Best Charging Strategy for LiFePo4

[adrian_sm]

What is the best charging strategy to keep my battery healthy?

I have a 48V10Ah Headway pack so about 500Wh worth. My one way daily commute uses around 180Wh.
At the moment I am charging at both ends, so doing a 40% DoD, charging immediately and staying on the charger until I ride back the other way.

The theory was this means I always have a full charge if I need it, but I very rarely take any detours. I was also under the impression that shallow discharges were better for the battery, but I have no idea if that is true.

Is this the best charging strategy to look after my battery? Or would I be better of doing a 80% DoD and just charge once a day, and rack up less battery cycles.

 

[mazott]

Using a 48V/10Ah Ping's battery, I've done in about 1,5 years 400 charges at 47% Dod in average. Almost never 100%.
Any discharge is refilled right after with a charge. While charging, the energy refilled is always about the same amount discharged, plus 10%. Cell balancing is achieved within 2-3 hours after full charge, and it happens once every 10 charges made. Healthy, so far.

 

[Gregte]

Everything I have read about lithium batteries says that a shallow discharge is better. This is of course also true of lead/acid. It is just the opposite with NiCd and NiMH.

 

[Ypedal]

energy in , energy out, personally i prefer to charge when possible and try not to drain deep cycles .

if the pack is hot/warm after a ride, i'd let it cool off before hooking up the charger.

Making a habit of shallow cycles is a good thing as the pack ages and one or more cells start to have less capacity, they will keep working.. if you regularly drain a pack 100% expect the weakest cell to die making the whole pack a doorstop.

In a perfect world all cells would age equally and start to fizz out at the same time.. in reality this has not been the case in general.... you tend to have one or more weak cells in a pack being the limiting factor of the whole pack.

 

[monster]

the exception might be if you are going to leave it for a long time then you might want to leave it at 50% DOD because this is better for long term storage.

 

[dnmun]

the argument about needing to totally discharge nicad or nimh is false. you can recharge them at any time with no damage. the memory effect can be allieviated if it is a concern by occasionally drawing the pack down close to voltages below 1V but that is usually done with a very low current discharge at those voltage levels to allow the crystals to dissolve.

the 1/2 charged storage i have heard is recommended for nicads, nimh has other high self discharge issues too, but i did not know it was recommended to store lifepo4 at 1/2 capacity.

remember that lifepo4 packs are different from electrochemical cells because they store the ions intercalated in the lifepo4 crystal which composes the cathodic material.

so state of charge should not effect it imo, but i have no other knowledge of that.

i also think that running your motor on charge stored at the very top of the crystal is easiest for the battery to handle over time which is how they get the high cycle numbers, because the ions do not have to pulled out of the crystal as much and the lithium growing on the anode doesn't structurally weaken the anode surface interface by crumbling off as the lithium pushes it around while falling out of solution and precitpitating around the anode.

hope that makes sense.

 

[Gregte]

the argument about needing to totally discharge nicad or nimh is false. you can recharge them at any time with no damage. the memory effect can be allieviated if it is a concern by occasionally drawing the pack down close to voltages below 1V but that is usually done with a very low current discharge at those voltage levels to allow the crystals to dissolve..

I too have heard this argument and agree that occasional deep discharges are all that is required for best longevity. However, a deep discharge every time will not hurt them as it will lithium and lead/acid. It is just not necessary.

You don't have to worry about true 'memory effect' but rather 'voltage depression', which has similar symptoms is the problem created by seldom or never deep discharging NiCd or NiMH.

 

[dogman dan]

I'm no sceintist, but the conventional wisdom seems to be you get more cycles out of lighter discharges. In other words, your theoretical 1000 cycles or whatever it is, is not used up by 1000 light cycles. 40% discharge would get you 2000 or more cycles, theoretically. It would be interesting to know which strategy pulls the maximum lifespan watthours out of a battery. Should be testable with about 100 battery packs and about 3 years of cycling them. :lol:

Oh well, I guess I wont be doing that test. I wouldn't worry about any cycles less than 80%, so if it is more convenient to just charge at home, do that. It can be nice though, to have the full charge at the end of the day, for the scenic route home on a beautiful evening. But I wouldn't sweat it much if the battery easily goes both trips on one charge.

