Announcement

Collapse
No announcement yet.

Lithium Battery Another Look

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Lithium Battery Another Look

    Well if you want to save time to get to the bottom line, Lithium is not ready for Solar with the exception of a very small scale systems mainly for toys. But if you want to know more and the challenges read on.

    Face the facts lithium batteries are very expensive and do not have the cycle life of lead acid batteries. Not only that are fairly dangerous with high fire risk involved which require some very complex Batty Management Systems to monitor the batteries for charge and thermal management. Coupled with short cycle life, high expense of batteries and BMS they cannot compete with lead acid batteries. Well that is the way it use to be and still holds true today with one exception of the LiFeP04 (aka LFP) lithium family of batteries that deserves a second look and holds some promise in the future.

    LFP or lithium Iron Phosphate uses a iron phosphate crystal on the cathode of lithium batteries. I won't go into all the details because it is not important to understand. What is important to know is it makes the LFP battery fairly stable or safe and does not require expensive Active BMS and can use rather inexpensive Passive BMS and no thermal management is required. More on that in a minute. The other important factors are its lower manufacturing cost and cycle life. Manufactures claim up to 2000 cycles which has yet to be proven, but if holds true is about on par with good quality Lead Acid batteries. The other advantage is significantly lower Internal Resistance which means efficient charge/discharge eliminating Peukert Effects to the point it can be ignored, and enables very high charge rates up to 1C and discharge rates of continuous of 3C and burst up to 10C for short periods of times measured in seconds. That is huge for the DIY Electric Vehicle Markets. Another huge benefit is the battery can operate without loss of cycle life between 20 to 100% SOC where lead acid is limited to 50 to 100%. They can also operate in PSOC ranges which are no good for Lead Acid Batteries. All PSOC means is the battery dies not require it to be fully recharged after every use to extend battery life. It can spend days below 100% days down to 20%.

    Now the down sides. They are still sensitive to over discharge. Go below 20% and they become a Brick. All it takes is one mistake and they are toast, very expensive toast. They are not as sensitive to over charges as the other Lithium chemistry in that they can take some over charge without bursting into flames, but they can still be turned into bricks from over charges. So with that said some form of BMS is required to protect your investment, most notable protection from over charging which can be done with Passive BMS. Cost is still problematic for LFP. Chi-Com made LFP's are in the range of 40 to 50 cents per watt hour compared to 18 to 22-cents for good 5 year FLA battery and about the same as AGM. Some of that cost can be offset because you can use a smaller LFP AH than FLA. It takes roughly a 70 AH LFP to equal a 100 AH FLA usable capacity. Having said that the Chi-Com LFP quality is questionable at this price range. There are other quality Nano-LFP batteries out there made by companies like A123 but will cost you in excess of $1/wh or 5 times what a FLA will cost you with no real added benefit.

    OK now that is out of the way, can you use LFP today. Well the answer is sort of on a very small scale toy size. Why do I say that? Because there is only a single manufacture that makes a Solar Charge Controller for Lithium batteries. That company is GenSun and they only offer 3 LFP controllers. One is made for a 12 volt Kyocera 140 watt battery solar panel (10 amps), 60 watt 12 volt battery panel (5 amps), and a 36/48 volt battery for golf carts using up to a panel with 8 amps under 40 volts Vmp. Not much to work with yet. Why you ask? Its the market folks as there is no demand yet for Lithium solar charge controllers. That is just the way it is for now.

    But what if you just cannot wait for a manufacture to make lithium compatible controllers. Well the answer is you had better know WTF is going on, otherwise stop reading because this is above your head and you would be playing with fire. So read at your own risk. LFP can be charged using either CC or CV method. I will not define what that means because if you do not know, you should not be reading this and playing with fire. A LFP cell is considered fully charged using a CV method of 3.65 volts and the current tapers down to .05C. In a solar system there is no reason to charge faster than C/4 although it can go as high as 1C with thermal management using a BMS. Armed with that you should be able to figure things out by modifying a conventional charge controller. With 3.65 voltage would mean you need a simple Float Type charger set to 14.6 volts if you want to go to 100% SOC. You wil have to figure out a way to shut off the charger when current trickles down to .05C.

    I also know one person is going to argue this but to get to 100% SOC you are going to have to use at least a Passive BMS. A Passive BMS is really nothing more than a simple circuit board attached to each cell that turns on when the battery reaches 3.65 volts and puts a BLEEDER RESISTOR across the cell to bypass and discharge the cell and remains turned on until the voltage falls below a set level. The cost of such a board runs about $12 each and will accommodate a 50 to 120 AH cell. Passive BMS is a different animal that is used with EV's where it takes power from higher energy cells and transfer it to lower energy cells. much more expensive and complicated.

