Lithium Cell Balancing at Nowty Towers - Resurrections

[nowty]

(nowtyproject for searching)
Having had my original thread shredded at St Elsewhere, and my re-loaded version shredded again at the other intermediate place,


I have decided to re-document my project a third time which may be of help to others. I will add bits to it over a period of time.

To be honest I think most folk are now going down the modular pylontech battery or other commercial battery as the cost of building your own battery bank is not much cheaper anymore but it may be of use to some folk who like tinkering with stuff.

Nowty Towers Batteries
The bulk of my lithium battery bank comprises of re-used Growatt batteries originally used in the Growatt SP2000 system. I bought up generally faulty ones a few years back around 2017. It always turned out the cells were fine, but the control boards were faulty and I managed to buy at an average price of a little over £100 per kWh so they were as cheap as lead acid at that time.

The original Growatt battery unit comprised of 2 x 24v quite handy module racks which were easy to pull out.
The gross capacity of each unit was about 4.8kWh with each 24v module rack being about 2.4 kWh.



Using a bench top power supply and a 1 ohm power resistor, I charged or discharged the module cell packs to manually balance them all up.




I first used them in a SMA Sunny Island 24v system and set them up as lead acid batteries. I set all charging voltage's, bulk, absorption and float to the same voltage which matched a SOC of 80%. Also set the charge temperature compensation to 0mv per degree, i.e. disabled that function.

I parallel’d up all the cell packs with fused croc clip links. Even without any active balancing I found that operating them conservatively between 20% and 80% SOC the cells remained within around 100mv of each other over a period of 6 months or so.

[nowty]

Each 24v module rack is made up of 7 cell packs in series.

Each module cell pack is comprised of 18 x Boston Swing Cells in parallel, each one being a 3.65v 5300 mAh battery.
https://liionbms.com/pdf/bostonpower/swing5300.pdf


A test review of the cells was published here in 2018.
https://secondlifestorage.com/index.php ... -test.5188


And a tear down (literally) of one of the cells.
http://www.yunergybattery.com/newsview. ... pjQLu7MKUk

[Joeboy]

Nice one Nowty, will take me some time to digest it all and understand about half of it.

Does this mean there's more to it than spurring off 15 of 3.2V from the Pylontechs?

I see the 15x3.2v in series gets me to 48v and with capacity of the modules that will be around 100Ah storage at 48v. I am not understanding what else is needed to ensure the batteries have a long life. I can set the max discharge rate to 80% via the goodwe software and both the Pylonechs themselves and the Victron are upper voltage limited. All operating with the parameters of these batteries?

Help me out here, why cant i just loop those 15 in in series, hook it up and let it run? (Ever hopeful).

[Tinbum]

The Growatt cells have a different voltage to most standard Li.

[nowty]

Joeboy, your pylontech are LIFEPO4 or LFP for short and so are the ebay ones your looking at so you need 15 of them to match. The quoted voltage of 3.2v is the nominal voltage of them but you typically charge them up to 3.5v which gives 54v for 15 cells. Remember the Pylontech manual says charge up to 54v. They will be damaged if you charge higher than 3.7v per cell.

As Tinbum says, my growatt batteries (I think they are NMC) have a higher nominal voltage of 3.65v and can be charged up to about 4.2v although I keep them at 3.85v with 14 cells to keep them and the pylotechs (15 cell) happy together.

I would not have started out this way but I did have two systems, one OnGrid and one OffGrid but I have now a single On/Offgrid system with a larger inverter so I amalgamated the two battery banks. It should become evident as I add more info to the thread and I did not want to add in the pylontechs to the thread at this stage.

[Joeboy]

Joeboy, your pylontech are LIFEPO4 or LFP for short and so are the ebay ones your looking at so you need 15 of them to match. The quoted voltage of 3.2v is the nominal voltage of them but you typically charge them up to 3.5v which gives 54v for 15 cells. Remember the Pylontech manual says charge up to 54v. They will be damaged if you charge higher than 3.7v per cell.

As Tinbum says, my growatt batteries (I think they are NMC) have a higher nominal voltage of 3.65v and can be charged up to about 4.2v although I keep them at 3.85v with 14 cells to keep them and the pylotechs (15 cell) happy together.

I would not have started out this way but I did have two systems, one OnGrid and one OffGrid but I have now a single On/Offgrid system with a larger inverter so I amalgamated the two battery banks. It should become evident as I add more info to the thread and I did not want to add in the pylontechs to the thread at this stage.

Thanks Nowty, I am set at 53.2v. It continues to blow my mind that such small shifts in voltage can carry represent so many Ah of power. I am enjoying the thread.

[nowty]

Thanks Nowty, I am set at 53.2v. It continues to blow my mind that such small shifts in voltage can carry represent so many Ah of power. I am enjoying the thread.

To answer your actual earlier question, yes you can hook up 15 of them ebay ones but you need to add a string fuse and cell balancers. Something this thread is about but have not quite got to the detail yet.

Li-On batteries have very flat voltage charge and discharge curves and LFP cells have incredibly flat curves so your right that very small changes in voltage can represent large amounts of power.

[Joeboy]

Thanks Nowty, I am set at 53.2v. It continues to blow my mind that such small shifts in voltage can carry represent so many Ah of power. I am enjoying the thread.

To answer your actual earlier question, yes you can hook up 15 of them ebay ones but you need to add a string fuse and cell balancers. Something this thread is about but have not quite got to the detail yet.

Li-On batteries have very flat voltage charge and discharge curves and LFP cells have incredibly flat curves so your right that very small changes in voltage can represent large amounts of power.

Brilliant , I will go into full on standby. Its all still floating in my mind so not real yet.

[nowty]

I then upgraded to a larger SMA Sunny Island 48v system.



So each original pair of 24v modules were re-connected in series with a heavy duty 10m㎡ 40A fused link, so that's my string fuses.



To make it easier to connect the parallel cables I made up some bus bar extenders out of copper strip.



I ended up with 10 x 4.8kWh batteries and some time later I split them up into two battery banks of 5 x 4.8kWh, each with DC isolators. I also parallel'd up every cell in each battery bank like I had done before.




Each battery bank is fused back to a common heavy duty distribution point, before going back to the Sunny Island battery inverter. The main battery lead back to the Sunny Island is 70m㎡, the intermediate leads to each battery bank is 50m㎡ and the each battery string parallel leads are 25m㎡.

[Stinsy]

.
(I find the squared character as in “4mm²” looks better that a superscript 2 as in “4mm²”)