BMS 48V questions with temp and size and communication

[mrzed001]

Hi newbie here

I have some questions about BMS
I want to make a 16s 100A 48V system (Fortune cells) mainly for UPS reasons. And safety first ...

1, Is there a way to monitor the temperature of all 16 cells? 16 temp sensor ?

2, a 150A BMS is enough for a 100A system (like this Daly https://www.aliexpress.com/item/4001326967928.html ) ... or bigger is better ?

3, if I want to double the battery Wh what is the best?
Make a second 16s pack (with own BMS) or make a big 2p16s pack (200A) with 1 BMS ?

For 1 pack I think 30-40A max for loading ... but if 2 pack I can not set 2x40A=80A in inverter because if first already full second gets all the 80A (I dont want to hammer it, slow is good)
So if 2 pack and 2 BMS can they communicate with inverter and lower the Amp ?


4, I want to use a MPP solar 5048 inverter. Here it says I can connect BMS to it with RS485 https://www.mppsolar.com/v3/pylontech/
First question is Daly or and other BMS is compatible with it?

Also I do not want to break the MPP warranty and solder its button like Will did

so the second question: can this do the same trick and shut down the inverter (in a battery cell high temp or cell low volt situation) ?

Thanks

 

[BiduleOhm]

1) That would be pretty expensive and useless as all cells will be roughly at the same temp. You can do a sensor at each end and one in the middle for example if you really want some redundancy

2) Bigger is always better. 150 A for a 100 A system is fine tho.

3) Best is two packs with a BMS for each.

3.5) That will be almost impossible without some DIY. And first you need to find an inverter where you can limit the power; I never saw one IIRC, they're kind of on/off devices, not half on.

4) I don't know, sorry. Best option would be to ask the manufacturers.

 

[mrzed001]

1) if one of the cells is (going)wrong, that will overheat (or produce more heat).
It takes a long time to one cell to heat all the 16 cells. Better to catch a problem at the beginning.
Maybe its V will differ from the others at that point ... but is it 100% sure ?
Extra heat is always a good sign to catch a problem before escalates.

Lithium is still dangerous a bit (and I do not want my house to burn down because of a faulty cell)

2) OK

3) OK

3.5) in MPP I can set the inverter to give 2-60A from utility or 10-80A from PV to battery.
For 1 pack 30A is enough. For 2 pack 30A is too low, need 60A. But if 1 is full then 60A for remaining pack is too much (ok not too much but I want slow load). I do not have a solution for this now ...

3.6) If I buy a 2nd MPP inverter (expensive) and put the 2nd pack (with BMS) to that ... can the output AC of the 2 inverter merged ?
(so not spit phase, and not 3 phase ... both AC output in the same sinus wave in one cable ... load balance maybe ?)
And they need total separated solar panels? So I can not share the panels output between them ?

4) I will do that. Just curios if someone else already did it, and what was the result

Thanks

 

[BiduleOhm]

1) But a cell doesn't do that unless there's another serious problem like a short circuit or a big over voltage; and both will be dealt with by the fuse and/or BMS far before it even escalates to that point.

And I thought I was the belt and suspenders champion here ?

3.5) That's the currents for the chargers, not the inverter.

3.6) Depends on the exact model. You need stackable ones to do that.

4) Yep of course, wait for other members answers

 

[mrzed001]

1) or a faulty cells last kick :D Can happen. Did you see a Tesla battery sitting alone on the table explode? I dont, but one of my friends did :D
OK LiFePO4 cells are more stable, but I would not put my (and family) life on it :D
If there is a way, a BMS to put a thermal sensor to every cell ... I will do it.

3.5) Yep, the charging current is the problem. Too low for 2, or too high for 1

3.6) OK, I will ask MPP ... now I have at least 7-8 question to them :D

 

[BiduleOhm]

1) Big difference between NMC and LFP chemistries... NMC will go to thermal runaway past a certain temp, LFP will not. Tesla batteries are NMC.

There's a way but it involves DIY. All commercial BMS I saw have 1 temp sensor, rarely 2, and mine has 3 (which I thought as overkill but I had a spare input on the MUX so...).

3.5) I would be more concerned with the discharge current as they are higher in most systems usually.

 

[ghostwriter66]

You are totally wasting money and effort on taking the TEMP of each battery ... if the LiFePO4 is going to go bad its going to have VOLTAGE issues WAAAAYYY before TEMP issues ....

1000's of LiFePO4's in our inventory -- totally abusing them daily ... heat - sun - arcing - etc -- and NOT one ever exploded ... YES we have them bloat .. BUT never explode ... its you're money ... but just don't want this thread to turn into the typical internet "a friend of mine's brother's nephew knew a guy who had a friend that saw a LiFePO4 explode and kill 100 ppl" stories ... LiFePO4 is safe -- otherwise I would not have my feet propped up on them at the office all day ...

 

[mrzed001]

1) True ... but the needle test simulates when the middle thin layer raptures. And with time even with care and no abuse ... it will. Murphy's law.

3.5) True, only load current can be regulated. Or if 2 pack, then discharge current is divided to 2. A 5kW inverter can have 100A from a single pack. But if 2 pack, then the max is 50A (from both). On the other side if I know what I plug in and how much A (W) it takes ... thats good enough.

You are totally wasting money and effort on taking the TEMP of each battery ... if the LiFePO4 is going to go bad its going to have VOLTAGE issues WAAAAYYY before TEMP issues ....

I am not ... because I can not find a BMS that has at least 8-16 temp sensors
If it costs extra 100$ .... I would not care.

I searched for data about cells going wrong, V decline ... can not find.

 

[BiduleOhm]

3.5) Ok, then perfect, you have all you need to stay under the max amperage you want

I am not ... because I can not find a BMS that has at least 8-16 temp sensors
If it costs extra 100$ .... I would not care.

That's why I said it'll require some DIY. For example you can use some electronic thermostats to measure the temp of each cell and have their outputs do something (opening a contactor, signaling the BMS if it has an input, ...).