18650 cells. What is the life-expectancy.

[march.brown]

I have eight 18650 cells and four torches ... The Romisen RC-U4 in the car will probably have its cell changed three times a year ... The three Solarforces will have their cells changed every four to six weeks approximately ... A spare will be put in the torch and the other recharged and put back in with the spares.

On average these 18650s will each be charged maybe six times a year ... So after five years they will each have been charged about thirty times.

My Solarforces (on Maximum) discharge down to 3.8 volts in ninety minutes , so still have about an hour left ... My intention is to repeat this discharge test about once a year and if the voltage drops below 3.7 volts after ninety minutes , that cell would be retired and a new one sourced.

Is any data available that can guess-timate the life expectancy of a lightly used 18650 cell or is life expectancy based on the life of laptop battery packs ?
.

 

[BeeMan458]

Unfortunately there's a whole charging culture that revolves around charging and discharging 18650's. If you hunt around this forum, you'll find lot's of information regarding the care and feeding of rechargeable Li-ion batteries.

As to cheaper one, like the ones I buy, 200-300 times. The better brands, I feel comfortable writing 300-500 times.

Hope the above helps.

 

[1 what]

BM458 you must get a better brand of "cheaper ones" than I do.
My record was a "middle level" (ie branded with a known brand) 18650 that died after less than 10 charges from a Pila.
These days I only buy AW's.

 

[BeeMan458]

No. Apparently, I bought the worst, evil rechargeable 18650 Li-ion's that one could ever buy on the face of the earth; UltraFire, protected, 3000mAh.

I think I hear them right now plotting to blow me up as I write this response.

 

[march.brown]

Unfortunately there's a whole charging culture that revolves around charging and discharging 18650's. If you hunt around this forum, you'll find lot's of information regarding the care and feeding of rechargeable Li-ion batteries.

As to cheaper one, like the ones I buy, 200-300 times. The better brands, I feel comfortable writing 300-500 times.

Hope the above helps.

.
My cells are all Ultrafire ... Six are grey colour 2400mAh protected and two are blue non-protected ... None are expensive.

They are all charged to 4.18 or 4.19 Volts on my Soshine SC-S2 charger.

Since my cells are only charged on average six times a year , I was wondering how many years they will last ... So the fact that they can be charged several hundred times doesn't affect me ... 300 charge/discharge cycles would last me fifty years ... I guess the cells would have died long before they were fifty years old ... There again , as I am now 72 years old , if they lasted fifty years I wouldn't know about it !
.

 

[jasonck08]

I think generally the capacity of the cells will drop about 10% a year. Also as a general rule of thumb, Li-on cells like to be stored at ~3.9v or 40% capacity. Leaving them at 4.2v for a while is not terribly good for the cell. I've had cheapo Chinese cells not last 10 cycles before, but I've had laptop packs last probably 300-400 cycles with Japanese cells.

 

[VidPro]

.
My cells are all Ultrafire ... Six are grey colour 2400mAh protected and two are blue non-protected ... None are expensive.

They are all charged to 4.18 or 4.19 Volts on my Soshine SC-S2 charger.

Since my cells are only charged on average six times a year , I was wondering how many years they will last ... So the fact that they can be charged several hundred times doesn't affect me ... 300 charge/discharge cycles would last me fifty years ... I guess the cells would have died long before they were fifty years old ... There again , as I am now 72 years old , if they lasted fifty years I wouldn't know about it !
.

