When will the 18650 explode?

[Jordi]

I have assembled a small setup 4S3P with 18650 li-ion batteries. They are connected to a MPPT tracker and a 40W solar panel. They feed small loads at around 50W max. in total. In order to do things properly I should have to add a BMS to each case of 4S batteries to keep them balanced but the cheap BMS's I bought do not really work properly, they should be advertised as space heaters instead.

I have fully charged all batteries one-by-one before adding them to the setup; they come from the same batch and are same age and use.
My question is, how often should I repeat this procedure to ensure the cells do not get too unbalanced?
I assume a bit of unbalance, even though not good for battery life, it should be tolerable.

 

[RCinFLA]

I assume you are referring to LiPo 18650 cells.

This issue that imbalance causes is at near full discharge or full charging where a single cell is over-depleted or over charged.

You need to have correct undervoltage and overvoltage sensing with series disconnect and the voltage values must be correct for the type of lithium-ion chemistry you are using.

Some of the cheap BMS's are not real clear what type of chemistry they are intended for. For LiPo you need low voltage cutoff no lower than 2.8vdc and overvoltage cutoff of 4.3v.

18650 LiPo cells are typically 2,700 to 3,400 mAH capacity range. There is no such thing as 10,000 mAH capacity 18650 cells you may see advertised.

General charging criteria for LiPo is charge current between 0.3 to 0.5 of capacity in amps maximum although there are cells designed to take higher charge rate. Absorb charge voltage maximum is 4.2 vdc. Current should decay at absorb voltage but there should also be a timeout timer to terminate absorb charging after about 30 minutes. It will degrade a LiPo cell if you leave it at 4.2v for long period of time, which is why you should not leave your cell phone on charger all night long. Some phones are smart enough to terminate absorb charging when current drops off.

Any cell that has dropped below 1.0v is considered bad and should not be charged as it has the possibility of metal dendrite shorts that will overheat cell and cause explosive outgassing fire if significant charge current is applied. Many laptop battery packs set a bit in an EEPROM memory chip to permanently disable battery to prevent charging a pack with a bad cell. It is very serious that they are not charged with a bad cell below 1.0 vdc.

Between 1.0v and 2.8v the initial charge current must be kept low, 50-100mA max, until cell rises above 3.0 vdc. This checks for dendrite shorts. If shorts are present, the low charging current will not allow cell voltage to rise and avoids overheating the cell. It should not take more than a few minutes for a good cell to rise above 3.0 vdc. If it takes a lot longer the cell has leakage shorts and should not be used.

Lithium ion LiPo batteries are easy to charge but difficult to charge safely.

 

[Jordi]

I assume you are referring to LiPo 18650 cells.

This issue that imbalance causes is at near full discharge or full charging where a single cell is over-depleted or over charged.

You need to have correct undervoltage and overvoltage sensing with series disconnect and the voltage values must be correct for the type of lithium-ion chemistry you are using.

Some of the cheap BMS's are not real clear what type of chemistry they are intended for. For LiPo you need low voltage cutoff no lower than 2.8vdc and overvoltage cutoff of 4.3v.

18650 LiPo cells are typically 2,700 to 3,400 mAH capacity range. There is no such thing as 10,000 mAH capacity 18650 cells you may see advertised.

General charging criteria for LiPo is charge current between 0.3 to 0.5 of capacity in amps maximum although there are cells designed to take higher charge rate.

Any cell that has dropped below 1.0v is considered bad and should not be charged as it has the possibility of metal dendrite shorts that will overheat cell and cause explosive outgassing fire if significant charge current is applied. Many laptop battery packs set a bit in an EEPROM memory chip to permanently disable battery to prevent charging a pack with a bad cell. It is very serious that they are not charged with a bad cell below 1.0 vdc.

Between 1.0v and 2.8v the initial charge current must be kept low, 50-100mA max, until cell rises above 3.0 vdc. This checks for dendrite shorts. If shorts are present, the low charging current will not allow cell voltage to rise and avoids overheating the cell. It should not take more than a few minutes for a good cell to rise above 3.0 vdc. If it takes a lot longer the cell has leakage shorts and should not be used.

Lithium ion LiPo batteries are easy to charge but difficult to charge safely.

Thank you for the great explanation. I seem to understand that If I have a high cut-off voltage and a low stop-charge voltage I may mitigate a bit the effects of the growing unbalance at least.

 

[RCinFLA]

Thank you for the great explanation. I seem to understand that If I have a high cut-off voltage and a low stop-charge voltage I may mitigate a bit the effects of the growing unbalance at least.

