Part Charge or Full charge? The Jury's Out!

flecc

Member
Oct 25, 2006
46,532
23,366
Do lithium batteries last longer with frequent part charges? I'm now not so sure due to the following accumulated evidence.

1) I owned three of the original eZee Li-ion batteries which were all treated with care, and since I'm not a commuter most of the charges were frequent and part only. Two failed within 6 months and one lasted only 10 months, while others using the same batteries on the same model bikes, but commuting and exhausting most of the charge each day, were sometimes getting 12 to 18 months use from a battery.

2) My experience with the latest type of li-ion on test hasn't shown any problems with fully discharging most of the time during the test program and it's still in use at 25 months. Others testing this type have in some instances reported failures by 18 months, despite most frequently part charging.

3) Panasonic intend their 10 Ah battery for the pedelec unit to be fully discharged every time any meter or capacity deviation is seen, and they make no mention of a need for frequent charging, instead publicising the full ranges that can be attained. They were also the first, and for a long time the only company giving a two year warranty without restrictive usage conditions attached, indicating no fear that full discharging does harm.

Obviously the first and second cases could be affected by differing conditions like terrain, rider etc, but the accumulation does seem to indicate that fully discharging doesn't do any damage and may even be preferable.

Conclusion: Based on the likelyhood that those part charging will typically carry out 2 to 4 part charges totalling one full charge, it could be that this average of 3 times taken to the full charge point could result in more stress than the stresses of one full charge point and one full discharge point when fully discharging, the latter numerically smaller.

My own intention for the future is to just use the content as convenient with no regard to any special charging pattern.
.
 

Mussels

Esteemed Pedelecer
Jun 17, 2008
3,208
8
Crowborough
My concern with full discharges is when the BMS doesn't monitor individual cells so the weakest cell in the pack would be discharged below a safe level often which weakens the pack and in turn affects the next weakest cell.
For BMS systems that monitor individual cell discharge and shut the weakest cell off before it gets damaged then that problem won't exist.
Maybe if the cells are matched during assembly and they all age the same this won't be a problem. :confused:
 

flecc

Member
Oct 25, 2006
46,532
23,366
I think they mostly rely on cells matching, and as you see, it doesn't appear to be a problem in practice. I think the race for higher capacity could be tempting manufacturers to push the BMS upper charge limit to the "safe" edge, making frequent charging just as damaging. Real safety and long life means ignoring the upper 20% and lower 20% of capacity and only using the centre of charge, but that will never happen on e-bikes.
.
 

Fecn

Esteemed Pedelecer
Sep 28, 2008
491
2
Warlingham, Surrey
I think you need a larger data set before attempting to reach a conclusion.

There will be inconsistencies in the chemical forumlation within the cells themselves which can cause some cells to fail before others regardless of usage pattern. Given that the manufacturing is not 100% consistent then whether a pack works well or not will very much come down to luck-of-the-draw when the cells get assembled into a pack.

I suspect that if you'd had another 4 identical batteries to test in exactly the same way, the results could have turned out totally differently.

I pulled apart a decade-old LiIon pack from my old Sony Vaio last year. That pack will have had well in excess of 1000 cycles in it's time. It had 12 cells in total. 7 of the cells were basically as good as new, giving within 10% of their original rated capacity. The other 5 were totally and utterly dead... short-circuit.

A couple of years ago, I bought 12x 18650 Li-Ion cells from Deal extreme. 6 of them went in a drawer. 6 of them went into service. All the cells got the occasional top-up charge, in keeping with good practice. After 12 months, one of the drawer batteries had died, and one of the in-service batteries had died. The in-service battery had received around 50 charges, and although I originally chucked it in my battery-recycling box, I later fished it out again and discovered that the problem wasn't the cell itself, but the little protection circuit on top. It had died in the sub-zero temperatures.

