So, I want to talk about this, but want everyone reading this thread to understand that I absolutely support the use of LiIon batteries. Also, note that I believe that the Chevy Volt is the Holy Grail automobile that will save GM, America's economy, bolster our national security, and help the environment. Those are pretty strong words coming from a RAV4 EV driver. It really is that good.

That being said, here comes the dirty little secret:

LiIon begins to degrade around the three year mark.

There doesn't seem to be much that anybody can do about it. It seems to be unrelated to charge cycles, state of charge, temperatures, charge rates, etc, and is only related to the calendar. NiMH, on the other hand, can last a decade or more, but degrades with high temps, with low SOCs, and with too many charge cycles.

What are the implications of this dirty little secret?

1. The Volt will cheat and lie (and I'm okay with it!). The Volt will start out with a 40 mile range. This is a lie. It's really more like 80 if the full SOC were used, which it won't be. The Volt will then cheat after the three year mark, and begin to expand the use of the SOC as the battery begins to degrade, thereby maintaining the 40 mile EV range later in the battery's lifespan.

2. During this battery's lifespan, strict control over temperatures, SOC, and charge/discharge rates will be maintained to put off the degradation as long as possible. This is a good thing, but may result in some strange behaviors, like occasionally not charging as fast; the generator running at strange times, etc. It's all for the good of the pack, and I'm all for it.

3. Pack replacement is inevitable. (This is okay, too). I actually breathed a huge sigh of relief the other day when GM revealed that pack replacement cost was built into the Volt's price. I knew that what they had been promising was essentially impossible, a LiIon pack that lasts forever. Either they were lying or they didn't know what they were doing, both of which are terrifying, because they spelled doom for the project. GM's revelation has actually restored my trust, as it should yours. They know what I've always known; simply ramp up production, let the cost come down, and the viable, profitable EV will become commonplace, even if the battery needs to be replaced at some point.

4. Toyota's love affair with NiMH is justified (somewhat). As long as Toyota believes that a E-propulsion vehicle must have a permanent battery, then NiMH is the way to go. It's not Toyota's lack of understanding of the LiIon that keeps them on the NiMH, it's the corporate culture mentality of perfection that keeps Toyota on NiMH.

5. Maybe pure EVs (with no Generator) should be built with NiMH until a better battery comes along. LiIon might get better, very very soon, but we need more plug-ins and we need them now. Yes, NiMH batteries cost a lot for the nickel, but the nickel is still there when the batteries are junked. They are worth gobs of money in scrap value. Imagine an old, dead EV selling for $15,000 for its scrap value. It makes the higher selling price easier to stomach.

Your thoughts?

Nate

That being said, here comes the dirty little secret:

LiIon begins to degrade around the three year mark.


I'm operating under the assumption that GM will choose A123 batteries & announce this in sync with their IPO.

Batteries from A123 will be significantly better than what is currently on the market. See these slides, which explain this as a function of oxidation potential:

http://www.a123systems.com/#/technology/life/lchart4/
http://www.a123systems.com/#/technology/safety/schart3/

This chart shows that 80% of retained capacity is reached at 7000 cycles. Keep in mind that these are full DOD (100%) and the Volt will be only 50-60% per cycle. This works out to about 500k miles driven in full electric mode.

http://www.a123systems.com/#/technology/life/lchart1/

The 80% point is at about 1000 cycles at 60C, full DOD. I would anticipate that an internal water cooling system on the Volt will be keeping the batteries much closer to room temp throughout the life of the vehicle.

Interesting graphics, Joshua.

The 80% capacity after 7000 cycle figure is at 25C, but the figures are nevertheless encouraging. The next slide showing performance at 60C seems to suggest that these batteries could indeed last 10 years as intended (50% DOD) for the Volt, even in hot climates.

500k miles is probably optimistic even for mild climates, given the other stresses the battery must endure.

My garage spends much of the day at around 60C/140F, so this is very good news for me!

I have my doubts as to whether the battery cooling system will run while the vehicle is unoccupied.