3-4years MAX. every UF cell i purchaced a few years ago is finished, the newer ones i wont know till they too are gone, if they changed anything since wont know for sure till they actually start lasting time.
Just watch for the clues, wont charge fully (very bad), self discharges, heats on charger (very bad) , wont recover from 0V (anode disconnect), Stinks, corrodes something, leaks tiny stuff on wrapper, suddenly rusts from corrosive stuff, and dont use them when you get those clues, because bad can be dangerous.
capacity testing is a good idea, but when they already wont charge proper , the capacity will be bad, and if anything was going to go horribly wrong, it is more likely to be on the charger itself. so if you spot things going badly there already, and the charger is still working, then they would be deemed finished, and the capacity testing will pretty much confirm it. Arggg, all i am saying is when it is going badly on the charger, you already dont want to be charging it then.


use em or lose em. if your just storing them store them down in voltage a bit, they might last a little longer.
naw you will live to 114years old, and will have replaced your cheap li-ions another 10 times. dont let mother nature fool you, she never gives up as easily as a cheap li-ion does

 

[march.brown]

3-4years MAX. every UF cell i purchaced a few years ago is finished, the newer ones i wont know till they too are gone, if they changed anything since wont know for sure till they actually start lasting time.
Just watch for the clues, wont charge fully (very bad), self discharges, heats on charger (very bad) , wont recover from 0V (anode disconnect), Stinks, corrodes something, leaks tiny stuff on wrapper, suddenly rusts from corrosive stuff, and dont use them when you get those clues, because bad can be dangerous.
capacity testing is a good idea, but when they already wont charge proper , the capacity will be bad, and if anything was going to go horribly wrong, it is more likely to be on the charger itself. so if you spot things going badly there already, and the charger is still working, then they would be deemed finished, and the capacity testing will pretty much confirm it. Arggg, all i am saying is when it is going badly on the charger, you already dont want to be charging it then.


use em or lose em. if your just storing them store them down in voltage a bit, they might last a little longer.
naw you will live to 114years old, and will have replaced your cheap li-ions another 10 times. dont let mother nature fool you, she never gives up as easily as a cheap li-ion does

.
Many Thanks for your reply , in particular your quote on my life expectancy ... If you are correct then I won't mind buying another ten sets of cheap lithiums ... If your quote was true , I would even consider buying the more expensive cells that last longer.

Anyway , I will probably just keep on checking the voltages every month or so and recharging when necessary ... Once a year , probably in March to celibrate my birthday , I will do a test discharge ... There are only eight 18650 cells to discharge and four torches to discharge them with , so it wouldn't take long ... I do this whilst I am on the PC and I check the voltages every half hour ... When the voltage has dropped from 4.2V to 3.8V , I will compare it with the original figures (90 minutes) and then I can make the decision as to replace the cells or not ... I doubt whether all the cells will need replacing at the same time , at least I hope not ... I will probably reduce the number of cells so that there are only two cells spare rather than four , either that or get another 18650 torch ! ... But I don't really need another , not really ... However , a nice grey Solarforce L2 would be nice as my three are in the lovely sand colour.

Thanks again for your reply and to all the others who also took time to reply to my thread.
.

 

[Niconical]

Based on your usage predictions, number of charges isn't an issue, it's the 5 years.
I'd say that by then they would be done or near done, whether it's 30 charges or 130.

 

[Black Rose]

Since my cells are only charged on average six times a year , I was wondering how many years they will last.

In your case, the number of charge cycles won't really matter, it's time.

As soon as a Li-Ion cell is manufactured, it immediately starts "dying" and the capacity loss is permanent...it's not something you can bring back by doing some charge/discharge cycles on.

The age and quality of the "base" cell that is used can also be a factor.
In your case with the Ultrafire batteries, the cells you have could've been several years old when they were made into "flashlight batteries".

 

[mdocod]

On proposed test concept:

That test routine for determining the condition of the cells sounds like a neat idea that could probably work pretty well if the power consumption of the discharge was expected to remain constant....

The problem I see, is that, as cells age, they are going to suffer from voltage suppression combined with an increase in cell resistance. Both combine to have the same total effect on the circuit, which is to have less available voltage under a given load.

If the flashlight in question had a very tightly regulated power consumption regardless of variations in input voltage, then the test would work great, but if it's running in a sort of "direct drive" fashion on a single 18650, then the reduced voltage performance of the cell under the load can actually end up translating to increased total runtimes with lower output.