I suspect all it takes is a fleck of dust in the chemical soup used during manufacturing to end a cell's life prematurely, and although usage pattern also plays a part, it may be less significant than people think.
 
Last edited:

NRG

Esteemed Pedelecer
Oct 6, 2009
2,593
10
Hi Flecc, this subject has been of great interest to me recently and I agree with you that there seems to be no need for any any special charging pattern. I don't use any on my commercial bike or on my DIY bike with RC Lithium batterys except for a balance charge as needed on the DIY pack when the cells drift more than 50mV apart.

However, I would add that there's plenty of practical experience on ES and many web references that show fully discharging Lithium based battery shortens the number of cycles considerably. I think caution needs to be exercised when analysing commercial battery offerings equipped with a BMS. The BMS is capable of capping the discharge and charge levels of the battery and while you may believe you are fully discharging it the BMS could have been set to not take cell voltage down below a certain level or set to not allow a cell go above a certain level....in effect capping the upper and lower limits means the battery is always part charged/discharged.

Whilst your findings are undoubtedly useful I think that unless you can test in a controlled environment no firm conclusions should be drawn IE: Temperature has a drastic affect on battery life, there's a natural ageing process in operation that affects some types of cells more than others, are the the BMS control parameters fully understood and what where the cells used... with an eye on quality and individual cell matching...

Edit, just blindly posted, didn't see Fecn's post but I agree a larger sample under more controlled conditions with a full understanding of what is being tested would be need to draw a conclusion. But as I said above there's a lot of evidence to the contrary of your finding Flecc.
 
Last edited:

flecc

Member
Oct 25, 2006
46,532
23,366
Yes, it's too early for conclusions, but the likely differences seem to be small enough for me not to bother with future precautions. As Fecn says, inconsistencies do exist and could affect the outcome just as much as charging patterns.

For someone like me with very variable usage, really consistent charging patterns aren't likely anyway.
.
 
Last edited:

john

Esteemed Pedelecer
Nov 1, 2007
531
0
Manchester
The BMS needs to protect the cells when charging too. If it doesn't do this so well then fewer charges could be better.

As the BMS will vary from suppliers and from year to year, it makes them difficult to compare.

NB crossed with posts above
 
Last edited:

Straylight

Esteemed Pedelecer
Jan 31, 2009
650
2
If David from Wisper is monitoring this thread, or indeed anyone else with a close relationship with a battery manufacturer, I feel it'd be interesting to get the opinions from the lab on this issue. As an end user, it's been a nagging doubt, and blind spot in my knowledge since taking ownership of my ebike.
 
Last edited:

brucehawsker

Pedelecer
Dec 17, 2009
119
0
We recommend half discharge and keeping the battery up to charge after use. Whether this is old wives or lab rat, but it seems to serve our clients well. Message is buy twice the Wh you think you will need....
 

HarryB

Esteemed Pedelecer
Jan 22, 2007
1,317
3
London
1) I owned three of the original eZee Li-ion batteries which were all treated with care, and since I'm not a commuter most of the charges were frequent and part only. Two failed within 6 months and one lasted only 10 months, while others using the same batteries on the same model bikes, but commuting and exhausting most of the charge each day, were sometimes getting 12 to 18 months use from a battery.


.

I have lost count of the number of original Ezee batteries I had (4 I think). I used them for commuting and mostly discharged them fully (they were always cutting out anyway so difficult to tell). I don't think any of them lasted more than 6 months without giving trouble and was the main reason for buying a bike with NiMH batteries.

I wonder if I am safe to come back to Li-ions yet? I would certainly like a bike that doesn't need to be charged at work.
 

lemmy

Esteemed Pedelecer
My non technical view on this is that you buy an e-bike for a purpose and the best thing to do is to use it as you wish regardless of what is good for the battery.

My use has little pattern, sometimes I'm doing many journeys of a couple of miles, sometimes I'm getting back under my own steam because the battery is done for.

