Perfomance data from LiFePO4 (LFP) Battery

PNjunction

Re: Perfomance data from LiFePO4 (LFP) Battery

Blackcherry04 wrote: »

Sounds like a good guess on your part, keep up the good thinking.

This didn't just come off the top of my head being an owner who has done his own lifepo4 battery autopsy to witness exactly what you have shown in that excellent pic you posted. Let's go further...

What is seen is typically what happens when you discharge a cell below 2.5 volts and OVER TIME, this destructive process happens. It doesn't happen immediately, but should be attended to asap.

The pic points out the peril of not doing proper monitoring, or not having any sort of Low Voltage Disconnect. Either at the pack level or individual cell level - we'll leave that for another thread. Still, 2.5v is a common LVD setting.

Note that under normal use, such as a housepower bank, the LVD would be at around 3 to 3.1v, which is basically about 90-95% DOD.

EV drag racers who do "bottom balancing", do so at 2.7 or if they are adventurous, do so at 2.6v. In either case, once the bottom balancing is finished, one should recharge immediately. You don't want to bottom balance and then leave them like this for a year in your garage. For our application, top-balance is the norm.

2.5v and below is where this damage happens on a much faster rate. I've mentioned that it is usually accompanied by case swelling, but that is not always, the uh, case. The vent is supposed to blow when gas buildup becomes too large. Note too that battery-cell banding is to prevent mechanical shifting of loose individual cells made into a battery, preventing undue torque on the terminals, and is NOT a way of preventing low-voltage swelling damage. While banding physically does this, the idea is to have it vent instead of blowing out the case. Also once a battery has swollen, the anodes and cathodes are no longer properly aligned, and the "overhang" is shot leading also to dendrite formation - very bad. Proper overhang (where the anode or cathode is slighly physically larger than the other) is something that counterfeiters and poor manufacturers get wrong.

This also points out the foolishness of buying used batteries since this problem is not always accompanied by tell-tale case swelling. If you were able to get to a cell in an extremely low-voltage condtion, you CAN bring it back to service, but ONLY if you apply very small recharge current - something like C/50 until the terminals reach the normal operating range of about 3v. Then and only then can you apply full recharge current. While the battery may be serviceable after this recovery, you haven't done it any favors, and should now be suspect from then on. Ideally you just learn your lesson and replace it.

So yes, buying used is a very bad idea. Unless you can vouch for the integrity of the seller, you could easily end up with one of these that don't have case swelling, but may have already vented.

Nothing new here about buying from reputable sources, be it lead acid or lifepo4.

However, what totally abused cells like this are good for are for making dramatic internet videos! Nobody spends that kind of money on good cells, and then abuses them to empty their wallet - unless they are just DIY'ers who don't know any better. Hopefully info such as this will keep the money in their wallet instead of on screen.

northerner

Re: Perfomance data from LiFePO4 (LFP) Battery

PNjunction wrote: »

What it really demonstrates is that nobody in their right mind does a direct short like that on purpose - however, even under these conditions of total abuse you witness the electrolyte cooking off and the battery itself doesn't catch fire. Later, what DOES cause the battery to catch on fire and melt is the HIGH RESISTANCE STEEL BOLT which melts the battery and the wiring infrastructure.

Ironically, although at first totally dramatic, it demonstrates that lifepo4 is not a firecracker unlike the chemistry of most laptop batteries. The fire department is standing by and instead of applying extinguishers to the whole mess, a fan is used to disperse the cooked electrolyte vapor.

Essentially, there is no defense for shoddy installation schemes that have no fuses or other safety features to help safeguard against these type of accidents no matter the chemistry.

A far better video of capacity testing under *normal* conditions:
https://www.youtube.com/watch?v=oJ8AfJsOhsM

Now we're talking.

I have no issue with performance of these LiFePo4 cells as demonstrated with the video link. What I do have concerns about is: 1) overall calendar life span (despite the high cycle life numbers posted for lower average DOD usage) and: 2)keeping the cells in balance long term.

