Mixing Lithium and Lead

[nowty]

Last year I bought several cheap second hand Growatts with 5 kWh batteries off fleebay, one of which I had been keeping as a spare. But my experience is once you buy a spare item, it never gets used so I started thinking of uses for it. The battery cannot be used without the Growatt controller as its operation is locked down. I was a bit bored the other day so decided to take the back off the battery and have a closer look to see if it could be easily unlocked.



It contains 14 battery modules in series and a proprietary protection circuit board. Each battery module contains 18 of 5,300 mAh cells in parallel so a total of 252 cells. Each cell is apparently two 18650’s in parallel in a single package.

http://liionbms.com/pdf/bostonpower/swing5300.pdf

First thing was to rip out the protection board and then the two battery racks conveniently come out to give 24v each. Then I had a crazy idea, my Sunny Island runs at 24v using forklift cells. Would it be possible to add them in parallel to increase my OffGrid capacity ?

It sounded like an absurd idea, mix lithium and lead ?, but the more I thought about it, the more it started to be a possibility. Charge and discharge graphs are different, but my operating range of lead is typically between 100% and 75% SOC. Voltage wise, this matches favourably with lithium between 100% and 40% SOC. Although I would only be adding circa 20% more Ah capacity, the actual usable capacity might increase by 50% or more. At high SOC, the lithium ones would provide most of the power so may extend the life of my forklift cells which are now 6 years old.

I took out the two lithium racks and connected them in parallel with my forklift cells when the voltages were roughly equal to prevent any initial current surges.





Then I reduced the charging voltage of my Sunny Island from the current 30.5v to 29.4v and set the temp compensation to zero. 29.4v gives the max 4.2v charging voltage per lithium module and is still good enough for the forklift cells.

Next day was full sun , so I measured the voltage of each lithium module regularly. Voltage increased more slowly than just with lead alone and the lithium was taking 75% of the current. When the voltage rose to around 28v the current was shared equally, then over 29v the lead was taking most of the current. The voltage of each lithium module was identical except for one with was 0.1v more, so cells looks nicely balanced so far. When the charging day was over at 29v, the lithium then discharged slightly and continued to charge the lead until the voltages stabilised at around 27v so continued the absorption charge of the lead even after the sun went down.

On discharge, from 27v down to 26v the lithium was supplying almost all the load.
From 26v down to 25v, the lead started to supply some of the load.
At 25v lithium and lead were supplying exactly half and then the voltage hardly moved down to 24.5v by the morning.

My first result is encouraging, no dramas , definite increase in capacity, better voltage stabilisation on discharge and batteries accepted the max charge from the Sunny Island all the time without throttling back. This is an advantage I had not even thought of.

Going forward I need to sort out some sort of protection circuit for the lithium as I cannot risk leaving it day after day charging the lithium to near 100%.

Watch this space     …………………………..........

[rogeriko]

There's hundreds on fleabay https://www.ebay.co.uk/sch/i.html?_from=R40&_sacat=0&_nkw=li-ion+protection+7s

[knighty]

if you search for BMS it will give you a lot of options

there's a hell of a lot to chose from... most will just monitor the cells and dump power into resistors (making heat) to discharge any cells which are over discharging

but... you can get fancy ones which will dump power from high cells into low cells


if you parallel up the balance wires, then you'll only need one BMS for the whole lot :-)

[linesrg]

nowty,

Now I know what the inside of my battery looks like!!!

I can see it wouldn't be that difficult to re-wire things internally to allow the cells to be charged by an external power supply. As I may have noted elsewhere simply connecting the positive and negative leads from the Growatt battery to a charger does nothing as the circuitry obviously needs a signal from the Growatt SP2000 unit.

I some very heavy duty (Albright) DPDT solenoids which can be used for this.

Regards

Richard

[biff]

Wow,
    Exciting stuff Nowty,  I wonder what it would take to put together a 120volt bank, ?
   Could I chain a series of Growatts together, ?
                                       Biff
    I guess the thing to do would be to study the suitable BMSs first and then try and find the way to keep them happy.
 I could sell my Lead acid friends..