I'm not aware of very many flashlights that run on a single 18650 that maintain [nearly] continuously even power consumption through a discharge. If your light diminishes in output through the run, then this test method won't work unfortunately.


-----------

On li-ion cell life:

Funny thing about trying to predict the life of a li-ion cell, is that, the only cells we really have good information on are the ones that have died. So we may be able to tell you which cells to buy or when to throw them out sometime in the past with a time machine, but as cells continue to improve, predicting the life of a cell bought in our little market is not an exact science. Average cell temperature and state of charge play a significant role in my understanding.



I have some AW 17670 size cells (BLUE label protected) from back in 2006 that still work very well. I just pulled one off the shelf to check it's voltage, (I think I topped up all of my cells about a month ago), it's holding 4.17V after a month of rest at over 4 years old. These cells are still in very good shape considering age and will probably continue to give me good service for several years to come. I use them occasionally in my headlamp but they didn't always see such mild service. When I first got them back in 2006, I was using them to drive incandecent lamps with ~1C discharge rates (based on cell capacity) for awhile there.

One of the easier ways to determine the health of a cell, is to see what kind of voltage it will hold, or sag to, after having been rested right after a charge. Cells in very good condition will hold around 4.20V for a very long time. A good quality unprotected cell charged to 4.20V and stored in a relatively cool place will often still read over 4.15V or better (as high as 4.18V) after a year of storage. As cells start to age, they will quickly settle down to lower voltages after their charge. Cells that are reading 4.10V a day after being charged to 4.20V are pretty beat up but can still be used. Cells that won't hold above 4.0V after a few days or even a week should be considered totally shot. I've noticed that as cells age, there seems to be a combination effect of both a semi-rapid voltage settling after a charge (first 24 hours) followed by what would be better described as semi-rapid-self-discharge that seem to cause the cell to continue to decline in a noticeable fashion for the days and weeks thereafter. (the voltage will just keep dropping and dropping, right down to ~50% state of charge over the course of a month or so).

As it wold turn out, the ability to hold a charge, which is reflected in cell voltage, is directly correlated to the condition of the cell in terms of age and wear and tear. When your cells are getting down around 4.00-4.05V after a day of resting after a charge, they are probably pretty far gone. Monitoring that weekly discharge rate thereafter seems to also be important. (Sometimes, a cell will settle to 4.10V, and then self discharge to 4.08 after a week, but if it's below 4.0V after that week, then things may be worse than previously thought).

---

UltraFire cells are going to be very hard to "predict." Looking at the history of cells carrying the ultrafire brand name over the last several years, we find that up until recently, most of them were total and complete garbage. Often they were under-performing from label capacity by as much as half, combined with massive variations in capacity from cell to cell. In the last year or so, we have seen some dramatic improvements in capacity and performance that are likely related to an improvement in the quality of the cells being chosen to slap the UF name on. UF cells still tend to be on the bottom rung overall for performance under a load, but the leaps and bounds made under this brand name (from super crap to less crappy) might suggest some better cell life to boot.


I guess what I am saying is that. A few years ago, UF cells were duds right out of the box, but there are some out there made in the last year or 2 that seem to be of a noticeably better quality. If the absolute best Name Brand LiCo cells on the market made today are good for 10+ years under mild conditions, perhaps the best UF cells offered recently will hold up for 2-5 years.

Keep an eye on them, dispose of them properly when you feel uncomfortable with them, and move on to something you can trust a little more.

Eric

 

[march.brown]

On proposed test concept:

That test routine for determining the condition of the cells sounds like a neat idea that could probably work pretty well if the power consumption of the discharge was expected to remain constant....

The problem I see, is that, as cells age, they are going to suffer from voltage suppression combined with an increase in cell resistance. Both combine to have the same total effect on the circuit, which is to have less available voltage under a given load.

If the flashlight in question had a very tightly regulated power consumption regardless of variations in input voltage, then the test would work great, but if it's running in a sort of "direct drive" fashion on a single 18650, then the reduced voltage performance of the cell under the load can actually end up translating to increased total runtimes with lower output.