I see that Balqon makes LiFePo4 battery packs using Winston cells and their own bms system. Does anyone know if the bms still requires monitoring and intervention when things start to get out of balance, or does the bms handle this? Or another question, is one better off buying cells and putting in a state of the art bms system, that keeps the cells balanced irregardless? I think that the cell balancing is the biggest issue I have with these batteries. I may consider going with a small bank (lower investment) and try them out before committing to a large expensive bank.

The link to the short circuit video demonstrates just how safe the battery is in the event of an accident, and that there is no fire. Wouldn't want to try that with a lead acid battery.

gww1

Re: Perfomance data from LiFePO4 (LFP) Battery

Northener
You ask about intervention. I did read on the outback forum that a differrent user then the one I mentioned earlier in this thread, was saying when his bms disconnected the loads from the battery due to low voltage, he had to manually intervene or they would not start charging the next day. It would seem something easy to fix but that is all I remember.
gww

northerner

Re: Perfomance data from LiFePO4 (LFP) Battery

gww1 wrote: »

Northener
You ask about intervention. I did read on the outback forum that a differrent user then the one I mentioned earlier in this thread, was saying when his bms disconnected the loads from the battery due to low voltage, he had to manually intervene or they would not start charging the next day. It would seem something easy to fix but that is all I remember.
gww

My question relates more about keeping the cells in balance, without manually having to intervene, but thanks for the info. Something else to consider perhaps?

northerner

Re: Perfomance data from LiFePO4 (LFP) Battery

Balqon uses Clean Power Auto electronics for their battery bms's. I have read other threads where users, who purchased batteries from Balqon, received no info or manuals about the BMS system. Apparently, they have what's called a Quanbuss BMS system used for controlling larger banks for vehicles, etc..

http://www.balqon.com/wp-content/uploads/2013/07/36_36bms_brochure_balqon.pdf

karrak

Re: Perfomance data from LiFePO4 (LFP) Battery

northerner wrote: »

My question relates more about keeping the cells in balance, without manually having to intervene, but thanks for the info. Something else to consider perhaps?

There are a number of BMS systems on the market that will automatically balance the cells with no intervention if you want to go down that route.

From my experience and what I have heard from others that have LFP batteries I don't think balancing is a big issue. If the battery is properly balanced before it is put into service (this is important) and the charging voltage is kept below 3.5 volts/cell (14 volts for a 12 volt battery) the batteries stay in balance for a long period of time. I last manually balanced my battery in April last year and it is still OK.

This is not to say that one should just "set and forget", just as you wouldn't with LA batteries. If you don't have some sort of cell monitoring which tells you the individual cell voltages which would be my choice, you should at least check the individual cells with a multimeter on a regular basis when the battery is at the end of the absorb charge cycle just before it goes to float.

Simon

PNjunction

Re: Perfomance data from LiFePO4 (LFP) Battery

northerner wrote: »

My question relates more about keeping the cells in balance, without manually having to intervene, but thanks for the info. Something else to consider perhaps?

The Housepower bms would do fine. Consider that these cells are intended primarily for heavy current EV acceleration. In our application, ie "Sub-C" they would be pampered provided you didn't undersize them to begin with. Sub-C operations means that they tend to stay in whatever state of balance you last left them in.

Still, you could raise your charge voltage level to get the cell balancers activating, and when done, you can then run conservatively under the balancing threshold and redo balancing whenever you felt necessary. Note that top balancing like this merely balances exposure to voltage, and not necessarily balance in capacity, since each cell typically has manufacturing variances in both capacity and IR. That is one reason I don't go bananas over small differences in balance voltage. For me, I get concerned when the variance starts to creep beyond 0.050v and would definitely shut down seeing anything more than 0.100v difference. Trying to achieve the craziness of 0.005v difference means you are spending too much time at high voltage for no net gain in relative capacity balance whatsoever. Ideally get cells matched at the factory.

I understand the you can enable / disable the balancing feature on the Housepower bms. Just use it when you feel like it, but certainly not every cycle.

