 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Solar : 48volt flat packs
  Page 5 of 5 |
|||||
| Author | Message | ||||
| Warpspeed Guru  Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
I have read that mini cycles are very bad for lithium life, they should be fully charged, then either discharged or left to settle, and then given another charge when appropriate. This happens quite naturally with things like cordless drills for example. There are really two separate issues here, as I understand it, for solar you would be best with one distinct charge charge and one distinct discharge cycle every 24 hours. The other issue is float charging, which is not recommended. Now lithium has a very flat voltage characteristic compared to most other battery technologies, but that is deceptive. Its true that during a continuous charge, or a continuous discharge the voltage does not vary much from start to finish. What they don't tell you is there is a fairly big initial jump up in voltage when charging, and jump down when discharging. So if you are mini cycling, the voltage will be going up and down by a significant amount as the current to/from the battery reverses. The flat voltage part is only during CONTINUOUS charge or discharge. So how can you float that ? Another frightening characteristic is that at the end of charging, the cell voltage rises at an ever increasing rate. It can go up 100mV in tens of seconds. If there is a series string, not all cells "take off" at the exact same point. You need to monitor every cell and cease charging immediately one cell begins to rise in voltage very quickly. I am told that permanent damage occurs above about four volts. Floating cells might be rather risky, if one cell suddenly "takes off" and has a much higher voltage than the others. And you cannot tell by just monitoring total string voltage. I do not know if that is the reason float charging is not recommended in principle, but from my own observation that would at least be one problem best avoided. Much safer to cease charging totally if the voltage across one cell starts going ballistic. Cheers, Tony. |
||||
Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 648 |
Hmmmmm, So you are suggesting one needs two lithium battery packs for off grid use, one being charged while the other is discharged, then packs swapped over the next day? That does not sound very practical to me, besides one lithium battery pack is expensive enough  .I have been doing "mini Cycles" ever since I installed my 200Ah/48v lithium battery. Oldest cells are passing 8 year of age and have about 50% capacity left. Two cells died so far from the mysterious short cell syndrome, replaced those. Unfortunately I took these faulty cells apart to see what went wrong (nothing visibly). I wish I had known about the dumping a capacitor charge trick somebody mentioned here. So the next cell that dares to go short circuit better watch out, I have a 30,000F, yes, FARAD, capacitor here to zap it with  . |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Not really Klaus, The way I did it, the system resets overnight when solar goes totally to zero. At dawn the battery starts charging with whatever energy is surplus after running the inverter. If its a foggy morning, the battery continues to discharge.... Eventually there is enough surplus solar to hopefully fully recharge the battery. The battery charger then turns off (via a flip flop that can only be reset overnight). Only one charge cycle every 24 hours, almost always. During the day solar drives my inverter directly, if its able, if not, any shortfall is made up by the battery. But battery charging cannot re commence once it hits full charge, until dawn the following day. If its a really crappy morning, there may be alternate charging and discharging cycles, but that is pretty unusual. I really do not know, I am certainly no expert. All I can do is tell my story of what I did, what I believe to be true, and what happened to me. Cheers, Tony. |
||||
Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024 |
Distinct charge and distinct discharge would be the most ideal situation to have and very common with cordless tools and a few other's. When charging these 18650 Lions 13s 90p and the other pack 13s 100p to 53.3v is 4.1V per cell some might be 4V some as high as 4.13 my upper limit is 4.15, if any get to 4.15 or more then its time to take action and deal with it, (always had plenty of time to catch it) connect a car headlight to that cell until bulk charge has finished, I have also left the headlight connected to a cell for the rest of the day (got distracted by some other job that needed doing) come back and the cell reads 4V, hasn't been a problem so far, for the few times this has happened. The only time a cell voltage gets a little high is when one of the others are low, this develops over time due to a few factors, older cells different brand and on a rare occasion (so far) a shorted cell in the pack of 90p which has just been a very slow drain over weeks until found and removed. The SCC only measures the overall voltage, it has no idea what the individual cells are doing. when on float 53.2v I have done voltage checks at different times of the day and any mini cycles have only made very small change's in cell voltage, well within the limits, I expect this is because the cell voltage is well below 4.2V and been 90p may also help. Never seen a fast runaway cell voltage, If it did I would have stuffed the whole lot up years ago. There will be a time when it gets worn out and much harder to control Most sunny days all day the battery is mostly bypassed, when I say mostly, I doubt the SCC is fast enough to react to motors starts and other sudden loads. On days with passing clouds, as it has been the last few weeks there would have been a lot of mini cycles which is not good. Now if I can divert all solar power to the inverter after batteries are full charged and block any charge going to the batteries, this would be the preferred way. Power would be taken from the batteries during motor starts if the SCC could not supply enough power, and on days with patchy cloud a lot more power would be taken from the batteries and not replaced until the next day. This can result in a very deep cycle, much deeper than I would be comfortable with, this depends entirely on the weather. This battery setup all ways remains cool when charging and discharging. What I haven't found any info on, is how many mini cycles will do damage to the cells compared to deep cycles. When charging a single 18650 cell or even 3s 1p I have seen the voltage rise faster when around 4.2v if left unchecked, well we know how that will end up. If I go to Lifepo4 battery setup in future, its going to need a lot better management from what I have read, the Lifepo4 looks to be a bit (a lot) more fussy than the Lion 18650 setup I use now, really don't like the idea of an expensive Lifepo4 cell's voltage going off ballistic. For some of the newer battery technology's the cycles may not be as much of an issue, don't know, haven't bothered to look yet, might be interesting to find out. Cheers Aaron Off The Grid |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
Just a quick pre breakfast internet search turned up this: https://enerdrive.com.au/2017/11/29/can-i-charge-my-lithium-battery-with-a-lead-acid-charger/ It shows the sudden very steep voltage rise for a four cell battery. That point where the voltage goes vertical does not always coincide exactly with all of the cells. Three of the cells might be still just beginning to wake up from the flat part of the curve, and the voltage across one particular cell can very quickly rise to damaging levels. Once a cell has "gone ballistic" it takes almost no charging current to keep the voltage from rising even further. Charging must be totally terminated as the article says. I terminate charging at 3.45v for the highest cell, which as you can see on that curve corresponds to 13.8v for four cells. Beyond that it starts to get really hairy. Edited 2020-12-26 06:06 by Warpspeed Cheers, Tony. |
||||
| oreo Regular Member  Joined: 11/12/2020 Location: CanadaPosts: 49 |
If I am floating at 3.45v, do I need to be concerned about hysteresis of charge/discharge?  I would argue that your charge controller needs to be matched to your BMS. As you near the end of the charge cycle, you can't be charging the battery at a higher rate than your BMS can bleed current to control the voltages for each cell. Otherwise you are not going to be able to keep the cells in balance. Nothing should be going ballistic. You want all the cells at the exact same voltage, when floating and when nearing their end of charge. Is this true for all lithium chemistries? Any details on what "very bad" means? Greg |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
They make no mention of charge or discharge rate. The change in cell voltage at a hundred milliamps will be very different to a sudden change of twenty amps ! That is very true. But suppose you are charging at ten amps, and one cell hits the max voltage. You will need to dissipate maybe 35 watts to hold that voltage. That is a lot of heat and power to get rid of on your cell balance board. A much more gentle way is to have continuous cell balancing, so that a small current is bled from the higher voltage cells. That tends to very slowly drag down the more active cells, and give the lower voltage cells a chance to catch up. Much easier to draw 100mA (or 330mW) for 24 hours, than try to have some kind of hard voltage clamping that needs large power devices and heat sinks at every cell. I use a 27 ohm one watt resistor that just runs warm to the touch. Fairly small, and not a lot of heat. Perhaps it needs both! The way I do it, any cell higher than the average of all cells gets its discharge load turned on. The discharger may be on continuously for days or weeks on the highest cell. Average cells cycle on and off fairly frequently. Particularly low cells may never get the discharge treatment. Over a long time span cell voltages definitely do even out. Very noticeable over a month. The difference between highest and lowest cell voltage may be reduced to a quarter of what it was a month ago. I really do not know, it is what I have read form various sources on the internet. It has been suggested that reduced life and long term capacity reduction is influenced by the number of charge and discharge cycles, as well as the depth of discharge. Its very difficult to quantify something like that in an exact way. Edited 2020-12-26 08:44 by Warpspeed Cheers, Tony. |
||||
Haxby Guru Joined: 07/07/2008 Location: AustraliaPosts: 423 |