Will GM will be able to announce anything at all that could materially affect the A123 IPO? I believe there are laws, or at the very least, SEC rules that might prevent an annoucement.

Guys, I don't see any timeline on those charts. What I see is inline with what I've stated before; that is, the calendar not withstanding, LiIon can be cycled over and over again with minimal issue. The real issue, and my point, is that LiIon seems to begin an inevitable degradation after three years, regardless of SOC, number of cycles, temps, etc. It's a calendar problem, and LiIon battery makers and testers tend to go with two year studies to avoid talking about the calendar problem.

I'm hoping somebody will come in and tell me I'm wrong.

Nate

@nater: I had to give you credit, you know what you are talking about. My biggest concern with the re-release of EV's is the control of battery temperature during discharging. My research with batteries has shown temperatures in excess of 140F when discharging the lower capacity Gel-Cel, lead acid batteries. High heat conditions result in loss of moisture from the Li-ion "paste". Overtime, this will cause the Li-Ion cells to lose a significant amount of charge capacity.

How do we (the EV community) prevent this? Simple, temperature guages that give instant feedback on the condition of the batteries. We can then turn on cooling fans for the batteries or stop a moment to allow them to cool down.

Finally, my biggest concern with Li-Ion cells is their tendency to die in laptops when fully discharged. I hope the onboard computer in the Volt will prevent this. :eek:

The real question is after 3 years how quickly does the battery degrade?

The expert I spoke to said it begins to drop off noticeably. His biggest concern was that all of the LiIon studies are two to three years, so even the dropoff rate is not well known. He said the degradation is linear over whatever the timeline is, meaning that it doesn't get worse at a faster and faster rate. However much you lose over the course of a month is about the same a year or so later. (This is good news; meaning it doesn't drop off all of the sudden). He thinks that the battery in the Volt is going to perform acceptably for about six or seven years; meaning that the EV range will start to be shortened after that. The good news is that replacement pack prices will drop before we reach that point, the pack is easily replaced (from an effort perspective), GM has fudged in an extra pack for each car, and, something not too many folks have considered, as long as battery efficiency can stay reasonable, the Volt can live on as a high MPG hybrid sedan, even if the battery is crap. Basically, the car can still be useful even with a worn out battery, as long as the pack functions efficiently, even if its overall capacity is severely diminished. It will be interesting to see how this plays out over the years. My RAV4 EV is now six years old and has 74,000 miles on the original NiMH pack. It still functions through the entire SOC without issue, and seems to have no noticeable loss in capacity. This isn't really possible with LiIon, but I'm okay with it, since LiIon is a more efficient chemistry (less wasted heat), is much more energy dense, and Lithium is readily available, whereas Nickel is becoming scarce. I welcome the changeover to LiIon, but will miss my old standby.

Nate

@Nater : The expert you talked to should look at what is happening on space applications. Companies like GS Yuasa or Saft can guarantee Li-ion technology for more than 10y with real time qualification periods over 5 years at space accounts (i.e. real calendar life tests over 5 to 7 years up to now). And the technology makes it actually without any severe change of behavior after 3y (why 3y? :rolleyes:).

@Bicster: My gut feeling: GM probably waits for A123 having finalized fund raising before announcing they go for Compact Power technology.

b456, I'll look into it as well. Do you have any whitepapers or other data I can look at? (Links?)

Nate

Finally, my biggest concern with Li-Ion cells is their tendency to die in laptops when fully discharged. I hope the onboard computer in the Volt will prevent this. :eek:

If you fully discharge a Li-Ion cell you will kill it. The protection circuit in the battery is supposed to prevent you from discharging it to the brink of death. With a laptop it can be difficult to accomplish that, because the manufacturer wants to maximize run time without much concern for longevity (beyond the warranty period). With an EV it should not be an issue. Manufacturers will build in enough of a safety margin. The 30-35% minimum SOC of the Volt is way beyond that margin.