 

[Ypedal]

With Nimh, nicad and LiMn a full charge means higher operating voltage.. LiFePo4 kinda fixes that issue with a very flat discharge curve, specially good on hub motors !!! that's one of the things a geared system can work around, as the voltage sags, flip a gear ratio..

But either way, imo, best way to make a lithium pack last as long as possible is to never allow it to drain past the lvc limit per cell ( for lfp 2.0v per cell aparently.. i stop at 2.3 as an absolute, only a few mah left at this point anyways. ) and with that in mind, keeping them fully charged and ready for use gets my vote, got packs going on 4 years !

 

[Gregte]

But the original question of the OP is;

is it better for the longevity of a 48v10Ah Headway Pack (LiFePO4) to charge it twice a day after it has lost 40% of its charge each time

or to let it lose 80% of its charge and only charge it one time per day?

In other words, is it worth the bother to charge it twice or will it last just as long by charging only once?

I believe the question is inspired due to the fact that some battery types will last longer doing it one way and other types will last longer doing it the other way, and perhaps some types don't care.

 

[Ypedal]

I'd be more concerned about the charger than the battery pack ( bumping around on a bike is hard on chargers ) get a 2nd charger so you can have one at each end and charge each time imo.

More shallow cycles are better than fewer deep cycles.

 

[adrian_sm]

Thanks everyone. Your right the original post was more about is shallow discharges better than deeper for an equivalent power used.

I bought some cheap chargers recently so I just leave one at each end.

It is all probably a moot point. Hopefully it will last so long that the reason I change is a new battery technology with higher density, higher output and cheaper.

I should go find that post that had a link to some research explaining the differrent battery technologies and see if I can get a clear answer at a chemical level.

 

[dogman dan]

Eh, stop worrying and ride with a big EV grin . Like Ypedal says, as long as you aren't riding to cutoff every cycle, yer good. Not that a good hunt for something isn't fun, too, but for sure if you are fretting, stop it and enjoy the bike. Ride to cutoff if there is a good reason to, and charge at both ends if it's easy to. Having two chargers is the way to go, I beat up two of em carrying them daily. Stuff would just shake apart inside and then either short out or simply not make contact.

 

[dak664]

A definitive answer would be good to have and could collectively save us a lot of money. Google hasn't found for me any good FAQ on the subject. Some factors seem to be:

1) Lithium battery life is more a function of time then number of cycles
2) Aging is faster, the higher the temperature.
3) Aging is faster, the higher the state of charge.
4) Aging is faster with larger *differences* between min and max state of charge.
5) Aging is rapid when maximum depth of discharge is used.
6) At present marketing favors high capacity ratings over long lifetimes. Military batteries charge to 0.1 volts less and avoid maximum DOD to increase lifetime at the expense of capacity.

So it appears for maximum lifetime you delay charging until just before use and then charge to just the needed capacity at some compromise rate that limits heating.

It would be great to have a three dimensional set (C,T,D) of three dimensional plots (x,y,t) characterized by C=charge/discharge rate, T=temperature, and D=time between cycles, with t=lifetime and x,y=min,max charge level. Longest lifetime should be around x=y=0.5, (half charge with the charger permanently connected to supply the load current), and shortest at x=0, y=1 (frequent deep cycle). The ridge of longest lifetime gradient connecting these points would indicate the best parameters for each particular (C,T,D). A really smart battery charger could know all this so when you connect the battery you would tell it you need 7 amp-hours tomorrow at 8 am, and it would delay charging as long as possible with a charge rate based on ambient temperature.

 

[dogman dan]

That's nice, but in the real world, you're gonna just keep it charged and ride it do a depth of discharge that depends on how far work is from home, or how long the fun trail is. Just ride it and grin. If I get close to 5000 miles out of my ping v1 while leaving on a charger 24-7 and storing it in a garage, while charging, at 120F, and get away with it, we have little to worry about. I've murdered nicads in drills treating them like I treat my ping.

 

[adrian_sm]

Yeah agreed. Ride it. Enjoy.