You know, you can demonstrate this first by purchasing a small 12v system like I did with my GBS lifepo4 batteries reviewed elsewhere here. My "learner" set is doing just fine thank you. Fortunately all the physics scale to those 10 times as large.

Blackcherry04

Re: Perfomance data from LiFePO4 (LFP) Battery

PNjunction wrote: »

You know, you can demonstrate this first by purchasing a small 12v system like I did with my GBS lifepo4 batteries reviewed elsewhere here. My "learner" set is doing just fine thank you. Fortunately all the physics scale to those 10 times as large.

lol..... Keep telling yourself that a 12 v , 40 amp hr battery on a bench being charged from your 5 amp charger on the grid is the same as a 48v, 500 amp hr bank with 64 cells off grid being charged from PV and a generator is the same physics scale. We've seen the advise you have for that, not even doable because it's talk with no experience.

Cariboocoot

Re: Perfomance data from LiFePO4 (LFP) Battery

Blackcherry04 wrote: »

lol..... Keep telling yourself that a 12 v , 40 amp hr battery on a bench being charged from your 5 amp charger on the grid is the same as a 48v, 500 amp hr bank with 64 cells off grid being charged from PV and a generator is the same physics scale. We've seen the advise you have for that, not even doable because it's talk with no experience.

In other words, systems of any type do not scale up linearly.

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

Just looking at the graph on page 5 http://www.balqon.com/wp-content/uploads/2013/07/35_35balqon_battery_2013.pdf an LiYFePo4 (if I'm right) and thinking about discharge between the limits of 20% and 80% DOD.

From their graph (watch out for the changing scale on the vertical axis) that seems to be a very safe range away from the drop off and a very narrow voltage range a bit < 3.2 to ~ 3.1V, which seems to be supported by the video https://www.youtube.com/watch?v=oJ8AfJsOhsM


Interesting that he got that additional output without hitting the 'knee', even though he started the video at ~4rs Hr in with 100+ Amps load, battery at 2.955 Vpc / 11.82V

Shut down at 2.9225Vpc / 11.69V 425Ah removed at 0.25C , Brochure states lowest V is 2.8 in table just above graph

PNjunction

Re: Perfomance data from LiFePO4 (LFP) Battery

My GBS cells follow the same pattern and voltages when discharging at 0.25C like he did, despite that fact that my 40ah battery is 10 times smaller. I'm conservative though, and stop sooner under normal use.

Blackcherry and Cariboo: perhaps I'm wording things wrong somehow. This is what I'm alluding to is that be it a 4, 40 or 400ah cell, they have the same characteristics - even very closely across a change in manufacturer.

karrak

Re: Perfomance data from LiFePO4 (LFP) Battery

westbranch wrote: »

Just looking at the graph on page 5 http://www.balqon.com/wp-content/uploads/2013/07/35_35balqon_battery_2013.pdf an LiYFePo4 (if I'm right) and thinking about discharge between the limits of 20% and 80% DOD.

When I asked Winston about the best operating conditions to maximise the lifespan of their batteries in our application they agreed that keeping the SOC centred on 50% was the best strategy, but went on to say that between the range of 3.65volts/cell and 10%SOC was OK. They also said to keep the battery voltage in the range 3.65 volts/cell to 2.8 volts/cell.

I am slightly more conservative than this and try to keep my battery in the range 3.45 volts/cell which equates to >90%SOC with the occasional increase to 3.6 volts/cell to balance the battery. On the lower end I try not to take my battery below 3.0 volts/cell (even under load) and will turn all battery loads off if the voltage on any cell drops below 2.8 volts/cell. As you can see from my figures the lowest level I took any cell down to in the past few months was 3.047 volts. It is very rare that my battery gets taken below 25%SOC. If I had a generator or was connected to the grid I would start charging the battery around 25%SOC and would stop charging around 50%.

