100 Ah 48 Volt LFP (LiFePo4) Battery Construction

Intro: 100 Ah 48 Volt LFP (LiFePo4) Battery Construction

Battery Usage.

This battery is meant to drive a 2500 watt Inverter or more producing 240 Volts AC for Homes ,Boats, Cars ,RV's etc.

Source the cells.

It has been found that the ethelene carbonate in the electrolyte/coolant of these types of LiFePo4 cells and others is causing dendrites to form . Dendrites short out and kill a cell. Now it has been discovered the manufacturers will change to alternatives and the cells previously made will suddenly get cheaper . This is happening now 9/12/19. So I suggest Lead Acid alternatives until prices stabilise again which will probably be 18 months to get all the old production sold to other mugs who do not expect this.

I bit the bullet and went with the best I could find on Ali-express even though its a punt at the best of times. The cells I settled on were 100 Ah from

Shenzhen Cozemate company Store !!!!! DONT BUY FROM THEM - THEY DO NOT HONOR THE WARRANTY . I HAD ONE CELL THAT JUST WENT TO ZERO VOLTS IN THE FIRST WEEK AND THEY DID NOT WANT TO KNOW. !!!!!

There is something wrong with these cells. I have just had a second cell die completely and will not take a charge and hold it . I have tried various methods to get the voltage back up but all end in the cell discharging to nearly zero volts. It shows a resistance measurement of 2.5 ohms with a multimeter but that might just be the residual voltage interfering with the meter.?( 9/12/19 )

I was pointed toward the fact that they are probably refurbished 120 Ah cells with new paper on the top and the serial numbers scratched off. This is indeed what these cells show and I suspected they were not new cells. (26/12/19)

The producer of the cells seems to be a large player in Jiangxi province China

Gangfengbattery.com

Cycle life:1500~6000cycles

Cost $2581 AUD for 16 delivered to my door.

Note- Ali-Express as the name implies seems strange in the way they do things. I ordered and paid for these with a credit card . Payment was accepted and off I went . A day later I received a message on Ali that my payment had been declined and I should send them details of my Drivers licence,passport,copies of the statements and a copy of the card. Now that may just be a scam operating on their site but I had no intention of sending any such data anywhere on the internet so I just ignored them. Thinking that was that I started looking around again . In the meantime I got a message from the seller giving me a tracking number and when I tried to find that number it was not there. Oh false number I thought . The seller is trying to keep the sale alive giving me that.

In fact the tracking number was correct and was with Startrack and the cells arrived by truck from Kingsgrove NSW marked PCA express. Star Track Express Pty Ltd.

I'm in Cairns in Far North Queensland and they took just on one week to get here after payment being sent from from NSW ? Packaging was robust and well done . The outside cardboard boxing was plain except for the addressing parts. The notices about Lithium and safety etc were completely isolated inside the box so no one knew what they were I suppose. They were marked "Ganfeng battery Technology Co Ltd"

There were two boxes and inside each two more ,each containing 4 cells in foam packing. Package had to be damaged 2 inches deep to even touch the cells. Weight was 19.3 kilos each box.Each cell weighs 2.13 Kilos.

A disappointment though - No Connectors and No Screws ? Screws are M6 in Stainless and 10mm should be long enough. I'll make bus bars from copper tube and use stainless plain washers. 1/2 inch soft copper tube is flattened then holes drilled appropriately.

A surprise - Terminals are Positive Black and Negative Fawn- A blob of red paint went on very quickly

Checking the voltage of each cell confirmed they were all essentially the same at 3.278 Volts each.

So what to do ? Discharge them to 2.5V and bottom balance ? Or Charge them up carefully as recommended by most manufacturers to 3.65 Volts the first time on receipt. I decided to charge them up with a BMS and an active cell balancer attached and did them in two banks of 24V nominal as that's what I'm set up for at present. I decided to just get each cell to 3.65 volts and start using them .

It should be understood that on initial charge to 3.65 Volts each cell magically lays down a protective SEI (Solid Electrolyte Interface) layer of electrolyte coated lithium ions reacting with the solvent in the electrolyte and depositing it on the carbon side of the separator.Later this protects the battery's electrolyte separator from damage from the electrons stored on that side when charged . Taking the cells back up to 3.65 Volts again afterwards is frowned upon as more plating with lithium metal occurs eventually sealing off the carbon and the charge carriers can't pass through as I understand it so far.

The 24V BMS I am using atm is 24V BMS

I will go to 48V BMS

The 24V active cell balancer is 24V ACB

I will go to 48V ACB

As I put more on this system I will update it. I am presently arranging side compression which is a must before using at large amps.

Update 10/7/20 System is performing remarkably well now and has been running the TV Computer and freezer for 10 months now flawlessly. Maximum draw to the inverter is about 15 amps and solar input maximum to the battery is also 15 amps.

STEP 1: Putting It Together and Charging

This is with 1/2 inch copper pipe connectors done and placeholder nuts and bolts . All going and charging at 8.5 Amps. Because of the weight ,(18 Kilos+) I'll keep it in two lots of 8 cells and later cable them in series for 48V .