I'm not aware of very many flashlights that run on a single 18650 that maintain [nearly] continuously even power consumption through a discharge. If your light diminishes in output through the run, then this test method won't work unfortunately.


-----------

On li-ion cell life:

Funny thing about trying to predict the life of a li-ion cell, is that, the only cells we really have good information on are the ones that have died. So we may be able to tell you which cells to buy or when to throw them out sometime in the past with a time machine, but as cells continue to improve, predicting the life of a cell bought in our little market is not an exact science. Average cell temperature and state of charge play a significant role in my understanding.



I have some AW 17670 size cells (BLUE label protected) from back in 2006 that still work very well. I just pulled one off the shelf to check it's voltage, (I think I topped up all of my cells about a month ago), it's holding 4.17V after a month of rest at over 4 years old. These cells are still in very good shape considering age and will probably continue to give me good service for several years to come. I use them occasionally in my headlamp but they didn't always see such mild service. When I first got them back in 2006, I was using them to drive incandecent lamps with ~1C discharge rates (based on cell capacity) for awhile there.

One of the easier ways to determine the health of a cell, is to see what kind of voltage it will hold, or sag to, after having been rested right after a charge. Cells in very good condition will hold around 4.20V for a very long time. A good quality unprotected cell charged to 4.20V and stored in a relatively cool place will often still read over 4.15V or better (as high as 4.18V) after a year of storage. As cells start to age, they will quickly settle down to lower voltages after their charge. Cells that are reading 4.10V a day after being charged to 4.20V are pretty beat up but can still be used. Cells that won't hold above 4.0V after a few days or even a week should be considered totally shot. I've noticed that as cells age, there seems to be a combination effect of both a semi-rapid voltage settling after a charge (first 24 hours) followed by what would be better described as semi-rapid-self-discharge that seem to cause the cell to continue to decline in a noticeable fashion for the days and weeks thereafter. (the voltage will just keep dropping and dropping, right down to ~50% state of charge over the course of a month or so).

As it wold turn out, the ability to hold a charge, which is reflected in cell voltage, is directly correlated to the condition of the cell in terms of age and wear and tear. When your cells are getting down around 4.00-4.05V after a day of resting after a charge, they are probably pretty far gone. Monitoring that weekly discharge rate thereafter seems to also be important. (Sometimes, a cell will settle to 4.10V, and then self discharge to 4.08 after a week, but if it's below 4.0V after that week, then things may be worse than previously thought).

---

UltraFire cells are going to be very hard to "predict." Looking at the history of cells carrying the ultrafire brand name over the last several years, we find that up until recently, most of them were total and complete garbage. Often they were under-performing from label capacity by as much as half, combined with massive variations in capacity from cell to cell. In the last year or so, we have seen some dramatic improvements in capacity and performance that are likely related to an improvement in the quality of the cells being chosen to slap the UF name on. UF cells still tend to be on the bottom rung overall for performance under a load, but the leaps and bounds made under this brand name (from super crap to less crappy) might suggest some better cell life to boot.


I guess what I am saying is that. A few years ago, UF cells were duds right out of the box, but there are some out there made in the last year or 2 that seem to be of a noticeably better quality. If the absolute best Name Brand LiCo cells on the market made today are good for 10+ years under mild conditions, perhaps the best UF cells offered recently will hold up for 2-5 years.

Keep an eye on them, dispose of them properly when you feel uncomfortable with them, and move on to something you can trust a little more.

Eric

.
Thanks Eric for your thoughts ... I will certainly take them on board.

If I am forced to buy fresh Ultrafire (or similar) 18650 cells every three years or so , perhaps I will be lucky enough to come across some of the "improved" ones ... Unfortunately , I wouldn't know how good they would be till they also died ... If they are going to get better quality (over the years) then I can't see the point in paying top prices for the "posh" ones.