It was more from interest that I started the tread.

Since I have a charger at each end, it is easy for me to just keep topping up and always have a full tank. But since the battery is the biggest running cost, I thought I should find out if this was a bad idea. But I couldn't find a decent answer from the usual google search.

But knowing my usual buying habits, I will probably end up upgrade before it dies.

 

[jag]

A definitive answer would be good to have and could collectively save us a lot of money. Google hasn't found for me any good FAQ on the subject. Some factors seem to be:

1) Lithium battery life is more a function of time then number of cycles
2) Aging is faster, the higher the temperature.
3) Aging is faster, the higher the state of charge.
4) Aging is faster with larger *differences* between min and max state of charge.
5) Aging is rapid when maximum depth of discharge is used.
6) At present marketing favors high capacity ratings over long lifetimes. Military batteries charge to 0.1 volts less and avoid maximum DOD to increase lifetime at the expense of capacity.

These all agree with conventional wisdom. Good actual test data seems hard to come by though. I suspect larger companies may have done tests for their internal use. Some anecdotal facts:
Some Thinkpad laptops allows you to choose battery charge/maintenance between maximum capacity and maximum battery life. The latter keeps the battery at a lower charge state.
Prius runs the battery across some relatively narrow band, avoiding both full discharge and full charge.

 

[adrian_sm]

Prius runs the battery across some relatively narrow band, avoiding both full discharge and full charge.

Avoiding a full charge is interesting. I vaguely remember someone, maybe Docbass, saying that a reduced charge voltage resulted in an increased life of x%. But I can't find the post for the life of me.

 

[dnmun]

the prius uses nimh batteries. this is lifepo4 you are talking about i thought.

nimh is a chemical cell, lifepo4 is an ion storage cell where the ions are stored in the lifepo4 polyanionic crystalline matrix of the cathodic granules.

i would recommend you always keep your pack fully charged, but not over charged. everyone's favorite number is 3.65V.

i would also recommend you avoid using the last 10%-15% of the battery on a regular basis where the voltage begins dropping rapidly. jmho.

 

[adrian_sm]

I am pretty sure the post I was refering to was talking about konions or A123, can't remember, but not NiMh.

What is the shelf life of LiFePo4 like. The charging/discharging might not be as big a factor as the slow inevitable change over time.

Found this link to A123 specs.
http://media.hyperion.hk/dn/a123/A123.pdf

Current test projecting excellent calendar life: 17% impedance growth and 23% capacity loss in 15 years at 100 % SOC, 60 °C

Also shows capacity losses for 100% DOD cycles of a single A123 cell
- 5% loss @ 25 deg C, 2.3A charge, 2.3A discharge
- 11% loss @ 45 deg C, 3 A charge, 5 A discharge
- 24% loss @ 45 deg C, 3 A charge, 5 A discharge

If I assume this is typical of LiFePo4 in general (bit of a stretch I know), then temperature and discharge rates are the biggest factor. Both of which I am not going to change in a hurry.

With researchers doing 10x better energy density and discharge rate batteries in the lab, and the look of this data, I'll be buying a better battery technology way before my batteries die.

 

[dogman dan]

I suspect a lower state of charge may extend lifespan by avoiding accidental overcharging, even if the chemistry is very different from others. I wouldn't sweat it if your charger shuts off at the correct voltage. It would only be a problem if you were charging manualy with a power supply that doesn't shut off .

 

[kZs0lt]

Prius runs the battery across some relatively narrow band, avoiding both full discharge and full charge.

Avoiding a full charge is interesting. I vaguely remember someone, maybe Docbass, saying that a reduced charge voltage resulted in an increased life of x%. But I can't find the post for the life of me.

I think you are referring to doc's post here: http://www.endless-sphere.com/forums/viewtopic.php?f=16&t=9268&start=0
It is about lithium-ion or polymer.
According to that post, I think a narrow band (avoiding full charge and deep discharge) does help prolonging the life of the cell, at least the calendar life. If we are speaking about how to get the most kw/h out of the pack over it's lifespan, there is probably a sweet spot of max. charging voltage, max. state of discharge, current limit, and temperature.