@Nater: could find interesting data in proceedings from battery conferences (i have AABC07 or 08 in mind)

In fact, I talked with a Tesla rep who had brought one of their cars to a recent family engineering day and he said they expected about a 5 year lifespan, and that was not due to cycling. The cells just won't last longer than that.

However, that may not hold true for new formulations. Part of the problem with verifying that is the new formulations haven't been around long enough to even measure their 10 year lifespan. A123 has had products available since 2005, for example, so you're not going to see any studies of their commercial cells that are longer than about 2 years at this point, anyway. They claim "10 year + projected calendar life" but since there's no cells that old there isn't any way to actually verify how long they could last:
http://www.a123systems.com/#/technology/life/

I guess the real question is how long was A123's R&D phase before releasing the chemistry that could be in the volt.

Here is an interesting article from MIT's Technology Review about A123:

https://www.technologyreview.com/read_article.aspx?id=20570

Registration may be required, or use BugMeNot

From the article:


Though a lithium-ion laptop battery might survive 500 complete charge-and-discharge cycles before its capacity fades, no car owner wants to buy a new battery every 18 months. According to A123's projections, however, its batteries should be able to deliver more than 15 years' worth of daily charges. And in addition to being safer than other lithium-ion batteries, A123's operate at a lower temperature, which makes it simpler to pack hundreds of them together into a large battery pack, Gray says.

There is more volt-y goodness inside.

very nice article. We can only hope that their estimates on battery life are correct. If not at least GM is factoring the cost of another battery in just in case.

@DaveP: indeed, Tesla packs are made out of standard design (18650 cylindrical cells - for power tools i bet), so that life time should be limited as these batteries (like laptop batteries) do not aim at enjoying 10y life time.

However, optmised design (i'm talking about cell design not chemistry) for high end applications (space, HEV) are able of much better level of performance in term of life.

The DeWalt 36v series of tools were introduced in 2005. So there should be ample 3 year old 26650 nano-phosphates available for testing.

That is a good theory Nater - but it should not be difficult to get hold of some 3 year old used batteries for testing. The manufacturing date may be on the battery too.

There is always silver-zinc. Better power (40%), but up front costs are high. All silver and zinc is recovered however.

There is always silver-zinc. Better power (40%), but up front costs are high. All silver and zinc is recovered however.

Aren't those batteries not rechargeable? That would be a pretty significant drawback I think!

So what about the new SCiB batteries from Toshiba, they advertise 10+ years lifetime? And a very short recharge time, which I think would heat up the battery considerably:
http://gm-volt.com/forum/showthread.php?t=1514

Aren't those batteries not rechargeable? That would be a pretty significant drawback I think!

Not the ones I read about. Company is touting them as better than LION for hybrids. I'm not that familiar with the technology, so I can't really comment (at least intelligently) beyond parroting the what I read.

OK, it turns out that SAFT is claiming 20 years calendar life, and A123 is claiming 15 years and 7000 cycles at 100% charge-discharge (DOD)!

Folks, its getting better by the moment.

I dont think LiFePO4 (LFP) and LiMnO2 (LMS) naysayers have much more leg to stand on.

http://www.osti.gov/bridge/purl.cover.jsp;jsessionid=17A353CA48047C8B430CE5BF CA4E0A59?purl=/925388-1KEcuJ/

See page 18.

Based on this document, LG batteries are in all respects inferior to A123 batteries, except for cost.

Based on this document, LG batteries are in all respects inferior to A123 batteries, except for cost.

If you factor in the need to replace the LG batteries before the warranty expires, will the LG batteries be really cost effective versus A123?

@mohsen:
1/ claiming long calendar life based on questionable extrapolation of accelerated cycle life test and demonstrating 10y+ calendar life in real time testing are two different things.

2/ I don't see the LG/A123 comparison in your (interesting) document. Where is it mentioned?

One of the reasonable explanations to the calendar clock based degradation that limits the life of these batteries could be due to the very slow but sure Brownian random motion diffusion. Through time, even with a barrier, diffusion can occur, and it is significantly based on time more than anything else. Some electronic circuitry are also victimized by diffusion through time. With the slow diffusion through time, things get mixed up rather than separated and it degrades the performance of batteries.