    The other PITA ass with using Passive BMS, or even if you do not use a BMS is you must initially balance all the cells and periodically balance the cells. After several cycles the balance grows to the point that forces you to balance them and you run the risk of destroying the batteries if you do not monitor them.. Remember I said just one over discharge below 20% SOC will turn it into a BRICK. Without a BMS and monitor you run a very high risk of doing just that.

    OK I think that is enough for now.
    MSEE, PE

  • #2
    Originally posted by Sunking View Post
    Well that is the way it use to be and still holds true today with one exception of the LiFeP04 (aka LFP) lithium family of batteries that deserves a second look and holds some promise in the future.
    I'm so glad you pointed that out. LiFePo4 differs significantly from other other lithium chemistries, yet most of the fear and bad reporting comes from abusing the other types such as LiCo02 etc, or gray-market factory rejects. Small LiFePo4's are used under motorcycle seats. BUT you'd be a fool to use any other chemistry than lifepo4!

    Now the down sides. They are still sensitive to over discharge. Go below 20% and they become a Brick. All it takes is one mistake and they are toast, very expensive toast.
    A properly designed low voltage disconnect will help here from going beyond 20%. However if they do, what bricks them is not doing a proper recovery charge *in time* - otherwise at very low voltages the battery is eating itself. You must get to them asap, and only apply about .01C charge current until a cell reaches about 3.2v, whereupon you can apply the normal charge rate.

    If you wait too long, or apply too heavy a recovery charge, they swell and brick that way. On the smaller scale, like a lifepo4 under your butt in a motorcycle, an Optimate Lithium charger knows this, and automatically applies a very small recovery current, all the while watching the internal resistance before deciding if the battery is bricked. Again, it all depends on how fast you get to the battery. Of course on a large scale, this would mean a substantial power supply as long as it can be programmed for no more than .01C recovery current.

    I must admit I have NO knowledge on how much of a hit you take in cycle life if you do this often. I have personally recovered a few well below 5% SOC, but it might take me a few years to figure out the damage, even though I was able to recover them. So they aren't exactly bricks since I got to them in time and did a recovery charge properly, but ideally I plan to learn from my mistake and never go there again.

    There are other quality Nano-LFP batteries out there made by companies like A123 but will cost you in excess of $1/wh or 5 times what a FLA will cost you with no real added benefit.
    That's correct. The A123 cells, are *high rate* cylindrical cells, and you pay dearly for that. The type appropriate for our use, like CALB, GBS, etc are not super high rate, and you get more bang for your buck. Typically they accommodate 1C, and in our application, nobody would even come close to using 1C for a significant solar bank if they design it right.

    In fact, for our usage, we are in what is known as a "Sub-C" application, and that has some significant benefits.

    Why you ask? Its the market folks as there is no demand yet for Lithium solar charge controllers. That is just the way it is for now.
    My Morningstar Prostar-15 pwm CC has been modified for no temperature-compensation (I merely pulled the temp-comp transistor). With only 3 preset voltages, I choose the "GEL" voltage of 14.1v for my nominal 12v lifepo4 GBS batteries. For a true 100% SOC charge, that would be 14.4v, but I don't want to go there. Limiting the charge to only 14.1v (3.525v per cell) is about 90-95% SOC is good enough, and gives me some unbalance headroom, if one considers 3.6v per cell the limit.

    Genasun limits their bleedoff balancing at 14.2v (3.55v per cell), so I'm pretty close. In fact, if I were to use a Genasun, I'd want to take advantage of their ability to reprogram the voltage you choose to something lower anywhere from 13.9 to 14.1v.

    The point here is that as long as you disable temperature-compensation, one doesn't actually need a "lithium" charge controller if you pick your voltages yourself. Thing is, I'm sure a Genasun would survive a boat-race around the world, whereas my Morningstar might not in that environment.

    Well the answer is you had better know WTF is going on, otherwise stop reading because this is above your head and you would be playing with fire. So read at your own risk. LFP can be charged using either CC or CV method.
    I agree - if it seems too much, just stick to lead-acid. It still works, and will serve you well when done right.

    You wil have to figure out a way to shut off the charger when current trickles down to .05C.
    It won't blow up, it is just that there is no need to keep charging any longer. Fortunately either the onset of night will take care of that, or one can do a workaround by using a controller that goes into float at no more than about 13.7v, lower is ideal.