Since I have only got eight 18650 cells at the moment , I will have to try to keep a log on their performance so that I will be able to estimate better their retirement date ... I will just number them from one to eight and keep a card index of charge dates and voltages for reference ... Good job there are only eight cells to monitor.

I won't be storing my cells at a lower voltage (3.9V) as it is sort of defeating the object of having spare cells ... They will also be stored at room temperature of about 60 to 65F ... I want to be able to pop in a reasonably fully charged cell when it is needed rather than have to charge the replacement up before use ... If I was going to follow the path of correct storage temperature and voltage , I would prefer to simply take the discharged cell out of the torch , charge it and put it back in ... That way I wouldn't need any spare cells (well maybe one) ... So I will be treating my cells rather badly ... If I get three years of this mis-use before I need to buy replacements then that's OK as they are not expensive cells ... Even some of the cells kept in the torches will be little used and I may not have to recharge them for months , so they are not being treated well either.

I chose to use 18650 cells because they had a lot of energy for the small volume , so perhaps I will have to just put up with having to replace them earlier than other more careful users ... My three Solarforce torches are the perfect size for me when I compare them with the Mag 2C that I used to carry ... To me a single 18650 is the best alternative in every way ... Since I don't need a torch for security work , I feel that I have (for my money) chosen the best light system ... I do also have my EDC (iTP A2) with me all the time , so I shouldn't be "sans lumiere" or "sans lumens", hopefully.
.

 

[jimbo627]

18650 Lifespan

I am a little new to battery revolution but am trying to get caught up as quick as I can. I have been reading as much as I can and I hope some of it actually sticks in my thick head, but I am still a little fuzzy on one point. I have quite a few 18650's that I have been using for years in hunting lights and I couldn't tell you how many times I have recharged them. I know with the Sanyo eneloops, they claim they can be recharged as many as 1500 - 1800 times but I have not seen info. on how many times the LI-ON batteries can be recharged. I recently purchased 4 of the ORBTRONIC 3400 mAH 18650's and would like to know what to expect from them and what is safe. Do they have a finite shelf life or are they good for a certain number of charges? Thanks for any light you can shed on this for me.

 

[ven]

Re: 18650 Lifespan

Tbh it varies on if you drain right down or top up regular,as a 2/3 drain and charge or a top up will lengthen cycles so to speak

Topping up regular and not running down to 3v,not storing at 100% charge,40% ideal all adds life.You will notice that (example here) your usual 2hr run time starts dropping to maybe 1hr after a time,then you know time to replace(as said example as 1hrs use is still useful for some applications)

Years and years for me,i have cells 4-5ys old issue free,not sure how many charges tbh........

Sure some will well out do me on that:thumbsup:

get a multimeter(if not already and under $10),check every so often is another way to spot any untoward issues,dont mean daily,maybe monthly,every 6 months.........

Normally they say 500 cycles,but "topping off" is not a cycle,in fact you could top a cell off 5-10 times and maybe equivalent to 1 cycle

 

[hikingman]

Re: 18650 Lifespan

ven,
When I top off a cell, I then run it on high in the light for about 2 minutes. Since I have many lights, 3 or 4 "topped off" cells may be unused for weeks or even months, subject to just some slow drain perhaps. Is this OK or not OK practice?

Dave

 

[UnderPar]

Re: 18650 Lifespan

Tbh it varies on if you drain right down or top up regular,as a 2/3 drain and charge or a top up will lengthen cycles so to speak

Topping up regular and not running down to 3v,not storing at 100% charge,40% ideal all adds life.You will notice that (example here) your usual 2hr run time starts dropping to maybe 1hr after a time,then you know time to replace(as said example as 1hrs use is still useful for some applications)

Years and years for me,i have cells 4-5ys old issue free,not sure how many charges tbh........

Sure some will well out do me on that:thumbsup:

get a multimeter(if not already and under $10),check every so often is another way to spot any untoward issues,dont mean daily,maybe monthly,every 6 months.........