That is based on every cell being monitored. Balancing is a separate item that corrects imbalance. For LiPo, the balancing dump on a cell, typically a resistor load dump, starts about 4.0 vdc cell voltage during charging. It is typically only 10-20 mA for an 18650 LiPo cell.

At a minimum, imbalance will reduce the extractable capacity of battery pack, limited by lowest state of charge cell and highest state of charge cell boundaries.

 

[Jordi]

Thank you for the answer again, apreciated. Seems there is no big risk on what I am doing then. The purpose of this setup is to use the 40W panel to feed a 50W load and the battery acts as a buffer for when a cloud passes by. So less capacity is not a problem, then I should full charge the cells individually once in a while to keep them a little balanced.

 

[meetyg]

Do yourself a favor, and don't burn down your house: Get a BMS.
JBD has some nice small 20-35a BMSs for less than 50$.
You can configure low and high voltage cutoff, and have some balancing. Bluetooth monitoring also included.

 

[OffGridInTheCity]

I've been operating an 18650 DIY lithium-ion Powerwall for 5 years now - and the main reason I don't worry is that I have a BMS that works (emergency cut-off, especially for overcharge) + can report on the voltages + operate within a safe voltage range.

I also have several small 7s7p 18650 lithium-ion DIY batteries in my APC UPSs but still use a BMS and.... check on them occasionally to see what the BMS is reporting. I expect them to be within <=30mv max difference kind of thing and make sure they're operating within safe voltage ranges. A key danger is overcharge, especially if sustained for many hours - this can start a fire for sure.

I use Chargery for small batteries with it's external display but you might go with Bluetooth / phone - easy to check now and then.

 

[MattiFin]

I assume you are referring to LiPo 18650 cells.

nitpicking but 18650 cells are typically not LiPo (lithium polymer electrolyte) but rather use liquid electrolyte. (4.2 volt cells are often referred as ”Li-ion” but LiFePo4 is also one possible Li-ion chemistry.)

dunno what these should be called. There is several possible chemistries around 4.2 volts and the common nominator for all of them is that they are not as safe as LiFePo4.

 

[OffGridInTheCity]

nitpicking but 18650 cells are typically not LiPo (lithium polymer electrolyte) but rather use liquid electrolyte. (4.2 volt cells are often referred as ”Li-ion” but LiFePo4 is also one possible Li-ion chemistry.)

dunno what these should be called. There is several possible chemistries around 4.2 volts and the common nominator for all of them is that they are not as safe as LiFePo4.

The OP specifically said "...with 18650 li-ion batteries...". Lithium-ion is significantly different than LifePo4 in several ways.
Lithium-ion is typically (assumed) 4.2v max and 3.0v-2.8v low for discussions unless called out.

 

[MattiFin]

The OP specifically said "...with 18650 li-ion batteries...". Lithium-ion is significantly different than LifePo4 in several ways.
Lithium-ion is typically (assumed) 4.2v max and 3.0v-2.8v low for discussions unless called out.

”LiFePo4 is also one possible Li-ion chemistry.)”

 

[OffGridInTheCity]

”LiFePo4 is also one possible Li-ion chemistry.)”

Retracted.

 

[augiedoggy]

LifePo4 is one possible chemistry of the "Lithium" family but not a subset of Lithium-ion. Lithium-ion and Lithium Iron Phosphate are 2 distinct chemistries under the Lithium family.

I got into a bit of heated discussion about this myself as I would agree with you but I was shown data and articles at the time that stated both are technically a form of lithium ion battery. maybe those articles were wrong? I dont know

 

[OffGridInTheCity]

I got into a bit of heated discussion about this myself as I would agree with you but I was shown data and articles at the time that stated both are technically a form of lithium ion battery. maybe those articles were wrong? I dont know

I think my concern with this thread and why I commented at all, is the OP specifically stated 18650 lithium-ion... which means a specific max voltage for charge, specific BMS characteristics, safety issues, and these types of things which are important to safely and successfully build/operate a DIY battery. LifePo4 almost certainly does not apply in this case. Yes, there is such as thing as an 18650 LifePo4 format, but its not nearly as common.

Googling.....
This Wikipedia link - https://en.wikipedia.org/wiki/Lithium_iron_phosphate_battery - starts it's first sentance with "
The lithium iron phosphate battery (LiFePO 4 battery) or LFP battery (lithium ferrophosphate) is a type of lithium-ion battery using lithium iron phosphate (LiFePO4) as the cathode material,".....

And this link - https://www.pnnl.gov/lithium-ion-battery-lfp-and-nmc - talks similarly.

And this link - https://www.electronicdesign.com/po...iumion-cells-and-chemistries-you-need-to-know

So it would seem I learned something today.