-- Battery Shelf Life --

[safe]

Battery Shelf Life

http://en.wikipedia.org/wiki/Lithium_ion

For Lithium based cells this is the estimated Shelf Life:

 

[safe]

This is a chart I made by applying the figures from above. This could tell you what to expect if you store your battery at typical 25 C degree temperatures. (77 degrees F) In summer you might age a little faster and winter a little slower, but it's actually the level of charge that's going to make a big difference. I didn't know that... so it's a good idea to charge up in the morning, go for a ride, then just leave the bike uncharged overnight. (don't let the charge go too low though)

The numbers across the bottom are YEARS and the left is CAPACITY remaining.

 

[Toorbough ULL-Zeveigh]

...so it's a good idea to charge up in the morning, go for a ride, then just leave the bike uncharged overnight. (don't let the charge go too low though)

That sux, always trying to find the Goldilocks zone;
Not to high, not too low, but juuust rite.
It seems like lithium batteries require almost as much maintenance as a woman.

 

[safe]

It seems like lithium batteries require almost as much maintenance as a woman.

They last about as long as an average American marriage. :lol:

 

[safe]

The Incredible Shrinking Battery

Like all the other chemistries the capacity matches the C rating over time, so as the overall capacity tends to go into decline the ability to pull amps also falls.

So if you design your bike so that it NEEDS to pull 40 amps (for example) and you buy just enough battery to barely pull that when new then within a short period of time your bike becomes really lame because it's reduced capacity cannot live up to the job it's designed for.

The smarter idea is to get more battery and design to use it less... far less... than it's initial capacities. That way you could have a battery that is five years old and down to maybe 40% capacity, but since you've left such a big overlap in your design it's still pulling like day one. It might lose it's range.... but it still would pull strong.

So this is stuff people ought to think about when they are dropping down big $$$ on Lithium... you might actually be better off with a little more battery than a little less, or you might be better using less power.

 

[safe]

Thundersky 3C 40Ah Shelf Life vs Peak Amps

So let's say you buy some Thundersky cells ($80 each from Patrick) that can at best pull 3C and they start with a 40Ah capacity and you want to be able to pull 40 amps, then how long will you be able to use the batteries?

It depends on how you store them (and cycle counts would add too, but my guess is that shelf life is about as serious a problem as cycle count) but it looks like I might expect a usable life of about five years, maybe a little longer if I make it an issue to leave the cells at a low charged state all the time.

Five years of ownership
Initial Cost - $1000 (12 cells)
Cost per Year - $200
Range - About 30 miles per ride
Typical Riding Per Year - 4000 miles (130 rides?)
So in the end you pay $1 for 20 miles or about the equivalent of 1$ for a gallon of gasoline.

So it's still cheaper than gas.

 

[fechter]

Is that relationship valid for all lithium chemistries? I assume they're talking about Co based cells. It may be entirely different (or not) for LiFePO4 or LiMn cells.

 

[safe]

Is that relationship valid for all lithium chemistries? I assume they're talking about Co based cells. It may be entirely different (or not) for LiFePO4 or LiMn cells.

I've read somewhere that it's really all about voltage, the higher the voltage the cell holds the faster it degrades. A123 have found a way around the problem by creating a cell that keeps the voltage low all the time and the result is longer life. So a more accurate way would be voltage.

Overall you want to keep the voltage at a certain ideal level and not too low or not too high because either way that increases degradation.

I'm looking forward to a time when a 10 year shelf life is easy to achieve, then you can afford to spend $1000 without feeling too bad about it. If I managed the Thundersky cells "perfectly" then it might be possible to get to 10 years for a 40 amp power demand, but if I mismanage them at all I'd expect much less. I've also read that some are having troubles with the cells actually rotting around the terminals. I hope that's been fixed.

 

[safe]

Chevrolet Volt

http://en.wikipedia.org/wiki/Chevrolet_Volt

GM plans to keep the lithium-ion battery in a state-of-charge (SOC) range of between 30% and 80%, with the on-board generator starting to recharge the battery at the 30% level.
...
General Motors said it will have the Volt on the market in 2010. To help spur battery research, GM selected two companies to provide advanced lithium-ion batterypacks: Compact Power, which would use cells made by its parent company, LG Chemical, and Continental Automotive Systems, which would use cells made by A123Systems. However, on August 9, 2007 GM established a more close-knit relationship with A123Systems so that the two companies could co-develop a Volt-specific battery cell. This cell was later unveiled at the EVS23 industry convention in Anaheim, CA. Work with CPI has continued at a rapid pace, and in late 2007 CPI delivered two fully-functional prototype battery packs to GM's testing facilities. On January 31, 2008, A123 and Continental delivered their first prototype to GM's European test facilities. GM will likely use both suppliers for the Volt, although this remains a matter of speculation.