It depends on the nature of the alloys and the chemicals used to assemble the batteries. Some diffusion are faster than others. Alloys and metals that are bonded together in solar panels have degradation through time and it is related to diffusion, but the rate of degradation is very slow, accounting to only about half a percent decline of original performance each year.

@mohsen:
1/ claiming long calendar life based on questionable extrapolation of accelerated cycle life test and demonstrating 10y+ calendar life in real time testing are two different things.

The DeWalt came out in 2005 - therefore A123 must have evaluation units in 2004 or 2003. That is almost 5 years of accelerated cycle/calendar life testing. I believe the extrapolation is credible and not a source of risk for GM and early customers (but maybe perceived a high risk by Toyota/Honda).


2/ I don't see the LG/A123 comparison in your (interesting) document. Where is it mentioned?

The comparison is between LiMnO2 and LiFePO4. Its so obvious that LiFePO4 is far superior to LiMnO2. LG is using LiMnO2.

Thank you Prof. John Goodenough of UTA who invented LiFePO4 (at MIT).

I guess the real question is how long was A123's R&D phase before releasing the chemistry that could be in the volt.

http://www.ge.com/research/grc_7_1_33.html

GE is heavily invested in 123 and Think. According to this article the Think is out now in Norway. This should give GM some idea of the success of the 123 battery in electric car operation if this is the same battery the Volt will use.

Eitherway unless these packs over heat they will still give you 40 miles range for pretty much the life of the car. Its going to rust apart before the battery packs give you less than 40 miles. I didnt know A123 was so well 80% after 7000 Cycles thats freaking amazing.
The chiense LiFePo4 cells maintain 80% after only 2000 cycles

Too bad GM wouldnt use a small Sollar cell with petelar effect cooler to help keep those puppies cool .

Eitherway unless these packs over heat they will still give you 40 miles range for pretty much the life of the car. Its going to rust apart before the battery packs give you less than 40 miles. I didnt know A123 was so well 80% after 7000 Cycles thats freaking amazing.
The chiense LiFePo4 cells maintain 80% after only 2000 cycles

Too bad GM wouldnt use a small Sollar cell with petelar effect cooler to help keep those puppies cool .

It would take more than a small solar cell, but we don't know what they are going to do. We know a photovoltaic roof will be an option, according to Lutz. But that won't help my car while it's sitting in my garage baking at 140 degrees F.

The comparison is between LiMnO2 and LiFePO4. Its so obvious that LiFePO4 is far superior to LiMnO2. LG is using LiMnO2.
Thank you Prof. John Goodenough of UTA who invented LiFePO4 (at MIT).


Wrong. Nobody uses LiMnO2. LG uses spinel based materials.

? Goodenough=UT, Chiang (A123)=MIT. Could not say they are working together

LiMnO2 = LMS = manganese spinel. That is the designation (LiMnO2). Checkout page 11 of my previous link from Argonne National Laboratory.

Goodenough used to be at MIT, when he invented the LiFePO4. This goes back to the 90s. Way before A123.

It would take more than a small solar cell, but we don't know what they are going to do. We know a photovoltaic roof will be an option, according to Lutz. But that won't help my car while it's sitting in my garage baking at 140 degrees F.


the roof covered should be enough to power around 1500 BTUs from a petaler effect cooler.

I don't think you'd want to mess with a Peltier cooler - they demand large amounts of power for small amounts of cooling. Also, the Peltier is just a solid-state heat pump - apply DC current and one side gets cool while the other side gets hot. Using it in this fashion, you'd have to dissipate the heat transferred from the "cold" side in addition to the heat created by the device itself.

You'd do better to just have the Volt run the fan on low to bring in cooler air from the outside or something.

@Mohsen:
Wrong and wrong.

LMS=LiMn2O4 (spinel)
LiMnO2= layered Mn oxide and indeed far to be stable.