But in the end, the machine is my servant and if my use shortens the potential life of the battery, so be it. This makes the discussion of what's best for battery life practically redundant - but interesting all the same.
 

Dynamic Position

Esteemed Pedelecer
Feb 28, 2009
306
2
My non technical view on this is that you buy an e-bike for a purpose and the best thing to do is to use it as you wish regardless of what is good for the battery.

My use has little pattern, sometimes I'm doing many journeys of a couple of miles, sometimes I'm getting back under my own steam because the battery is done for.

But in the end, the machine is my servant and if my use shortens the potential life of the battery, so be it. This makes the discussion of what's best for battery life practically redundant - but interesting all the same.
I agree with your view lemmy. The Panasonic battery comes with instructions regarding the ambient temperature ranges for charging only, so as long as we keep to these instructions the BMS should be able to keep control of matters to give the battery as long a life as possible.
There are other variables which are just as likely to affect a the battery life other than when to charge, the ambient temperature when the battery is discharging is known to affect e-bike battery performance, but does this drop in performance in freezing conditions also translate to a reduced battery life? Hitting a pot-hole is more likely to deplete the life of an e-bike battery quicker than the charging regime employed!
 

onmebike

Esteemed Pedelecer
Jan 3, 2010
499
1
West Essex
This is just an observation. After reading every bit of info I can find on the care of Lithium batteries including the battery university, I've come to this conclusion.
Nowhere can I find an in depth description of what is deemed to be a deep discharge? Battery university says discharging, anything less than 70% of the batteries total capacity cannot be regarded as a deep discharge.
It would therefore appear a deep discharge of an E-bike battery is impossible because of the cut off point imposed by the controller or battery management[approx 31volts for a 36volt battery] No where near 70% of capacity.
Most of the information state's frequent partial charging is better than a deep discharge followed by a full charge, mostly due to heat generation of the latter. But as said above it doesn't appear possible to deep discharge e-bike batteries, at least not by the battery university's definition of deep discharge.
Heat is apparently the lithium batteries biggest enemy. Batteries should be stored longterm 40% charged in a refrigerator. Which seems to fly in the face of many forum members experience that the cold kills battery capacity.
Another common claim on many of the article's read is that it makes no difference how lithium batteries are used because they do not suffer with memory effect like nickel based cells.
One thing that does stand out is that most article's written on the subject refer to smaller consumer item's such as mp3/4 players, mobile phones, camera's and laptops etc.
E-bike batteries seem to vary in construction, some using many smaller cell's others less larger one's. Not being privy to battery construction methods I can only guess that the basic construction for say a 36volt 10ah battery is something like 10 x 3.6volt 10ah packs in series, each pack consisting of maybe 5 x 3.6volt 2ah cells in parallel, which is a total of 50 cell's.
For true balance charging and battery management each and every cell should be monitored but I doubt they are because the circuit required would be too elaborate so I guess that just the 10 main packs are crudely monitored for balancing purposes and the charger uses delta peak detection to assess when the battery has reached capacity.
Conclusion, Read the manufacturers guide, use the supplied charger and hang on to the guarantee if your lucky enough to have one. If all else fails you still have the pedals.
 

NRG

Esteemed Pedelecer
Oct 6, 2009
2,593
10
The DOD will depend on the BMS and how its setup. There are some very sophisticated ones available but I suspect there are also some much simpler ones that don't perform all the necessary cell monitoring functions, balancing and protection that's needed. As for the term 'deep discharge' I take that to mean anything below 90% or to a point where cell damage is of concern.

