Best LifePo4 charge controller settings known to man for Maximum Service life and Minimum battery stress!!! 5,000-10,000+ cycles?
- Thread starter Go2Guy
- Start date
Ampster
Renewable Energy Hobbyist
I believe you can get at least full 2500 cycles or 5000 half cycles from Grade B cells with a reasonable use pattern. Perhaps there are differences of opinion about what "reasonable" is? I do not have long term data to prove my assumption but having a six year old EV with less than 10% degradation gives me hope.
You are confirming my point. You don't need to baby your batteries and try to squeeze every last cycle to get 2,500 cycles. This is 7 years of reasonable use to get to 80% capacity.
Ampster
Renewable Energy Hobbyist
Yes, in one of my iterations I used seven year old Nissan Leaf modules that still had 80% left. I used them for another two years and by then it was such a Frankenpack because of the modules that I added. I randomly tested a few modules and was surprise at how much capacity they had left. I recently sold them to someone who was going to add some to an RV and others to some electric scooters.
Curious if you have you studied any of the manufacturers spec sheets? Typically, cells are rated for at least 2000 cycles at full C rates. EVE makes high cycle claims depending on C rates.
If I only got 1000 cycles out of a LFE cell I would be severely disappointed. Granted the cells don't need a lot of babysitting but the consensus is keeping the cells voltages between the knees is one of the best ways, if not the best way, to extend cycle life. Of course ideal temps also enters into the equation.
curiouscarbon
Science Penguin
- Joined
- Jun 29, 2020
- Messages
- 3,017
cell above 3.0 V
cell below 3.65 V
avoid long time above 3.45 V
cell above 0 C
cell below 35 C
this recipe seems to be decent to get plenty of capacity and life from LiFePO4 cells
cell warming up too much seems to be a factor in early cell degradation
cell below 3.65 V
avoid long time above 3.45 V
cell above 0 C
cell below 35 C
this recipe seems to be decent to get plenty of capacity and life from LiFePO4 cells
cell warming up too much seems to be a factor in early cell degradation
Ampster
Renewable Energy Hobbyist
One of the factors that can create warming is high C rate charges or discharges.
sirslayer
New Member
- Joined
- Mar 20, 2021
- Messages
- 73
I said 1000 cycles because of the constant charging of 3.65 volts per battery. I believe even at that voltage, my lithium phosphate should last longer then that. I do drain them down to 2.7 volts and they will last for many years . I think most folks do baby your batteries and I’m struck that most folks are not using their batteries to the max to save money and let them slowly die from shelf age.. well at least they still look new …
Ampster
Renewable Energy Hobbyist
I resemble that comment about shelf age but I do try to stay active and not spend much time on the shelf. I hope to outlast my batteries.
It all depends on where you are standing.
I happen to work with laboratory tools and proper calibrated tools are very important. And after all checking was done all necessary terminals and connection inspected it goes through cycle discharge and charge while continously recording data on individual cells resistance and voltage with total pack voltage and resistance. And for residential use this was usually in 60% of max A bms is designed for if mosfet. Residential laboratory test are always done at 75°F.On EV battery packs there are different types of laboratory tests that include various different temperatures.@Go2Guy - I hate to speak for @Steve_S (he does it pretty well himself) but I think the three main points he made - if I were to reduce them to bullet points - are:
I'm sure if I didn't get that right Steve will correct me.
- Temperature matters, because colder cells will accept current (absorb energy) slower than warmer ones. This can be important when you have more than one parallel battery being charged.
- With this chemistry, millivolts matter. In the case of AGMs or any other lead acid, you can get things to the nearest 1/10th of a volt and it's probably OK. Not so much with LiFePO4.
- Recognize that there are voltage drops throughout your system, even with large wires. Because of #2 (millivolts matter), you need to know and account for those drops from your SCC and Inverter/Charger to your battery at different charge currents and as the charge cycle finishes.
I have seen many YouTube videos where people don't really take good measures for voltage and resistance under load and also imagining thermal camera for any temperature raises because of weak terminal conditions. Calibrated volt meters that can measure three decimals are very important for lifepo4 chemistry. Discharging voltage drops even not very important are necessary to lower resistance. But for charging it is very important to adjust for mV drops across wiring and compensate this with MPPT.
Mining for all the metal ingredients that make up all kinds of batteries is harmful to the environment as well.
We all gotta pick our poison.
Absorb time can be easily figured if you have run battery cell charge and discharge tests.
Example if you are after 3.330 to3.450 V absorb will be necessary if you want to bring lifepo4 chemistry to max energy at this conservative voltage level.
Also many need to understand that with this chemistry relying on flat voltage curve Voltage is not possible to determine +/- 3.000-3270 depends on AH battery cells . Columb meter (AH meter) would be necessary and one that highly accurate. But even then there should be algorithm in the BMS to know when battery was not used for quite some time ( Charging ot discharging) temperature between parallel banks to compensate for battery drift. Highly experienced members here can live with only good balancer, considering they have calibrated voltage tool to measure mV and low resistance quality tool. Regarding putting multiple batteries in parallel and hooked to side terminals make sure you feed one wire from the top post terminal and on another post from the bottom to the energy user's.
Example if you are after 3.330 to3.450 V absorb will be necessary if you want to bring lifepo4 chemistry to max energy at this conservative voltage level.
Also many need to understand that with this chemistry relying on flat voltage curve Voltage is not possible to determine +/- 3.000-3270 depends on AH battery cells . Columb meter (AH meter) would be necessary and one that highly accurate. But even then there should be algorithm in the BMS to know when battery was not used for quite some time ( Charging ot discharging) temperature between parallel banks to compensate for battery drift. Highly experienced members here can live with only good balancer, considering they have calibrated voltage tool to measure mV and low resistance quality tool. Regarding putting multiple batteries in parallel and hooked to side terminals make sure you feed one wire from the top post terminal and on another post from the bottom to the energy user's.