Goodenough=University of Texas since 25y+=Hydro-Quebec=Phostech-Sud Chemie

MIT=Boston=Chiang=A123

A123 different from Sud-Chemie

Get reliable info instead of trolling this thread please!
(and stop believing all what you read...even from ANL)

thanks!

b456 -

Goodenough was a Prof. at MIT - or a research assoc. there. He used to work at the Lincoln Laboratories.

Never said that MIT owned the LiFePO4 patent. You hard of reading, ey? Or dementia creeping in?

And since when do you know better than Argonne?

A123 not same as Sud-Chemie? Now first thing you said that makes sense.

I don't think you'd want to mess with a Peltier cooler - they demand large amounts of power for small amounts of cooling. Also, the Peltier is just a solid-state heat pump - apply DC current and one side gets cool while the other side gets hot. Using it in this fashion, you'd have to dissipate the heat transferred from the "cold" side in addition to the heat created by the device itself.

You'd do better to just have the Volt run the fan on low to bring in cooler air from the outside or something.

Unless you use both functions of the Peltier module ... remember if there is a temperature difference between the two sides they will also generate electricity , while moving the heat from the hot side to the cold side .... so parked in a sunny spot when the temperature difference gets large you not only move some heat out of the car but also generate electricity at the same time ... even if only small % of the energy moved is converted to heat.. it's a two for one deal.

Also because of this secondary side of the Peltier effect modules ... you can use them to recycle some of the waste heat that otherwise would have been thrown away from the exhaust of the generator ICE or the radiator for the ICE....

A cleaver engineer would design several if not all of these functions into one heat pump system... although given the time crunch for the Volt... I doubt this type of innovative heat pump system would be developed or used.

---------------------

As for the batteries... all chemistry options have their issues... weather it is time period that causes a break down... or a number of cycles that causes a break down ... or a depth of cycle that causes a break down ... or a rate of cycle that causes a break down ... or a temperature that causes a break down ... or a etc....etc..... no one battery is good for all applications anymore that all people like the same foods... from all that I have read I am still in the NiMH camp myself... but I recognize that for many people the issues that come with Lithium batteries like the A123s , and other chemistry variants are acceptable or minor compared to what those batteries offer them.... like most things in life it is pros and cons which weigh different depending on the application.

Unless you use both functions of the Peltier module ... remember if there is a temperature difference between the two sides they will also generate electricity , while moving the heat from the hot side to the cold side .... so parked in a sunny spot when the temperature difference gets large you not only move some heat out of the car but also generate electricity at the same time ... even if only small % of the energy moved is converted to heat.. it's a two for one deal.
How much energy will you create with a 40*F temperature difference, using a reasonably sized (and not cost prohibitive) Peltier device? Enough to even run a small fan? Plus, once you started cooling the interior of the car, the temperature difference would decrease, reducing the energy output from the Peltier.

How much energy will you create with a 40*F temperature difference, using a reasonably sized (and not cost prohibitive) Peltier device? Enough to even run a small fan? Plus, once you started cooling the interior of the car, the temperature difference would decrease, reducing the energy output from the Peltier.

How much you get depends on how you use it.

As a waste heat recovery device for the engine ~60% of the energy from the combustion engine is wasted out the radiator and the tail pipe as heat... from the combustion engine generator even if the Peltier device only recovers 1% of that waste heat... you are recovering over 3kW of energy... that would have otherwise been thrown away.... that is a significant amount of energy.

As far as spending energy to cool the car Peltier modules are not a high efficiency way to do it... but they are a way to make use of that energy that would otherwise be wasted.. or would otherwise cost energy to get rid of.

As for cost ... cost is always 100% subjective... a map is always less $ than a GPS system... but people want the convenience so they are willing to pay for it... a $10 Radio gives you music far cheaper than a iPod, but people pay for the iPod and all the music that goes into it because they want it.

Like anything you use it first where you get the most bang for your buck and you only use it even in that first place when it gives you enough of what you want to justify the cost to you to have it.