[Westie]

Assuming the SI is set for charging LA batts  SI will trickle charge as is normal for LA,  so isn't there a danger of overcharging the Li batteries during trickle charging?

I would be thinking about protecting against fault current flowing between banks  by adding a fuse between them?

Would be nice to get a schematic of the Growatt BMS, if the 'authorisation' handshake could be faked the BMS would probably still operate.

[Fionn]

Would it not make more sense to connect the pack in parallel with one of your existing growatt units, even if at the module level?

If you're looking for a cheap but useful BMS, this one is worth a look:
https://www.aliexpress.com/item/14S-Smart-Li-ion-Battery-PCB-board-Bluetooth-Mobile-BMS-for-48V-58-8VLi-Ion/32794186005.html

The majority of the ones on eBay have no adjustability on their settings so you can't vary the over voltage and under voltage limits.

Was there any fusing between the parallel packs in each module out of interest?

If you held onto the existing BMS it might be of interest to someone with a bare Growatt and other batteries.

[nowty]

Project Updates,

Lined the shed with cement board to prevent the odd fire mishap. It won’t survive a full-on lithium fire, but it will mitigate against scorching by blown fuses, loose connections, etc.

Added an isolator.

Added fuses for each parallel battery string.

Picked up another cheap Growatt, so I now have four battery strings.




Added a DC isolator to the lead batteries so I can just run off the lithium if I want to. In fact, in future, I am thinking of only running on lithium during the winter months.



I have been manually balancing the batteries by,

Using a 50w resister to reduce down any high cells. I found one cell which was 0.2v too high and bled it off for a couple of hours with a 1 ohm resistor.


Using a bench low voltage supply to charge up any low cells, I found one which was 0.4v too low. Five hours of individual charging brought it back up in line with the others.



Apart from those two cells all other cells were found to be within 0.05v of each other and all are now  keeping balanced after several charge / discharge cycles.


I have been manually isolating the lithium when the bank voltage hits around 28v (cell voltages of 3.8v and 80% SOC), and then continuing the charge on the lead up to 30v. At the end of the charging day I reconnect the lithium bank.
I have ordered a heavy duty normally closed DC contactor relay which will be opened and closed by a voltage sensing relay to perform this task automatically.

[biff]

Wow Nowty,
         I am green wif impreszions. You have some crazy good gear. Yours is the battery thread of the century. Real porn.
       Biff

[knighty]

you should really parallel up each cell in the packs

you could just join the balance connectors ?

should really have a fuse on each one too!


advantage is if one cell pack is a bit weak it should average out a bit with the others :-)

[nowty]

you should really parallel up each cell in the packs

you could just join the balance connectors ?

should really have a fuse on each one too!


advantage is if one cell pack is a bit weak it should average out a bit with the others :-)

I actually did this with the first two strings, i.e. connect the balance connectors together via pin headers, but I was concerned that I would inadvertently short one out so I cut them off.

I am experimenting using some cheap crop clip test leads to parallel up in a couple of places to improve the balancing but I was only going to do this temporarily, then monitor cells over a long time to see how much the cells  drift. I agree that permanent linking should require fuses.





And today my heavy duty DC relay arrived but I am still waiting for the voltage sensing relay to wing its way from China.

[fourfootfarm]

Croc clips pretty much are fuses!

Really interesting stuff will be keeping this in mind for future upgrading

[nowty]

Picked up another two Growatts which were sold with red lights on the battery pack so were cheaper than usual. Batteries were fine, one had a low cell which needed some individual charging and the other had batteries down to 10% SOC so just needed charging up a bit. I think I might do away with my lead acid and just run on the lithium.


Other changes,
Upgraded the isolator as the original one was running a little hot.

Added the isolating relay and tested it at a charge rate of 80 Amps. I was surprised as it did not even spark when breaking the connection.
But I am still waiting for my voltage sensing relay from China, so i am still having to disconnect the lithium manually.

I am currently making up some fused battery links to connect all the cells together.

[knighty]

how big is your lead acid pack?

20kwh of LiIon is quite a bit?


how's the usable kwh or LiIon vs lead acid compare?

[biff]

Neat work Nowty,
                        Biff