    While there is no need to float lifepo4, you CAN do this at these voltages since there is no current flowing when the float voltage is actually lower than the totally charged battery voltage. And, by not being held at the high absorb voltage no damage is being done. It can actually act as a safeguard against small parasitic loads you may not be aware of.

    I also know one person is going to argue this but to get to 100% SOC you are going to have to use at least a Passive BMS. A Passive BMS is really nothing more than a simple circuit board attached to each cell that turns on when the battery reaches 3.65 volts and puts a BLEEDER RESISTOR across the cell to bypass and discharge the cell and remains turned on until the voltage falls below a set level. The cost of such a board runs about $12 each and will accommodate a 50 to 120 AH cell. Passive BMS is a different animal that is used with EV's where it takes power from higher energy cells and transfer it to lower energy cells. much more expensive and complicated.
    Um, that must be me. I prefer single-cell charging, and then assembling into a pack myself, yada yada. However I could easily take a middle-ground, and use the passive cell balancers (typically they kick in at 3.55v) for a few rounds. If I felt that they presented a point of failure in themselves, one could merely remove them after they are satisfied that the batteries are balanced, and then merely run the bank at a lower charge point, like 14.1v (3.525v per cell for example). A few ways to skin this cat.

    After several cycles the balance grows to the point that forces you to balance them and you run the risk of destroying the batteries if you do not monitor them.. Remember I said just one over discharge below 20% SOC will turn it into a BRICK. Without a BMS and monitor you run a very high risk of doing just that.
    Yes, by all means monitor them. However, unlike the EV's that these cells were basically intended for, we are running in a "Sub-C" application, and they tend to stay in the state of balance you last left them in. Actually, if you run through enough cycles, you will witness what some would call a "self-balance" of sorts if you drive them into into absorption. But that takes many cycles. Again, on the low end, an Optimate Lithium charger does this by oscillating the charge into and out of absorption many many times automatically.

    And, without getting too far into it, on paper there is no such thing as self-balance. However in practice when you drive the cells into absorption, because each cell has a slightly different manufactured capacity and internal resistance, over time they will tend to balance up. Not exactly close, and I'm not recommending just slapping these cells together without an initial reasonable balance. Here again, not driving these to 100% SOC, and instead shooting for say 90-95% SOC instead with a little headroom may ease the obsession over balance.

    LiFePo4 is not for everyone, and I cannot recommend them as a universal replacement for lead-acid. However, the more you know about the differences in their chemistry from the other types, and also the difference in our application (sub-C), it may help some move beyond the hype - both pro and con, and make a more reasonable decision based on knowledge and not fear or advertising.

    Sorry if I went too far, but I feel that LFP operations can be managed in a few different ways depending on your risk factor - like anything.

    Comment


    • #3
      Originally posted by PNjunction View Post
      Um, that must be me. I prefer single-cell charging, and then assembling into a pack myself, yada yada.
      Yes it was you I was referring too, and it was out of respect. I am glad you chimed in. My whole point both from economics and technically LFP is not ready for prime time. The only affordable LFP are the Chi-Coms as you noted like Calb, GBS, Sinopoly, and so on... Just a few short years ago all those batteries were made under a different names like Thundrsky and the quality was so horrible they had to change their names. Point I am making here is I do not trust the SOB's as they already got caught red handed. Something does not add up if you claim 2000 cycles and only a 1 or 2 year warranty. However I will admit I am now an owner of 32 GBS 100 AH cells sitting in a modified golf cart, and man is she fast, scary fast for a vehicle with no real brakes.

      I know we disagree about BMS, and that is OK. You and I can get away with no BMS, but John Doe public cannot, and that is where you and I differ on opinion. For LFP to go mainstream in solar a BMS is going to be required to protect the investment. Economics and proven reliability are not there yet. Real life comparison comparing Apples to Apples; a 48 volt 100 5 year FLA battery will cost you $950 to $1000 and you get a real 3/7 year warranty backed up by a company that has been around a long time. A 16S 70 AH LFP (48 volt @ 70 AH) will cost you $1600 to $1700 and comes with a questionable 2 year warranty from a Chi-Com company only in biz a couple of years, and known to have committed consumer fraud under a different name. You don't have to be an economics professor or an engineer to figure out which product to buy.

      So until the economics and trust issues are resolved, LFP is not going to go mainstream anytime soon. On the technical side of charge controllers is not a problem as it can be done today and is less complicated than Lead Acid chargers. There is just no demand for it. Demand will be there when the economics and trust issue is resolved. Unfortunately no USA or European company can really get into manufacturing of LFP batteries as of yet. No one can compete with Chi-Com and India companies.
      MSEE, PE

      Comment

      Working...
      X