Simon

reefbum

Re: Perfomance data from LiFePO4 (LFP) Battery

I was looking at Lifepo4, from the information I was able to find Lifepo4 has a memory effect which coupled with no long term usage reports from the RE wold prompted my purchase of FLA.

Scientists at the Paul Scherrer Institute PSI, together with colleagues from the Toyota Research Laboratories in Japan have now discovered that a widely-used type of lithium-ion battery has a memory effect.

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

From 2013

http://www.psi.ch/media/why-lithium-ion-batteries-fail

ADD:

Materials in lithium ion battery electrodes expand and contract during charge and discharge. These volume changes drive particle fracture, which shortens battery lifetime. A group of ETH and PSI scientists have quantified this effect for the first time using high-resolution 3D movies recorded using x-ray tomography at the Swiss Light Source.

Blackcherry04

Re: Perfomance data from LiFePO4 (LFP) Battery

Watch out for the sharp knives. I posted this one since it's LiFeP04 specific.

http://phys.org/news/2013-04-memory-effect-lithium-ion-batteries.html

northerner

Re: Perfomance data from LiFePO4 (LFP) Battery

Impurities in the LiFePo4 material can also affect it's aging process.

http://www.eng.uwo.ca/people/asun/Paper/Nature%20of%20LiFePO4%20aging%20process%20Roles%20of%20impurity%20phases.pdf

Looks like the memory effect will not a big deal, when using LFP batteries for off grid usage, when batteries get regular daily cycling.

Blackcherry04

Re: Perfomance data from LiFePO4 (LFP) Battery

northerner wrote: »

Looks like the memory effect will not a big deal, when using LFP batteries for off grid usage, when batteries get regular daily cycling.

How about supplying us the location of that quote. Sounds like BS damage control. I though one of your big points was they didn't have to be charged daily to 100%

northerner

Re: Perfomance data from LiFePO4 (LFP) Battery

Blackcherry04 wrote: »

How about supplying us the location of that quote. Sounds like BS damage control. I though one of your big points was they didn't have to be charged daily to 100%

It's in the article of the link you provided

According to Petr Novak, Head of the Electrochemical Energy Storage Section at the PSI and co-author of the publication, the study disproves a long cherished misconception: "Ours is the first study that has specifically looked for a memory effect in lithium-ion batteries. It had simply been assumed that no such effect would arise". To acquire knowledge via research is often a fruitful mix of speculation and diligence: "Our finding results from a combination of critical investigation and careful observation. The effect is in fact tiny: the relative deviation in voltage is just a few parts per thousand. But the key was the idea of looking for it at all. Normal battery tests usually run deep, and not partial charging/discharging cycles.

Read more at: http://phys.org/news/2013-04-memory-effect-lithium-ion-batteries.html#jCp

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

So what do you think this tidbit from near the end means, as it is not in 'normal' electro-speak,

The effect is in fact tiny: the relative deviation in voltage is just a few parts per thousand.

you 2 guys are too fast for me

I got out of the article that partial charge and discharge would lead to the memory effect.

If the battery is not fully charged, a certain number of lithium-rich particles that have not made it over the barrier will remain

then they say, about car batteries,

It is indeed absolutely possible that the effect could be detected and taken into account through clever adaptation of the software in battery management systems,

so it might just be a toss of the coin...???

Blackcherry04

Re: Perfomance data from LiFePO4 (LFP) Battery

northerner wrote: »

It's in the article of the link you provided

Thats NOTwhat you said, you have no clue if it is or isn't.

Blackcherry04

Re: Perfomance data from LiFePO4 (LFP) Battery

Kind of Interesting.

Attachment not found.

northerner

Re: Perfomance data from LiFePO4 (LFP) Battery

Blackcherry04 wrote: »

Thats NOTwhat you said, you have no clue if it is or isn't.

They are just referring to repeated partial charging/discharging of the battery. Most off grid battery usage involves heavier charging and discharging. ie (you go through a long period of discharge and then recharge usually once per day, so generally no repeated charging/discharging) They also say that the memory effect is very small and is what I meant that it sounds like it would not be a big deal, if using these batteries for off grid use.