This might be something to consider... maybe you NEVER want to fill the battery to full? Maybe you are better off getting a large enough LiFePO4 battery and then do like the Volt and keep it between 30% and 80%.

 

[safe]

Early Charge Cutoff?

Why not just shut off the charging the moment that the cell voltage reaches 3.8 volts (or less)?

(the published full charge is 4.25 volts for Thundersky cells)

After all... a truly "full" cell is bad for it.

Doing this (max 80% charge) alone would extend the shelf life by two years, but it will obviously shorten your range by 20%.

 

[OneEye]

The degradation at a high charge state is due to oxidation of the free lithium. From what I've read, the LiFePO4 chemistry is not supposed to have the self-oxidation problem that LiCo, LiMn and other Lithium Ion chemistries have. Supposedly it will have a 5-10 year shelf life/service life regardless of state of charge. If true that would be awesome. I hate that all my ancient laptops (P3/600) have no battery capacity despite not being run through many charge/discharge cycles. The batteries aren't worth re-buying or re-building as they won't see enough cycles (ever) to justify the cost. They only rarely go mobile.

 

[safe]

Supposedly it will have a 5-10 year shelf life/service life regardless of state of charge. If true that would be awesome.

But think of the counter evidence...

If the Chevrolet Volt has plans to do a 30% low and 80% high pack range then they are likely onto something. And they are also planning to use A123 cells. So unless there has been a change in philosophy (which is possible) I'd tend to think that higher voltage always makes things worse. Even the older Lithium types are okay if kept at low charge and low temperature for long periods.

If you simply take an ordinary older Lithium cell and keep it at 25C / 77F and keep the charge to 50% you can get ten years if you are willing to go down to 30% capacity.

A123 seems to have found a way to NEVER raise their voltage above 3.6v and that might be their secret. Keeping the charge low will not hurt... that's pretty certain...

 

[OneEye]

I'm not sure, but should your decay curves be logarithmic? Also, I thought the battery industry considered 80% of original capacity the cell/battery end-of-life. Individual users can decide for themselves if it is even worth it to carry around something that is only 30% of its original effectiveness.

Obviously anything on the Chevy Volt is somewhat speculative, but I thought the reason they intend to keep the SOC between 30-80% is it extends the cycle life (as opposed to the shelf life) of the cells. Dicharging and recharging alter the cathode and anode at the microscopic level, leading to cell degredation; deep discharges/full recharges do so to a greater extent. These changes are a different phenomenon than the propensity of most Lithium chemistries to self-degrade simply in a storage (0-cycle) mode.

Clearly LiFePO4 is new technology, and unfortunately there is not enough authoritative literature and independent test results available to the public to validate its operating parameters. If it lifes up to half of its claims I will be impressed.

 

[safe]

Obviously anything on the Chevy Volt is somewhat speculative, but I thought the reason they intend to keep the SOC between 30-80% is it extends the cycle life (as opposed to the shelf life) of the cells. Dicharging and recharging alter the cathode and anode at the microscopic level, leading to cell degredation; deep discharges/full recharges do so to a greater extent. These changes are a different phenomenon than the propensity of most Lithium chemistries to self-degrade simply in a storage (0-cycle) mode.

Should it matter?

If the Chevy Volt people think this adds value then I would tend to want to follow their lead. It could be a little of both that the higher voltage causes damage during rapid cycling and it also degrades shelf life. I have read that voltage is the key culprit, but it was not from a reputable source. It would be nice to find someone from inside the Chevy Volt team to comment on it. Know any corporate spies?

 

[OneEye]

Should it matter?

I would suggest it matters a great deal. As you point out, in many Lithium Ion chemistries self oxidation at high state of charge cuts your range by 20% per year regardless of usage. For many applications, this 0-cycle degradation will have a greater impact than the per-cycle induced degradation.

This hasn't always been the case, Lead-acid has a long-shelf life / short cycle life relatively speaking, which has made it easy to wear out the batteries faster than they degrade due to shelf-life issues. As the cycle life of lithium batteries has increased the shelf life has become more significant.

If the LiFePO4 chemistry is (relatively) shelf-stable at a high state-of-charge, the price-weight-range-cycle life-shelf life optimization decision e-bike builders/buyers/users have to make changes dramatically.