When I compress them slightly I'll remove the connectors and file just a little sideways clearance into the connectors with a dremmel or chain saw file..

I used paddle pop sticks to separate the cells and allow air to flow for cooling.Not much with just 1 paddlepop ,about 1mm but I'll try 2 glued first on the second string. The compression sides are set up not to compress which might damage the cases but simply a thread or two tight just to restrict expansion. Its plywood and stainless rods through.

It is cheaper to use threaded stainless rod than to buy the rod and thread it yourself . Cost me $15 for the rod and $80 to get the ends threaded. I could not get them started myself as stainless is soft but hard to cut.

STEP 2: Connect to Your Inverter

I am going to use this as two strings of 8 cells giving 27 volts roughly and driving a 2500 watt inverter. Flat out it will drawing 100 amps @ 27V . Giving roughly 10 amps @ 240 V.

STEP 3:

Some shots of the insides of a cell , more to come

Comments

[Nismoboy34]

Thank you for the write up . I'm amazed that you have had two cells fail . This does make me wonder what kind of manufacturing standards the supplier has .. they should not fail so early and with two in a batch gone bad means they have developed shorts through the ion seperator .. confirm they are inscribed gfb then a serial number on them and have a qr code

[tytower]

The quality of these cells is definitely bad i
I have had a third cell fail by some sort of electrolysis action inside the cell eating through the alloy base in numerous places and electrolyte leaking out . Very nasty . Still trying to work out the cause . It does not seem to be an external contact with anything else.

Alternatively it could be the glue used to fold the heatshrink wrap over on the base.

It is on a cell positioned on the end of the series where the positive connects to the inverter . This position on all types of batteries seems to fare the worst and I think it must act as a buffer while the battery is charging . Electrons are forced out of its pos terminal during charging and in during discharge as the completion of the circuit..

[tytower]

Thank you for your comment. The cells are inscribed but I cannot read the QR code on them however they are marked GFB FFH4 D3 and a line under that has been scratched over to make it unreadable. Down the side reads AO51852402964 .

Do you know anything about the cells?
They will not take or hold a charge but show a resistivity of about 2.5 Ohms . I have pulled the first one apart and its internals can be seen at the url below but you will have to join to get access. I found no breaches visible to the naked eye on inspection of the separator so personally I think something else is going on here. The electrolyte level seemed right but the face of the copper plate seemed darkened maybe burned in a sort of camouflage pattern to look at.
I purchased a pack of 4 cells from elsewhere and replaced them . I also have 2 spares for the future. The battery is now working very well.

https://www.facebook.com/groups/1620056458221395/?ref=bookmarks

[ElectroFrank]

Your opening statement: "This battery is meant to drive a 2500 Kilowatt Inverter . . ."

That's a small power station ! I think you mean "2500 Watt" or "2.5 Kilowatt" !-)

[tytower]

OOps ,corrected thanks.

[BruceN15]

You have the battery bank wired up incorrectly. since you watch will prowse you should know his advice is wire banks in paralel FIRST than series second. you wired it backwards with series first than paralel second. it's possible that contributed to the bank going out of balance and destroying 2 cells. you need to take 2 cells and parallel them together, do that with the rest of them and THAN series those. look on will prowse's website at his wiring diagrams. that would also mess up your BMS setup becuase your bank would require 2 BMS, if it was wired the other way only 1 BMS is required, assuming I understand your setup correctly

[tytower]

Thanks for replying but you don't want to believe everything you hear until you see it for yourself . You think you are right but you are not. If you parallel the cells first and one shorts out it will take the other down with it and the single BMS will turn off. That's not the end of it . If you are not instantly on hand the two cells will continue to discharge to zero volts and that will destroy them both.

If they are in 2 strings of 8 cells with a bms on each string ,each cell is protected by its own bms and the remaining string powers your system until you get there. A BMS is $20 and pretty standard and reliable now so two are cheap . One cell is $150 so lose 2 and you are down $300.

Keep in mind too that lead acid car batteries etc have been wired this way for many, many decades . When you paralleled them it was 6S 2P all the time but you did not have the luxury of a BMS !

The failed cells I had were because they were used cells, refurbished and sold with the chinese governments sanction on ali-express. I pulled them to pieces and verified that.,
You can see my photographs in detail by joining below and going to photos.
https://www.facebook.com/groups/1620056458221395/?ref=bookmarks

[NeilRG]

This was a very informative presentation! I guess among people who know battery packs as you do compression rods are a known requirement.. I guess the reason is to prevent distortion of the inner structure of the battery?

[tytower]

Thank You .Yes . If the cells get hot through heavy discharge or charge the electrolyte can gas and swell the cells like car batteries gas off with high charge voltage rates. LFP cells though are sealed so they swell out and eventually blow the vent. The alloy holds that swollen state because it is relatively thick.

I believe once an area gasses it never recovers but I may be wrong in that if it proves to be a fact that the fluid the cells are filled with is not just a coolant but is the true electrolyte it might be possible for outside electrolyte to flow back into the damaged area.