Take this graph of a gen3 Lipo cell, under load when the voltage hits 3.2v the cell is all but done....the voltage is past the knee of the discharge curve and taking it further is bad for the cell. A good BMS will at this point shut the cell/pack off...others may allow it to fall to 3.0v IE empty, which is a risk as this is the point where cell damage can occur especially if there is no individual cell monitoring and you have a weaker cell in a series that may be brought down below 3v...

http://static.rcgroups.com/forums/attachments/4/4/9/7/4/a705206-102-LVC example.gif?d=1137597622

There's no 'buffer' for the weaker cells with the LVC set to kick in at 'empty'

So yes, an E-Bike battery can be deep discharged and I'm wondering if some of the early life failures reported before are due to poor BMS implementation and/or poor cell quality.
 
Last edited:

flecc

Member
Oct 25, 2006
46,532
23,366
Heat is apparently the lithium batteries biggest enemy. Batteries should be stored longterm 40% charged in a refrigerator. Which seems to fly in the face of many forum members experience that the cold kills battery capacity.
The loss of capacity due to cold is only a temporary effect while the cold lasts, the capacity returning with returning warmth. I rarely use my laptop battery so it lives in the fridge at about 5 degrees C and at over five years old it's still fine when I warm it up for the odd trip away.

Batteries are chemical devices of course, and most chemical reactions require heat, the cold suspending or slowing reactions and therefore slowing ageing.
.
 

Mussels

Esteemed Pedelecer
Jun 17, 2008
3,208
8
Crowborough
But in the end, the machine is my servant and if my use shortens the potential life of the battery, so be it. This makes the discussion of what's best for battery life practically redundant - but interesting all the same.
It may be redundant for you but my commute uses half the battery's capacity each way so I have the choice of half or full charges, if one or the other increases the life of the battery considerably then it is important. I have charged my battery about 600 times so far, that could have been 300 without inconveniencing me at all.
 

flecc

Member
Oct 25, 2006
46,532
23,366
So yes, an E-Bike battery can be deep discharged and I'm wondering if some of the early life failures reported before are due to poor BMS implementation and/or poor cell quality.
The BMS I've seen on nominal 37 volt batteries have mostly cut off at around 32 volts, so leaving a small cushion, though not enough to protect a failing cell of course.

However, the failures of the earlier generation lithium batteries were almost all due to the inadequacy of the manganese cathodes which had hurriedly replaced the fire causing cobalt based ones. Those early manganese variants simply couldn't deliver current well enough and were near the low voltage borderline from new when subjected to high loads. Therefore only a little ageing brought them into the cut-off point under maximum load.

This applied to the batteries that both Harry and I have mentioned above as failing in only months. Meanwhile, others riding the same bikes and stressing the batteries less in flatter areas were getting far longer lives, Russ our administrator one of them, showing that failures were not due to individual cell intrinsic design or manufacture.

Following development, later manganese cathodes greatly reduced this problem and today's compound cathodes are infinitely better at maintaining an adequate voltage floor under load. One day, with the iron based cathodes of LiFePO4 more fully developed, attention will increasingly turn to replacing the carbon anodes which could be bettered. Plenty to look forward to.
.
 
Last edited:

Wisper Bikes

Trade Member
Apr 11, 2007
5,803
1,781
65
Sevenoaks Kent
If David from Wisper is monitoring this thread, or indeed anyone else with a close relationship with a battery manufacturer, I feel it'd be interesting to get the opinions from the lab on this issue. As an end user, it's been a nagging doubt, and blind spot in my knowledge since taking ownership of my ebike.
I put exactly this question to PP and AE last month, they were both of the opinion that it makes very little difference whether the batteries are fully discharged or topped up frequently. The range achievable from the battery during it's useful life will be about the same.

All the best

David
 

lemmy

Esteemed Pedelecer
It may be redundant for you but my commute uses half the battery's capacity each way so I have the choice of half or full charges, if one or the other increases the life of the battery considerably then it is important. I have charged my battery about 600 times so far, that could have been 300 without inconveniencing me at all.
My point is that since there is no solid irrefutable evidence, even from the manufacturers, of what is best for the battery, the best bet is to do what you want.

In your case, charge at work or not, whatever suits you best, since the effect on the battery is a matter for conjecture.