What may be a bigger deal is the actual calendar life of these batteries, which is still unclear (and of course dependent on how they are treated as well).

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

too bad there wasn't a 40% to 80% cycling to reflect PV type use :cool:

rather than 0% to 100% etc

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

and looking at those 3 graphs isn't the BMS/Motor supposed to keep the battery above 0% SOC?

(that is like the 50% soc for FLA's, point of no return)

unless you trickle charge them for a while, so you would be ICE powered till you got the SOC high enough, or in our case 'gen time'

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

Interesting research paper, especially the graphs on the page following with a charge to 50% and a 40% discharge, more like our type of cycle..

http://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0CB8QFjAA&url=http%3A%2F%2Fwww.researchgate.net%2Fprofile%2FMohammad_Farkhondeh%2Fpublication%2F265475829_Mesoscopic_Modeling_of_Li_Insertion_in_Phase-Separating_Electrode_Materials_Application_to_Lithium_Iron_Phosphate%2Flinks%2F545bc8f10cf249070a7a7ec8.pdf&ei=_tbKVKqFAYmuogSOl4Jg&usg=AFQjCNGJN-iZKbyvkVg3vDLezJCBFNZuxw&sig2=q3ZINQG9osdbqF1y4hCktQ&bvm=bv.85076809,d.cGU

westbranch

Re: Perfomance data from LiFePO4 (LFP) Battery

Have a look at ReefBum's comparison
http://forum.solar-electric.com/showthread.php?26203-Battery-Cost-Compare&p=211864#post211864

northerner

Was just on the Balqon site and noticed a couple of changes. First of all, it appears that all the various sizes of cells they sell are listed as "in stock". If that's the case, perhaps the long periods of waiting for shipments from China has been dealt with?

The other is they all got quite a bump in cycle life rating. Not sure if these are new cells, or maybe they're trying to make the old ones look more appealing?:D I still am skeptical of calendar life, so not sure if it will really make a difference anyway.Anyways, they are all now rated for 3500 cycles to 80% DOD, 7000 cycles to 50% DOD, and 17,500 cycles to 20% DOD, all at a discharge rate of 0.5C. Sounds to me like they want to sell more batteries?

westbranch

from the 700Ahr battery page, http://www.balqon.com/store-2/#!/700-Ahr-Lithium-Iron-Phosphate-Battery-In-Stock/p/11906026/category=2736691 listed as a LiFePo4 battery but...

Exclusively available from Balqon is Yttrium Iron Phosphate batteries which improve battery life, safety and thermal characteristics. Iron phosphate is one of the safest batteries in the Lithium battery family mainly due to a strong oxygen bond with phosphorous that reduces thermal runaways at high voltages. Yttrium a rare earth metal is known for its thermal characteristics, most often used on high performance spark plugs to reduce damage from heat and oxidation from elements under high temperature. Yttrium batteries operate at lower voltage of 4.0 volts max and have longer life than a LFP battery which operates as high as 4.2 volts and more prone to thermal runaways.

I tried to make some sense of their numbers but not close enough for me so I sent an inquiry about cycles claimed and cycle life ...

northerner

westbranch wrote: »

from the 700Ahr battery page, http://www.balqon.com/store-2/#!/700-Ahr-Lithium-Iron-Phosphate-Battery-In-Stock/p/11906026/category=2736691 listed as a LiFePo4 battery but...


I tried to make some sense of their numbers but not close enough for me so I sent an inquiry about cycles claimed and cycle life ...

Funny thing is the previous cells also had yttrium, as they were Winston cells. Not sure if these cells are something new? Also, note that the shelf life is rated for 5 years. Really has one wondering what the calendar life is?

westbranch

Concur about vagueness... what does this mean?

5 years at 60% State of Charge (SOC)

Does it need to be depleted to 60% to store or that it will go down to 60% after 5 years???

Self discharge rate 1-2 % per month

1% per month = 12% / yr = 60% dod after 5 years...