Normally they say 500 cycles,but "topping off" is not a cycle,in fact you could top a cell off 5-10 times and maybe equivalent to 1 cycle

I just learned something here. Thank you Ven!

 

[ven]

Re: 18650 Lifespan

ven,
When I top off a cell, I then run it on high in the light for about 2 minutes. Since I have many lights, 3 or 4 "topped off" cells may be unused for weeks or even months, subject to just some slow drain perhaps. Is this OK or not OK practice?

Dave

I am no expert Dave ,but its not far off what i do,i top my cells off, allow a few minutes from charge(can put in light or leave out just i do not power on straight away) ,power up for a minute or so then thats it.Few weeks is fine,i think its more if its stored at near 100% for months on end it wont help the cell life.I have a cut off of around 1 month,if its not been used at all i will use the light for a few minutes for example or cell thats in question.It may be 2 months............no biggie to me

Depends also on how many cells you have,sounds a little nerdy but if you have lots ie 30+ and maybe you find you are just using 3 lights,then its worth storing the other cells at 40% charge or rotating them....................please note that is just an example.:thumbsup:


Topping off is good practice imo,as topping off the cell is not counted as a charge cycle,so "5 top offs" maybe 1 cycle.........if makes sense as its not a deep charge.

Month imho is fine,i would not leave longer than 3 months personally,some are just more cautious, i am presuming here we are using good cells,maybe protected cells.

Digression here........
Chances are a new cell will come out or a deal on some 3600mah cells maybe next year..................guess i am trying to say before the cells goes duff from lots of recharging and wont hold charge,you will have probably replaced it anyway with a new one...........

Dont get hung up too much,you know probably more than me:thumbsup:.

Bad practice is draining cell right down regular,storing for long times at 100%...........if you dont do that then your fine and better life will be yours to behold:thumbsup:

Sorry for waffle,would have been easier just to say that


:twothumbs

 

[ven]

Re: 18650 Lifespan

I can give an example of my VV deivices on which get used pretty much all the time apart from when asleep :laughing:

I use unprotected cells,not smoked now(cigs) for 14 month,but turned into a little VV mod geek as in i like my tanks/mods but thats by the by

The cells until recently(now use 3400 unprotected pany in one device) have been samsung and sanyo,only 2600mah,get used daily,topped up daily,swap around and rotate.
When charged its from around 1/3 to 1/2 left in cell(never measured voltage tbh before/after) but could do if required.
These cells are around 14 months old,daily top ups,used daily and still spot on today.No issues,no short use down to cells not holding charge.It is probably equivalent to using an 18650 for around 1/2 the charge a day but with higher Amps,my amp are around 7.2A on average to 8A how i set it up,this is more than most lights draw unless modded but for a shorted time,just regular if makes sense.So 8A for around 6 seconds,then again after a short break for a few minutes at a time.

It really is a "how long is piece of string" question,its down to how you look after your cells,what they are feeding,for how long they are feeding the device(time on turbo for example) to what level of power(is it 200lm or 1000lm),but 5yrs general use from a cell is not unheard of.I have older cells up to 4-5yrs old,still fine.........heavy use,deep charging reduces life,top up after shorter use will prolong life........

But its not like cells are $100 each,so just general common sense,looking after and all will be happy:thumbsup:

 

[ven]

Is any data available that can guess-timate the life expectancy of a lightly used 18650 cell or is life expectancy based on the life of laptop battery packs ?
.

Old question i know just to point out :laughing:
In simple terms the answer would be NO as totally different,laptops tend to be left on,battery is always being charged(in general unless battery saver options where it allows it to drain down to 80% then top up).

Also if you buy an expensive laptop with genuine battery,the cells will be "usually" better than the $20 replacements on ebay.These are cheap crap so wont last,the 2hrs unplugged will soon be 10 minutes before drained..........

So dont go off that,totally different imho,others opinion may vary..........thats just my view:thumbsup: