Why not LiFePO4?

12357

Comments

  • CariboocootCariboocoot Banned Posts: 17,615 ✭✭
    Re: Why not LiFePO4?
    westbranch wrote: »
    OK just so I get this right, I have a 21 W load x 24 hrs = 504 W hrs x 2 (50%, FLA no reserve) = 1008 Wh /12V = 84 Ah battery x .84 (Fla to LFP) = 70.56 Ah LFP battery is needed. Correct?

    That's about right.
    Personally I don't think the LiFePo advantage is in the ability to run a smaller battery bank, especially with the still huge cost difference. That slightly smaller capacity battery will still cost you more money, and will not necessarily last any longer. To my mind the advantage is in the flexibility of SOC without battery damage; something many people can make use of and may well find worth the extra price.

    This is similar to the MPPT advantage not being in the increased output power but rather greater flexibility in design (and ultimately over-all lower cost for systems of a certain size and above). People focus on the wrong advantage and don't consider that to be worth the extra money.
  • Blackcherry04Blackcherry04 Solar Expert Posts: 2,490 ✭✭✭
    Re: Why not LiFePO4?

    I'd like to know how Lithium Depletion effects the overall life of them and where power fade and capacity fade comes into play. I sized in a extra 15% into my current FLA bank to see if they remained healthy , how long I could hold off replacement if the capacity remained high enough to run my loads.
  • ReedReed Solar Expert Posts: 55 ✭✭
    Re: Why not LiFePO4?

    The battery bank is composed of four Mazanita Micro batteries in series. Each battery is composed of four China Aviation Lithium Battery Corporation 180 Ah 3.7 V nominal cells so that each battery is a 180 Ah 12 V nominal battery.

    The four batteries in series makes a 48V nominal 180 Ah battery bank (or equivalent to a 720 Ah 12 V nominal battery). This is just about 9500 Wh.

    Cell Bank
    Absorb 3.6 57.6 V
    Full charge current 0.05 9 A
    Max cell volt 3.9 62.4 V
    Float 3.4 54.4 V
    Disconnect 3.1 49.6 V

    We use a TriStar (Morningstar) TS-MPPT-45 Controller. This is quite sufficient since the power comes from the two banks of 235 W panels at around 90 V (maximum noted power was 1300+W) and this gives an amperage of around 14 A.
    Reed and Elaine
    ZoNiE wrote:

    I decided to forge ahead on all 12 pages of this sitting in an airport, and glad I did, despite the early kibitzing.

    Again, glad this kept going. This Lurker is still here and learning...

    Go for it! :cool:

    Also, to Reed and Elaine, sounds like you have if figured out. Forgive me if I missed it, but what size is your bank? (AH/volts?) what charge controller do you use, or, if you have a blog somewhere with all of this, please redirect me.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?
    I'd like to know how Lithium Depletion effects the overall life of them and where power fade and capacity fade comes into play. I sized in a extra 15% into my current FLA bank to see if they remained healthy , how long I could hold off replacement if the capacity remained high enough to run my loads.

    As per your chart in post #102, you can see that while under load, be it large or small, the overall slope of the discharge doesn't change, but you can have a voltage drop when running over 0.5C continuous. This is unlike lead where due to Peukert, that grade of that slope changes radically under high current draw.

    As a solar user who wants to make it through the night, we'll NEVER be pulling 0.5C or more, which is what EV, RV, Marine might be pulling - still the overall capacity is not outrageously different, it is just the voltage drop. Yes, if you magnify the graph, you'll see what might appear to be a minimal peukert effect of something like 1.06 or less. Again, we'll never go to those extremes of current to make it a big issue.

    Consider that when compared to lead, an 80% discharged lifepo4 battery should be STOPPED at around 12.8v (using a 12v battery as a convenient example), whereas with lead under load, you generally START out at 12.8v or so and drop down from there. If you equipment is voltage sensitive, or runs more efficiently at higher voltages under load, then that is a plus that is sometimes not considered from an operational standpoint. So too is running a panel higher in the power-point of the charge curve with lifepo4, since we are not dragging a panel below 12.8v upon charge, but starting at that level. Combine this with internal resistances even smaller than the best AGM's, you have more efficient charging taking less time - which is the whole mantra of solar - do it in less time. So when one is considering the "price" of each cycle, take this into account.

    The anodes and cathodes of a lifepo4 battery do not "shed" any material under charge discharge, so theoretically they should last forever. The biggest contributor to loss of cycle life is the growth of the SEI layer. This is a small layer of inorganic compounds coming from the electrolyte that block the receptor sites of the anode and cathode material and grows each time you charge or discharge. Initially, it is a protector, since without any sort of SEI layer, the battery would only last a month under use. However, over time, this layer can block so many receptor sites, that capacity and performance falls to less than 80% capacity, and one usually desires to replace the battery.

    High heat and high voltages (even within spec) contribute to the acceleration of the parasitic reactions and their generation of inorganic clogging compounds and electrolyte oxidation.
  • Blackcherry04Blackcherry04 Solar Expert Posts: 2,490 ✭✭✭
    Re: Why not LiFePO4?
    PNjunction wrote: »
    As per your chart in post #102, you can see that while under load, be it large or small, the overall slope of the discharge doesn't change, but you can have a voltage drop when running over 0.5C continuous. This is unlike lead where due to Peukert, that grade of that slope changes radically under high current draw.

    As a solar user who wants to make it through the night, we'll NEVER be pulling 0.5C or more, which is what EV, RV, Marine might be pulling - still the overall capacity is not outrageously different, it is just the voltage drop. Yes, if you magnify the graph, you'll see what might appear to be a minimal peukert effect of something like 1.06 or less. Again, we'll never go to those extremes of current to make it an issue.

    tbc...
    Thats not what I said about loading, thats the power curve of a new healthy battery. I looked at some information that show what they do in the EV market , I'd like to know if the RE market fairs any better about how they age. Are you on grid power with yours ?? Was that Samlex charger for charging your battery ?? How many amp hrs is it ??

    I know how to use google, thank you.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?
    ..I looked at some information that show what they do in the EV market , I'd like to know if the RE market fairs any better about how they age. Are you on grid power with yours ?? Was that Samlex charger for charging your battery ?? How many amp hrs is it ??

    That's the ironic thing - these batteries are normally intended for EV use, yet we don't use them like they do and will live a pampered life. That is, while EV reports are useful, those guys are running typically down to 90% DOD instead of 80%, may be suffering from bms failures, or other forms of cell abuse like having high-resistance connections under accelleration well above 3C, melting cell terminals while drag racing, using dimestore voltmeters to monitor their packs, etc. :)

    Our relatively simple low-voltage 4S/8S/16S, 12/24/48 volt batteries in a simple series string, is vastly simpler in care and feeding. Tip: get the largest cells you can to fit your application to cut down on cell count and additional jumper wiring complexity. Ie, reduce point-of-failure nodes.

    A more reliable resource would be the guys using them as marine house-banks, solar or not.

    If you are looking for 6 year+ plus reports, (taking into account amateur blunders) I'm afraid your only recourse is to join the party. I think you'd do VERY well. Even though the technology has been stable for over 15 years, it was only in the last 5 or 6 years that an average human could obtain them thanks to the likes of GBS, Winston, CALB, etc.

    Yet, work still goes on today with improvements in the electrolyte to reduce the parasitic reactions, and slow down the growth of the SEI layer. Everyone seems to have their own special recipe, some of which never make it to market anyway, so I'm dealing with what I can obtain NOW, and so far so good.

    I don't even want to speculate why the major lead manufacturers here didn't adopt the technology for large prismatics for 10 years. Anything here, unless it comes from the CEO, is just amateur crystal-ball gazing. At any rate, it doesn't matter now.

    Re the Samlex - I really like it, mostly because it is a high-quality CC/CV charger that I can set down to 14v. It is only 15A, but will charge my 20 and 40ah batteries just fine. Even with the 20ah battery, 15A from the samlex, (which quicky runs down to about 10A when near 13.7v or so), causes zero drama or heating. If I didn't have the 20ah battery, I would have chosen an even larger Samlex, but I have the time on AC to not make it that critical. You won't notice much of a difference in heat until you go higher than 1C, which I've done to a few smaller Shorai and other A123-internal powersports batteries. Noticeable there, but not too alarming.

    My solar CC is just a simple pwm Schneider / Xantrex, which I can set the voltages to very accurately, and of course NO temperature compensation. I'd like a Genasun, but I don't like their choice of 14.2v default charging, and won't pay for custom reprogramming down to 14v max. But that's just me.

    At any rate, I'll NEVER be able to achieve even .5C with solar on an appropriately sized battery for house-bank use, so I take that into account. The way we use these things, I expect to easily meet or exceed the specifications from the manufacturer because they are intended for much more heavy duty use from an EV.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    Sudden death syndrome:

    Like all products, some things just fail right from the factory and need to be addressed from that standpoint. Try not to buy junk, or deals too good to be true, like gray-market used / abused / diy stuff. Make sure your dealer is authorized, reputable, and knowledgeable. This shouldn't be anything new to us, but there are temptations out there preying on those who just want to go lifepo4 because they think it is "cool". Caveat emptor.

    There are also reports of sudden death, and aside from outright manufacturing problems, ABUSE is the culprit, either from too much heat or too high a voltage (too low too!), where one day you have full performance, and the next day - zilch!

    What is typically happening is due to abuse, growth of the SEI layer is very fast, along with electrolyte oxidation. Unlike lead which warns us with tell-tale signs of poor performance, lifepo4 will give you all it can until at that very last recharge, where enough of the receptor sites is smoothed over with inorganic materials, and there is nowhere for current to flow. Let's just say that with an abused cell, the small amount of receptor sites that ARE available, are working their guts out, until a full-closure occurs on the very next recharge. Lights out.

    Moral - treat your cells kindly. A *minor* abuse here an there won't go immediately noticed, but in the large scale of things, one just wants to cut down the accumulative affects of anything. That is also the reason why I am playing it conservatively with lower non-drop-in voltages, and conservative DOD cutoffs. Unlike an EV'er, I have that luxury.
  • ReedReed Solar Expert Posts: 55 ✭✭
    Re: Why not LiFePO4?

    PNJunction - Been lurking on and off with the Aussie Grey Nomads forum for a while. Finally joined this last week. They are far more likely to go for innovative developments. Most of the Grey Nomads have solar and a number of LFP. Have been in correspondence with one guy with 1200 W on his 28' or so TT (caravan) and LFP. Many more are using DC air conditioning and refrigerators.

    He goes with your idea of not needing a BMS and has set the controller to cut out at 27.2 V (exactly one half our 48 V nominal float of 54.4 V). As you have noted, the EV world needs BMS with their much higher power requirements. Manzanita Micro packages for the EV crowd and ours came with a sophisticated BMS. Since we already have it, we shall continue to use it. It probably allows us to get to float somewhat more quickly. Absorb is 57.6 V but that is quite a bit below the max of around 62 V.
    Reed and Elaine
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    Lifepo4 from a "system" standpoint in low solar-insolation periods:

    Because a lifepo4 can easily handle up to 1C of charge, (although most manufacturers like .5C for optimum life - I have no data between the two), they are also good candidates for low solar insolation areas.

    Consider: FLA batteries - typically limited to C/12 to C/8 charge current, with C/10, aka 0.10C as the middle ground. Even if you had the means for a larger array, you are limited to C/8 for the most part.

    AGM: Conventional lead-calciums are limited to about 0.3C maximum inrush current. TPPL agm's, like Concord, Odyssey, Optima etc have no practical maximum inrush current as long as voltage is controlled, which is fantastic. I won't get too far into that here as it has been covered elsewhere.

    HOWEVER, to get the most life out of agm's, you NEED at least 8 hours of float to actually finish a charge, despite the amount of initial charge current. This float is not related to any sort of storage or maintenance afterwards. But to actually finish a single charge cycle, it requires 8 hours or more of float, something we don't have with solar usually. That means compensation using techniques like extended-absorb, IUIA charging algorithms etc. AGM users kind of know this going in, where the reason you want to hammer tppl's like Odysseys is to actually finish the absorb cycle and get to float (or other compensation) as fast as you can before the sun goes down.

    With lifepo4, it is pretty straightforward. If you have the means, you can get your solar array up to .5C, or perhaps 1C in extreme cases.

    Normal power-budget planning should handle most normal solar-insolation issues, BUT if you are in the the pacific northwest for example, insolation can easily be widely unpredictable. In this case, if you have the means, a much larger solar array can easily be handled by lifepo4, and with no need for any sort of float or compensation algorithms. Lifepo4 eases the problem of poor solar insolation, or those who want a large amount of autonomy, and keeps things simple.

    This is one area that EV'ers don't have to think about. :)
  • zonebluezoneblue Solar Expert Posts: 1,218 ✭✭✭✭
    Re: Why not LiFePO4?
    PNjunction wrote: »
    Our relatively simple low-voltage 4S/8S/16S, 12/24/48 volt batteries in a simple series string, is vastly simpler in care and feeding. Tip: get the largest cells you can to fit your application to cut down on cell count and additional jumper wiring complexity. Ie, reduce point-of-failure nodes.

    I wonder if LiFePo4 changes some of the thinking around system voltage? If much of the bigger system equals bigger system voltage logic is predicated upon FLA charge rates, and pack discharge rates to support inverter loads, then surely all that changes here?
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?
    Reed wrote: »
    He goes with your idea of not needing a BMS and has set the controller to cut out at 27.2 V (exactly one half our 48 V nominal float of 54.4 V). As you have noted, the EV world needs BMS with their much higher power requirements.

    I think we need to define "BMS" as a battery-management-system, which incorporates an HVD, LVD, and optional balancing. These differ from just simple "balance boards" that are tacked across the terminals.

    So in that case, I'm not really against using a BMS for the HVD or LVD functions, but question the need for balancing if one is able to operate conservatively, and cells are reasonably the same.

    My BMS consists of
    * HVD controlled by the setting on my charger or charge-controller.
    * The LVD consists of either watching coulombs/ah, my brain (dangerous) or the little powerwerx emergency-vehicle relay. My hard limit is no cell lower than 3.1V under load, which at a low 0.1C discharge rate, allows me enough time to catch it even with nothing more than a Fluke multimeter.
    * Balance consists of a reasonably close SOC among the cells (about 0.050 mv or less) at top of charge (no more than 3.6v under charge regardless) and a pack level of no more than 14v, or 3.5v per cell calculated.

    Still I understand those who desire to use more automated methods like yourself.

    I can pick lifepo4 apart and only concentrate on the negatives, like any battery chemistry, but for me they are outweighed by the positives in a solar application. Like you, I think that if more people got some hands-on with a small exploratory prismatic battery, instead of bench-racing specs, pricing, politics and so forth, they would have a different perspective.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    For those who want to see a chart from the manufacturer about cycle life, in this case GBS cells reaching 2000 cycles:

    http://www.elitepowersolutions.com/docs/Cycle%20Life%20Testing%20of%20a%20GBS%20cell.pdf

    Thanks to elitepowersolutions for providing this.

    Note that these are 100% DOD cycles (which does NOT imply zero voltage!). At 80% DOD, we should do much better.

    However, this test does not take into account calendrical aging, or the exacting science of measuring coulombic efficiency as done at Dalhouse University by Prof Jeffrey Dahn's group. It also means that the lab knows what they are doing, and not hammering them with full current immediately while deep down in the discharge knee.

    For bench-racing purposes, it is a bit reassuring, since I'll never go to 100 DOD willingly.
  • zonebluezoneblue Solar Expert Posts: 1,218 ✭✭✭✭
    Re: Why not LiFePO4?

    Bit short on details. If thats knee to knee... 0.5C charge and discharge, 25*C? Then pretty decent.

    What exactly determines the calendar life?
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    Yep, that's knee to knee, actually well into the discharge knee.

    The best I can determine from what I've gleaned from Rick Suiter on the EVDL list, GBS charges CC/CV to 3.6v, and lets current drop to C/20 before terminating charge. It is then discharged to make sure that it meets the capacity and IR specs. Then it is discharged to a very low 2.5v, and allowed to sit for some time. At this very low voltage, any cell with high internal resistance, leakage or other problems will show up early if the cell takes a nosedive and doesn't maintain near 2.5v.

    EV users will be familiar with this process for "bottom balancing", although they never go as low as 2.5v - usually no less than 2.6 or more commonly 2.75. They will also watch the ocv for awhile to look for those taking a nose-dive, as at this low voltage, that is where these kinds of problems are more easily detectable. Usually they have all the cells in parallel when bottom balancing, but some are known to do it individually in series to each one - it just takes much more time.

    If a cell is stable at 2.5v for a period of time, then GBS will recharge to about 50% before making them available. I'm sure other manufacturer's do much the same.

    My take on calendrical life - since we don't have a decade's worth of consumer level testing behind us, all these tests prove is that you can repeatedly hammer-cycle, and if that's your application then ok. But MY question is how stable is the electrolyte 10 years out? How well are the separators doing 10 years out, even with the best of care?

    One thing I'm doing is trying to read between the lines of the spec sheets. Just because there are specs doesn't mean that they are optimal for our application. For instance, although one can easily take these cells to 1C charge, most manufacturers recommend no more than 0.5C. But they don't say why. Does it shorten the life of the electrolyte in the long run? Do the separators degrade long term? Fortunately, as solar users who don't design for teeny-weeny capacity, we will always be under .5C charge/discharge, usually well below.

    This kind of stuff makes me want to go into the business myself. Problem is, I'd want to make them so good that I've never have any repeat customers, and go bankrupt quickly. :)
  • sub3marathonmansub3marathonman Solar Expert Posts: 300 ✭✭✭
    Re: Why not LiFePO4?
    PNjunction wrote: »

    One thing I'm doing is trying to read between the lines of the spec sheets. Just because there are specs doesn't mean that they are optimal for our application. For instance, although one can easily take these cells to 1C charge, most manufacturers recommend no more than 0.5C. But they don't say why. Does it shorten the life of the electrolyte in the long run? Do the separators degrade long term? Fortunately, as solar users who don't design for teeny-weeny capacity, we will always be under .5C charge/discharge, usually well below.

    One thing I've been wondering, is there is some minimum charge rate for the Lithium cells?

    Many times it is recommended for the PbA to have a minimum of 5% charge rate, 0.05C, but I haven't seen any such minimum recommendations for Lithium.
  • inetdoginetdog Solar Expert Posts: 3,123 ✭✭✭✭
    Re: Why not LiFePO4?
    One thing I've been wondering, is there is some minimum charge rate for the Lithium cells?

    Many times it is recommended for the PbA to have a minimum of 5% charge rate, 0.05C, but I haven't seen any such minimum recommendations for Lithium.

    AFAIK there is no such minimum rate. But anything less than the self-discharge rate will not do you much good. :)
    SMA SB 3000, old BP panels.
  • Blackcherry04Blackcherry04 Solar Expert Posts: 2,490 ✭✭✭
    Re: Why not LiFePO4?

    This makes my head hurt. 1,000 cycles, 2,000 cycles ( 2.7 years - 5.5 years of daily solar cycling, but maybe 10 years 3,600 cycles ) to some level, but they will do better than a 20-25% loss in capacity if I cycle them to 50% dod and recharge them to somewhat less than 100% is what I am getting from all this. They sound like FLA batteries at 3x the cost without the distilled water. They are light and will probably fit some people needs.

    Attachment not found.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    You chopped off the top half containing some major information. :)

    This is at 0.5C charge / discharge. You will NEVER do that with FLA. Of course, doing 0.5C with lifepo4 with a housebank means one did a bad job of calculating their power budget. Running larger capacity than EV'ers do (this chart was meant for EV'ers basically), and low currents relative to that application increases the lifetime.

    All I'm saying is to keep in mind OUR application or relatively low voltage and low current compared to EV use. Also keep in mind the higher efficiency from a systems standpoint of much higher voltage under load, much lower internal resistance, with the ability to do that throughout the *entire* capacity of the battery - well 80%. Will your inverters or other infrastructure run more efficiently, cooler, and need replacement less often?

    Take a tip from the guys who have been there for years already - this argument is basically over, and those that want to run lead can do so. Those that want to run lifepo4 usually do so after some first-hand experience, and never go back, DESPITE the initially higher cost.

    Basically the cost issue has been discussed to death in both camps, and is just a waste of time, and merely serves to increase the signal to noise. Sure they cost more. Pick your poison.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?
    One thing I've been wondering, is there is some minimum charge rate for the Lithium cells?

    Only from a practical standpoint, but yes I'd shoot for 0.05C as the minimum. However you should still charge according to your load requirements over time and solar insolation to make sure you don't exceed 80% DOD. I personally wouldn't risk puny currents keeping me skirting 80% DOD.

    Also know that if you are charging at around 0.05C, you basically will have no absorb, so in that case you'd set your voltage to no more than 3.45v x cell. (In the case of the teeny 20ah GBS cells, that is actually 3.38v).
  • Blackcherry04Blackcherry04 Solar Expert Posts: 2,490 ✭✭✭
    Re: Why not LiFePO4?

    Here, I chopped it off to save space, as if knowing it makes a lot of difference. What it does show is the capacity fade is about the same as FLA batteries. How many amp hrs @ 12v do you think it takes to make a pot of coffee ( 10 Cups ) and keep it warm for 2 hrs ?? Hint, using your charging / capacity scheme it's a $500 coffee maker @ 50% dod.

    Attachment not found.
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    You have to keep the test conditions the same. Show me a mobile coffee-cart running from FLA batteries, running at 0.5C, down to 100% DOD by the time the day is over, all the while getting 2000 cycles out of it.

    You know we're getting nowhere. All we are doing now is bench-racing specs, desperately trying to find an edge case to prove a point and win an argument based on that, rather than exploring the possibilities of using lifepo4.

    Stick to lead - that's fine by me. I still have some myself and have no intention of throwing them out the door and *healthy* skepticism is fine. What you won't find me doing is cheerleading lifepo4 in the lead-based threads adding to the signal to noise for those who have no intention of running it.

    Quite frankly, when I set out with an exploratory large prismatic lifepo4 battery, I did so from a negative standpoint to prove to myself that they were bunk. But once I got some hands-on, that standpoint did a total 180 degree turn. It will for you too, it will just take a little longer than expected. :)
  • PhotowhitPhotowhit Solar Expert Posts: 5,653 ✭✭✭✭✭
    Re: Why not LiFePO4?

    As I said before, save me money and I'm all for it!

    I do appreciate the pioneers, I purchased a panel and a couple golf cart batteries, supplemented with charging off a van for an extended trip in the Ozarks nearly 30 years ago. I was glad to 'play' and understood it would be more expensive than just having a battery bank charged off the van.

    I do see relative applications where it would be cost effective, but don't feel they have reached the 'off grid' home use, yet.

    Thanks for working with them, perhaps when I need to replace my battery bank, they will have matured and gotten cheaper.
    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • gww1gww1 Solar Expert Posts: 963 ✭✭
    Re: Why not LiFePO4?

    I just heard on the news that tesla is building a plant to build lipo batteries. Maby more compatition and mass production will reduce the price.
    gww
  • westbranchwestbranch Solar Expert Posts: 5,183 ✭✭✭✭
    Re: Why not LiFePO4?

    From what I read he is looking at monster storage 'peaker type' battery banks.
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • PNjunctionPNjunction Solar Expert Posts: 762 ✭✭✭
    Re: Why not LiFePO4?

    The batteries we would use (GBS, CALB, Winston, etc) have already matured. :)

    That's the secret!

    The Tesla cells are high-rate cells, the types that we don't use. We're talking 10/15C or more, and work is always going on to extend the life and capacity of those.

    To the relatively low-rate lifepo4 cells we use (.5C to 1C max recommended), with a properly designed house bank of adequate capacity, our usage patterns of 0.1 to 0.3C are more of an annoyance to the cells than anything else. In their current state, they are really over-built for what we do.

    One wants to keep their eye on the ball, erm, the application in mind. You've got guys out there wasting big money on using high-rate cells in low-rate applications, either by not knowing that there IS a difference, or they just purposely under-size their capacity for whatever reason.

    To justify large prismatic lifepo4's over lead now, in a housebank application, means that one should look at ALL the positive characteristics, and see if the current price justifies those characteristics. If lead works, then that's ok too!

    You may be waiting forever for advancements in EV battery technology to improve, but for us, the relatively low-rate cells have already been around for 15 years, but only commonly available to us for about 8 or so.

    I'm just saying - if you keep your eyes on EV high-rate cells, you may wait longer than necessary if lifepo4 interests you, since in each case either the rate, or the sub-chemistry has nothing to do with ours.

    A silly example would be like an operator swearing off flooded cells and saying that in all cases agm is better, only to find out that the flooded battery he chose for deep-cycle was in fact, an SLI type. We spend a lot of time trying to get newbies to know the difference here, but when it comes to li-ion, most mistakenly lump their characteristics all together. :)
  • karrakkarrak Solar Expert Posts: 326 ✭✭✭✭
    Re: Why not LiFePO4?

    We have been using LiFePO4 batteries in our off grid system for 17 months. The batteries replaced some ageing NiCd batteries which were around 70% efficient.

    Regarding pricing, i did a costing of our batteries here, http://forum.solar-electric.com/showthread.php?19079-Alternative-to-Lead-Acid-Batteries&p=178256#post178256 . Although the upfront capital cost of the lithium batteries are higher than lead acid batteries, here in Australia the overall cost of the both types of battery work out in the same ballpark assuming 3-5,000 cycles from the lithium batteries which seems plausible, only time will tell.

    Regarding safety and ability to handle accidental abuse i think lead acid batteries, especially FLA batteries have the advantage here. With overvoltage, the lead acid batteries split the water in the battery into hydrogen and oxygen which either recombine back to water or the water can be added if the batteries are not sealed, the Lithium batteries don't have this 'safety' mechanism and overcharging will end up heating the battery, damaging the electrolyte and structure of the battery and if taken too far can result in the venting of the flammable electrolyte which can catch fire. Discharging the batteries too far will cause damage to both types, my perception is this damage is more severe with lithium batteries. With the cost of monitoring electronics coming down and now being only a small part of the total system cost I think it is now reasonably easy to guard both types of battery from this sort of abuse.

    The big positive we have found with lithium batteries is convenience. With the old NiCd batteries we could never be sure what the SOC of the battery bank was, with the new lithium batteries due to their high current efficiency (around 98%) we just have to keep track of the current going into and coming out of the battery to know the SOC. The fact that lithium batteries are not degraded by not keeping them full means we don't have to oversize our solar array or have a generator to keep them full and don't have to worry about this. The lithium batteries are also maintenance free, no having to check and add water, and no corrosion of the terminals. Having said that, the lithium batteries should be monitored and checked. Even if the lithium batteries only last five years and thus end up costing more than lead acid batteries would have, the convenience of them makes me think I might buy them again.
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • PhotowhitPhotowhit Solar Expert Posts: 5,653 ✭✭✭✭✭
    Re: Why not LiFePO4?

    It appears you have 4 strings in parallel?

    I haven't heard if this is a problem with LFP?
    Home system 4000 watt (Evergreen) array standing, with 2 Midnite Classic Lites,  Midnite E-panel, Magnum MS4024, Prosine 1800(now backup) and Exeltech 1100(former backup...lol), 660 ah 24v Forklift battery(now 10 years old). Off grid for 20 years (if I include 8 months on a bicycle).
    - Assorted other systems, pieces and to many panels in the closet to not do more projects.
  • karrakkarrak Solar Expert Posts: 326 ✭✭✭✭
    Re: Why not LiFePO4?
    Photowhit wrote: »
    It appears you have 4 strings in parallel?

    This is not of my choosing, it is a combination of supplier stuff-ups and time constraints. I will probably start an 'discussion' :-) by saying my preference would be for 2P, mainly because my experience of other battery technologies is that individual cells do not all fail/age at the same rate and if you have a failure of one cell you don't loose the whole battery.
    Off-Grid with LFP (LiFePO4) battery, battery Installed April 2013
    32x90Ah Winston cells 2p16s (48V), MPP Solar PIP5048MS 5kW Inverter/80A MPPT controller/60A charger, 1900W of Solar Panels
    modified BMS based on TI bq769x0 cell monitors.
    Homemade overall system monitoring and power management  https://github.com/simat/BatteryMonitor
     

  • SkykoSkyko Solar Expert Posts: 121 ✭✭✭✭✭
    Re: Why not LiFePO4?

    Will you get 3000 to 5000 cycles out of LiFePO4 in the warm temperatures of Australia?

    I have read somewhere that they age faster in higher temperature environments (but so does lead acid).
  • zonebluezoneblue Solar Expert Posts: 1,218 ✭✭✭✭
    Re: Why not LiFePO4?

    I got the impression that if you have to parallel them, you do it at the cell level. I suspect the reason for this is that it minimises the cell count for monitoring/balance purposes if you go that route. Lots of fuses though.

    Agreed, total lifecycle cost is now about the same as AGM. Other advantages mean LFP is now ahead, so dont see what all the fuss is about. Availability is the thing in my mind, once they are easier to get theyll take off.

    Sounds like a neat motorhome upgrade project. Will look out for a small set to pair with the upcoming brat controller, whatever i can get under the 400 dollar customs threshold.
    1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
    http://zoneblue.org/cms/page.php?view=off-grid-solar


This discussion has been closed.