells" correct ("cells" are actually a module of 2 cells in series parallel), two stacks of 7 "cells".in series then series the two stacks in series=48V

Modified: Inside each box will be two stacks of 7 "cells". Each stack will be 7P "cells," in parallel and the two stacks will be series connected making a 7P2S pack, 7.4V nominal. correct

7 modified boxes will be put in series to make a 7P14S 48V battery with an installed BMS. Correct

You will have 4X of the described batteries all in parallel. correct

I believe that is the best way. I also believe you should "interleave" the "cells" as mentioned in my prior post. Please explain! I think the modified above is the same?

Take stock packs 1-7, break them down.

Each modified pack will have two stacks of 7 with a single "cell" from each of the 7 packs.

While you can't be certain that your 4X finished 7P14S batteries have the same capacity, you will have a high degree of confidence that EACH 7P14S battery's "cell" groups are nearly identical.
 
Stock: Inside each box is 14S "cells" correct ("cells" are actually a module of 2 cells in series parallel), two stacks of 7 "cells".in series then series the two stacks in series=48V

I believe you are incorrect. I believe they are in parallel.

Based on what David Poz says in his video. He measured approximately 3.8V between the terminals of each stacked unit indicating all cells within the stackable unit are in parallel.

View attachment 40405

Stock: Each stackable unit is actually FOUR cells in PARALLEL. The 7 stacked items are in series.


Henceforth, let's refer to EACH stackable unit as a cell since it has the voltage of 1 cell.

So, each stock box has 14 cells in it, two stacks of 7 with all of them in series.

Interleaving... You're going to mix cells from all packs into each stack as follows:

Disassemble stock packs 1 though 7. Stack all 14 modules from each pack and keep them separated - 7 stacks of 14 cells.

Build modified pack 1:
Stack 1 - assemble with 1 cell each from the 7 packs in parallel.
Stack 2 - assemble with 1 cell each from the 7 packs in parallel.
Connect stack 1 with stack 2 in series.

Repeat for modified packs 2-7.

When complete each 7 parallel cell stack should have identical capacities even if the 7 stock packs had different capacities.
 
Hello again!
I double checked, and the stock batteries are configured in series.
Stock battery =48V
Pack one x7 cells in series 3.7V nominal 3.7*7=25.9V
Pack two x7 cells in series 3.7V nominal 3,7*7=25.9V
Pack one and two in series =51.8V for each stock battery.

My question is, if I have different voltages for each stock battery, and I interleave all cells in parallel, as you said this will give me the same capacity and the same voltage for each battery (I really like this) but the individual cells will have different voltages before being placed in parallel.
Ive attached a spreadsheet, where I made two linked tables, so I can feed each cell voltage in the left table and it will populate the right table, and tell me if I'm above or below a 10% discrepancy of the newly made battery mean.
My question is, do you think maintaining less than a 10% discrepancy from the mean voltage of each new battery is acceptable? Id like to charge all the cells independently, but being time constrained this wont be an option.
Thank you again for your time!
 

Attachments

  • LG Chem inerleaving.zip
    13.9 KB · Views: 11
Don't over-think it.

Re-sort by cell voltages and build from "like" voltages:


Cell V1
Bat 1
3.25​
Bat 2
3.43​
Bat 10
3.47​
Bat 7
3.55​
Bat 13
3.56​
Bat 22
3.57​
Bat 21
3.6​
0.35
Bat 12
3.61​
Bat 27
3.61​
Bat 3
3.65​
Bat 11
3.66​
Bat 6
3.67​
Bat 23
3.67​
Bat 8
3.69​
0.08
Bat 20
3.69​
Bat 4
3.77​
Bat 14
3.78​
Bat 28
3.83​
Bat 25
3.84​
Bat 18
3.85​
Bat 19
3.86​
0.17
Bat 5
3.89​
Bat 26
3.91​
Bat 17
3.95​
Bat 16
3.98​
Bat 15
3.99​
Bat 9
4.12​
Bat 24
4.15​
0.26

That minimizes deviations within each cell group, and as mentioned, after a week or so, they will be within 5-10% of each other in true state of charge - possibly even less of a difference as this chemistry has a very high SoC to voltage correlation. I haven't personally tested this, but have on other chemistries. Once you start charging, they will equalize very quickly as charge is absorbed much quicker by any cells at lower SoC.

It's extremely encouraging that each battery has identical cell voltages to X.XX accuracy.

I would recommend you charge each 48V battery individually to 56.7V. If all voltages are below that, you can select a lower voltage provided it's higher than the highest. Once all packs are within 0.2V of the same voltage, connect them all in parallel. All charging should be done with the BMS installed and protecting the cells.
 
Sorry, I should have explained! The spreadsheet is only an example, I have no idea of the voltages until I travel. Although, the last time I was there I did manage to check one stock battery and the 14 cells within the battery were within +/-0.2V. They have an installed BMU so it seems to balance quite well.

I will do as you suggest, please tell me if I'm on the right track!

Split the 28 stock batteries into two lots, 14 stock batteries. Open 14 of the stock batteries and interleave the individual (3.6V) cells, to form 14 modified batteries, matching the voltage of each cell as close as I can. At this point, In each modified battery there will be a total of x2 packs of 7 cells with slightly different voltages per cell.
Parallel 7 cells in each pack, this will give me two packs of 7cells in parallel in each modified battery, and each pack voltage should be the same ~3.7V. Series the two packs = ~7.4V for each modified battery.

Do the same for the other 13 stock batteries. I now have 14 modified batteries each with a voltage of ~7.4V. Split the modified batteries in to two groups of 7, Install BMS wires in each group of 7 modified batteries (total 14S) check voltage of each modified battery, (which should be the same) and series the 7 modified batteries to give me x1 14S battery. (7.4Vx7=51.8V) Connect BMS.

Charge the modified battery to 56.7V. Repeat the process with the other 3, once all 4 batteries are at 56.7V parallel and all should be good...

Does this sound accurate to your thinking?
 
±0.2V within a pack is really bad. Staggeringly so. If you mean ±0.02V, that's much better. If it's really ±0.2V, you might want to re-think this whole thing.

Okay. I see what you did. You'll build 2X batteries at a time using 14 packs and constructing each modified pack with 14 cells instead of the 7 I recommended. That will make those 2X 48V batteries the same capacity - good idea.

I would definitely gather all data first (all 392 cells), sort and select the packs according to ascending cell voltages. Look for outliers. Formulate a backup plan if something looks ugly. If you find one or more dead modules, you can always only build 2 or 3 batteries and pursue resolution with batteryhookup.

Don't make too many compromises. Outliers are where the shit will hit the fan 99% of the time, and they should be avoided. If a pack has all cells within 0.02V, but one is 0.1V out, consider not using that pack.
 
Roger that! Thanks for the help, I'm going to start with 2 packs and finish the the other 2 at a later date. 28KW is sufficient, until I get the hot tub....
 
Back
Top