 

[safe]

If the LiFePO4 chemistry is (not) (relatively) shelf-stable at a high state-of-charge, the price-weight-range-cycle life-shelf life optimization decision e-bike builders/buyers/users have to make changes dramatically.

Yes is does.

That's why this thread will either:

Awaken people to a stunning new reality.

Make people laugh because we thought something that was never true.

...until we get a definitive answer we just don't know.

 

[NickF23]

If the Chevrolet Volt has plans to do a 30% low and 80% high pack range then they are likely onto something. And they are also planning to use A123 cells. So unless there has been a change in philosophy (which is possible) I'd tend to think that higher voltage always makes things worse. Even the older Lithium types are okay if kept at low charge and low temperature for long periods.

Why do you think this is to do with calendar life issues and not cycle life? Toyota used a similar charge/discharge protocol for their nimh prius packs. My understanding is that their reason was to do with maximising the cycle life of the battery, specifically the 2000 cycles of a good qulaity nimh battery would not be good enough for a pack that was warranteed for 8 years. Nimh aren't known to have a specific 'calenadar life' issue as lithium consumer cells are.

If we presume lifepo4 cells give 2000 cycles this might not be good enough for 40 mile range Phev like the volt which could easily do 600 cycles a year in normal use charing twice per day to do 80 miles. This isn't good enough if for marketing reasons GM wants the battery pack to last the life of the car.

Valance, saft and many other large battery companies claim to have 10 year plus shelf life their lithium traction/automotive batteries. Intersestingly the 'shelf life' problem may not even apply to Thundersky lithium cobalt packs, some of the data collected such as from http://www.solarvan.co.uk/ on the Thundersky list shows rising resistance but no drop in capacity after 2 years. Perhaps its only a problem of cheap consumer lithium cells where long life isn't considered commercially important?

 

[safe]

http://www.solarvan.co.uk/why.htm

"This picture from the 2003 Lithium-Ion UK battery conference shows myself and the other enthusiasts, who participated in the first UK bulk purchase of Thunder Sky cells."

My understanding is that the newer Thundersky's are supposed to be better than the first one's they made back in 2003. Patrick would know more about it. He's said that he's had no problems with his batch which I think they got in 2007. Plus the LFP cells are LiFePO4, so it's a totally new chemistry for them.

 

[safe]

Using Logic

If the act of raising a cell to a high voltage during cycling causes degradation and also when it drops too low it causes degradation while cycling then it only stands to reason that sustained high voltage would degrade as well as sustained low voltage. Think about it... the longer a cell is in a low voltage situation the more degradation it sustains... why would a behavior be asymetrical and not act the same way for high voltage?

Higher voltage and higher temperature is "without doubt" a negative factor with pre-existing Lithium cells. We have no debate about this part and can assume it's a fact. But it's a fact with a different chemistry.

The argument would have to be something along the lines of a RATE OF CHANGE within the cell causes damage and it's not the tension inside the cell that does it. An argument for long shelf life would use the math concept of the "first derivative" so that a constant voltage gave a zero result, but an increasing voltage would give a linear result.

http://en.wikipedia.org/wiki/Derivative

Got that?

In order for a constant high voltage to NOT do damage and yet cycling DOES do damage you need a circumstance where it's the rate of change that does the degradation.

The "C" rate would have to trump the shelf life rate.

I have heard that you can destroy a cell in a hurry if you OVERCHARGE it, so obviously damage does happen up there.

 

[safe]

An Inconvenient Truth?

It might simply be a truth that is awkward for people to deal with. That ideally you want to let your LiFePO4 cells sit at 50% charge when not being used and only charged to full moments before you use them. It makes charging a more complicated issue because it becomes something that has two stages to it. My guess is that short periods of full charge do only as much damage as the time that the cell remains at that charge just as low voltage damage is measured in time there.

We all need chargers with a LOW and HIGH voltage setting to represent STORAGE or RIDE mode?

 

[safe]

Simple Guidelines

http://www.batteryuniversity.com/parttwo-34.htm

Avoid frequent full discharges because this puts additional strain on the battery. Several partial discharges with frequent recharges are better for lithium-ion than one deep one. Recharging a partially charged lithium-ion does not cause harm because there is no memory. (In this respect, lithium-ion differs from nickel-based batteries.) Short battery life in a laptop is mainly cause by heat rather than charge / discharge patterns.