There was a great paper that I read a long time ago by a gentleman who now works at tesla. I wish I could find the link. His PhD was on maximising the life of 18650 lithium cells, and his test procedure was sound, with thousands of test samples in custom test equipment. I was working for another uni at the time and we were contemplating setting up the same charge/discharge apparatus. What I took away from it, in dot points, was: Don't charge all the way up to the max voltage Don't discharge all the way down to min voltage Once fully charged, avoid resting the cell for a long time before discharging Once fully charged, don't subject the cell to warmer temperatures for any length of time The cc/cv charge concept is over simplified. When batt is low, ramp up the current till batt hits the shoulder of the curve, only then should you charge with the full rated charge current. Also taper off the charge current before hitting the CV part of the charging scheme. Heavy "micro cycles" (eg. Regenerative braking in a car) are fine as long as batt is in the shoulder zone. These rules seem to be backed up in what I have seen in practice on YouTube vids where people post their logs while supercharging. The supercharger never hits the batteries with 50kw or more in one hit. Everything is ramped up/down smoothly. Tesla's have a "pre-supercharge" mode where they cool/warm the batts in anticipation of being supercharged. Ultimately, The batteries are like a little baby. Have to feed them right, don't let them starve, and keep them warm or cool as necessary  The best way for them to last a long time in our case is to have as much storage as we can afford. This keeps them in the butter zone, and means that each cell has less stress applied to it. More cells, less load per cell, happier cell, less energy going in/out per cell. Again this is backed up in the EV world where Tesla's are doing a lot better than other makes when it comes to the longevity of their packs. Tesla's have packs starting at 60-70kwh while the Nissan leaf has for a long time been limited to a 24kwh pack. So naturally the leaf pack will be cycled 3x as hard vs the tesla pack for the same km travelled. Leaf packs are very well known for their fast degradation compared to Tesla's |
||||
| Haxby Guru Joined: 07/07/2008 Location: AustraliaPosts: 423 |
Trusting the company building the cell is another big issue. Has the company even been around for 10 years? Do they plan to be around for another 10 years? I would only trust Panasonic. They have a track record. Any new player will have a hard time competing with a company that has a solid past. |
||||
| Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024 |
Great  This is how I have been treating them. I agree the Panasonic's are good, have been running well for me, there are some others that are doing well so far, Samsung cells have been used in power tools and notebooks. Cheers Aaron Off The Grid |
||||
| Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 648 |
The cells in my battery bank are LYP technology, like Tony's. Here is what the manufacturer recommends: I am charging & discharging them to levels recommended by the AU supplier. There is an active cell balancer which works *all* the time. There are two charging sources,, one is a grid tie inverter working backwards. this dumps unregulated charging current from a max 3KW solar array (2.2KW was the highest I'd seen as the panels face in 3 directions). This charging is controlled by a voltage sensing relay and is terminated when the bank reaches 55.5V. The other charging source is a 60A Outback MPPT. This is set to go to snooze mode at 56.5V. Today I watched the cell voltage monitor as the bank reached fully charged. At that stage the individual cells vary a bit with the best up to 3.65V and the lowest around 3.47V. All are still well below the 4.0V maximum allowable cell voltage. That rather large cell potential difference changes quickly as the active cell balancer (Deligreen) starts shuffling charge. Once the charging is in snooze mode the cells vary from 3.5V (best) to 3.48V, keeping in mind that half the cells are 8 years old and the others only 4 years. The mini charge cycles start when I turn on machinery but quickly stop as the battery voltage rises again. A steady draw, like the A/C seems to maintain the battery voltage at 3.5V/cell, more or less the charger supplying enough current to keep the bank charged while solar input is available. Only time will tell performance in the long run. The two short cells so far were before I bought that commercial balance, was using a home brew version instead. |
||||
| Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 648 |
I seem to be unable to include the hyperlink in the post above, any clues how to do that? I tried the 4th box above but there must be some additional trick to it. |
||||
| Revlac Guru Joined: 31/12/2016 Location: AustraliaPosts: 1024 |
The box comes up with "url text" just enter a name of what you want (link here or something) then ok, the next box you put the full address in, then hope it works. Easy to mess up the copy and past, I missed the HTTPS bit sometimes, check the link works after you post it. Hope this helps. Was that Deligreen? Edited 2020-12-26 20:59 by Revlac Cheers Aaron Off The Grid |
||||
| Murphy's friend Guru Joined: 04/10/2019 Location: AustraliaPosts: 648 |
Thanks Aaron, I'll try that. Forgot about that 'naming' of the link My battery specs: winston LYP battery OK, that worked  Deligreen is the brand name of my active balancers. |
||||
| Warpspeed Guru Joined: 09/08/2007 Location: AustraliaPosts: 4406 |
That third curve "charge capacity" is pretty much exactly what I was seeing. Charging at 0.5C or usually a bit less (say <30 amps for 60Ah battery) the cell voltage stays pretty flat up to about 3.45v, then starts to noticeably rise at an ever increasing rate. I doubt if charging to say 3.5v or more will make a whole lot of difference to stored capacity, but it does start to get a bit hairy as one or two cells might suddenly "take off". As Haxby says, staying safely in the flat part of the curve at both ends cannot be such a bad thing to do. Cheers, Tony. |
||||
| Page 5 of 5 |
|||||
 Print this page |