Batteries with fuel gauge (laptops) should be calibrated by applying a deliberate full discharge once every 30 charges. Running the pack down in the equipment does this. If ignored, the fuel gauge will become increasingly less accurate and in some cases cut off the device prematurely.

Keep the lithium-ion battery cool. Avoid a hot car. For prolonged storage, keep the battery at a 40% charge level.

Consider removing the battery from a laptop when running on fixed power. (Some laptop manufacturers are concerned about dust and moisture accumulating inside the battery casing.)

Avoid purchasing spare lithium-ion batteries for later use. Observe manufacturing dates. Do not buy old stock, even if sold at clearance prices.

If you have a spare lithium-ion battery, use one to the fullest and keep the other cool by placing it in the refrigerator. Do not freeze the battery. For best results, store the battery at 40% state-of-charge.

_________________________
Created: February 2003, Last edited: September 2006

Figure 2: Effects on cycle life at different float charge levels (Choi et al., 2002)
Higher charge voltages boost capacity but lower cycle life.

 

[OneEye]

It raises an interesting question, for (insert battery here) does it matter which 50% or 60% of the stored charge is used for cycle life?

Is there a difference between utilizing the energy between the 100% and 50% SOC and recharging to 100% vs using the energy between the 80% and 30% SOC and recharging back to 80%? I'm guessing that using the energy stored between the 50% and 0% SOC and recharging back to 50% is the most risky and probably the most damaging, but for various cells what are the impacts. Safe's chart from battery university seems to suggest the top end of the range is dangerous, but this may be partly due to the oxidation potential in high SOC states, and also partly because you are utilizing more than 50% of the theoretical charge when going from 100% SOC to (say) 30% rather than moving from 50% SOC down to 30%.

For lead-acid, I think the charts suggested you could get the most energy in and out of a battery by utilizing the 100% to 50% range, using less than that meant you got a few more cycles from each battery, but less overall energy.

 

[safe]

Lead Acid is easy... just keep the battery 100% full all the time and never drain it much and it lasts the longest. This isn't helpful in getting much out of the battery and the Peukert's effect makes things even worse.

Lithium cells seem to prefer a 30% to 80% SOC for long life and minimal cycle degradation too. They seem to fit together, the wider the swings within a cycle the longer you spend time in the extremes. We know that both extremes are bad for Lithium.

So you end up with:

Best Storage:
Charge should be at equilibrium in the middle.

Worst Losses:
Either far below the middle or far above the middle... imbalanced voltage and state of charge levels.

Lowest Cycle Losses:
Cycle within a tight spread above and below the equilibrium point in the middle state of charge.

The Chevy Volt is using numbers like 30% to 80% as "ideal" and my guess is that they used some computer analysis to project that as the most cost effective range to be in. However, changes in technology can change those range data points, so ideally you would want a charger that was ADJUSTABLE so that you could adapt to different possibilities like this one does:

http://electricmotorsport.com/store/ems_ev_parts_chargers_zivan.php

Got an extra $500 laying around burning a hole in your pocket?

(actually for what you are getting it's a very good deal)

 

[PJD]

Well...I'm ready to buy an expensive LiFePO4 pack, and I assumed this shelf life problem didn't apply to them.

This complicates battery management! With lead acid, they were charged upon arriving at work, then upon arriving at home, and left on float overnight for commuting and errands first thing in the morning.

But with lifepo4's Now, here's the scenario...

1 get home from work at 6:30 pm - pack discharged 90%.

2. Plug in the charger, charge for an hour or so, get up from supper or other errands to check the voltage, charge for another hour - check again - 50% charge - done for now.

3. Leave them partly discharged overnight, then get up at 4:00 AM to plug the charger into the scooter so it will be fully chrged for the next days commute; get back in bed, then get back up for the 8:00 commute.

4. Repeat.

Not convenient at all!

 

[EMF]

I don't think the LifePO4 have a shelf life problem. Do some searches on Google and see for yourself. LifePO4 uses shelf life as a selling point. Safe found a table somewhere on a "lithium" cell ( I guess-the table does not mention the chemistrry unless I am blind) and is dumping them all in one catagory. If I had a battery like the one he pasted the chart on in this thread - I would think the best strategy would to put the damn things in the refrigerator. Because as the ambient temp rose the batteries suffered on the shelf.

But, I don't worry about such things and I would not even bother with keeping them cool or fretting over the charge status and just kick myself in tha ass for buying a cheap Lithium cell pack. Luckily, I have A123 cells so this "data" does not apply.