My cells cycle between 3.5V and 3.0V, with a 400ah pack i have regularly discharged 300ah from fully charged during a cycle.
They have never been charged when over 35°C.
They are approaching 11 years old now with no discernible difference in performance from new.
In that time i have seen many others using more aggressive parameters (especially disregarding high voltage/temps) have their batteries fail.
As for shelf life - i know of packs from 2007 that are still going - shelf life is unknown as it is untested in fractional C LiFePO4.
You can ignore the actual real life experiences of those that have been using these cells for a decade, and find out for yourself how to make a cell that should last for 15 years now last for 6 years.
Are you running cells in parallel and how many?
I use 2 cells in parallel to make 400ah
squowse
Solar Enthusiast
- Joined
- Jan 4, 2021
- Messages
- 605
Great post and guide . ?---ORIGINAL TITLE----Best charge controller settings to achieve 10%-90% usage on lifepo4 ?
EDIT-UPDATE and the ANSWER to this question. . . This post is a accumulation of all the great and wonderful information given to me by all the great people on this thread. I consolidated all of this into this post and is the result of many hours of research and consulting with the people of this forum. Again thanks every body!!!!! This would not be possible with out you guys. if you believe i missed something or want to amend something please comment on this thread but please do so in a way that can be consolidated and added to this post , . . this is really not my post per say but an accumulation of the knowledge of the people of this forum and thread. .
BEST SOLAR SETTINGS [SO FAR] FOR MAXIMUM LIFE 5,000-10,000+ cycle life
PREFACE The Best settings possible are actually to run from 14%-90% SOC and this has little to due with voltage [besides staying out of the high and low voltage knee] however this is not possible with out a columb meter,or even better a charge controller that operates off of this principle. . Most charge controllers are designed for lead acid and being that the case we have to enter in voltage numbers numbers, so even tho this is not the ideal way to go about it, its the only way for most of us with cheaper and/or older charge controllers. . Here best voltage numbers possible known by the great contributors of this group to achieve the maximum service life possible and put the batteries under as little stress as possible.
BULK/ADSORB 3.45-3.52 [heavily debated] but universally agreed no higher than 3.52 for max service life [3.52 is what victron recommends for max service life
[lower voltage= slower charge times IE [C rate] which is Good but potentially longer in absorb phase which is Bad]
[Higher voltage=faster charge times which is more stressful on cells, BUT less time in absorb phase which is good ] 3.5v-3.52 seems to be a good compromise fast and slow charge times NOTE: Lower voltages means charge rate will be reduced at high SOC and it will be harder to reach 100%. which is great if your shooting for less than 90% state of charge. !!!IMPORTANT NOTE: While the above mentioned data is recommended for minimum stress on the cells it is important that the voltage you set activates your BMS, to keep your cells balanced, if not your BMS becomes useless and will never actually balance , most BMS activate around 3.4 VERIFY at what voltage your BMS activates at . if you are not sure, it just may be better to higher voltage to ensure that its activated[Will Prowse goes all the way to 3.6v OR another option is to keep an eye on it and make sure the cells stay balanced.
FLOAT 3.35
3.35 volts is a good float voltage for high reserve capacity and minimal cell
degradation [3.35 is not really debated if maximum service life is the goal 3.4 if you want a little bit more reserve. Also 3.4 is recommended by many OEMS, however according to some data at 3.35 the cell is not really in a state of stress while anything higher than this it is. . your choice a little bit more conservative or a little bit more reserve ]
CUT OFF 3.1V [but a more ideal verifying the cut off voltage is is to run down battery pack under average load you will be running and then take note of what cell voltage drops off first, once the first cell drops off in voltage , measure the total pack voltage and set low cut off voltage to that measured total pack voltage.
NOTE: In certain high load situations you may get get voltage drop causing the system to "cut off" and shut down under high load even if its not necessarily towards the end of your desired capacity. . , if this is the cause. you may have to set the cut off lower voltage. . 3.1 is a conservative number 3.0-2.9 is recommended for high load situations, but remember this post is all about having the data to make educated choices if your goal is maximum service life and achieving a balance between functionality and maximum service life
OTHER NOTES:
TEMPERATURE-"""storage at high State of charge and high temperatures promoted such severe losses that the cells in question were unable to recapture capacity that they had lost reversibly""".DO NOT STORE IN HIGH SOC especially in HIGH HEAT, i dont have all the data yet but it seems operating under 32F or over 100F could/will causes damage[especially under 32] as a good precaution in short if your not comfortable the batteries are not comfortable, [this is a super generic but a safe conservative recommendation] As solutions to this conundrum people will run thermostat controlled heat pads and heat blankets for the winter [search the forum] and use water tanks/pads or ac for high temp operation [search the forum] .
Why temperature of lithium cells is important
Temperature is an important parameter that influences the aging of a lithium battery. At 35°C a lithium cell ages twice as fast as it does at 15°C. So if the battery’s useful life is 15 years at 15°C, it will be 7.5 years at 35°C. I summarize...diysolarforum.com
The Degradation Behavior of LiFePO4/C Batteries during Long-Term Calendar Aging
Abstract: With widespread applications for lithium-ion batteries in energy storage systems, the performance degradation of the battery attracts more and more attention. Understanding the battery’s long-term aging characteristics is essential for...
VOLTAGE LOSS Every connector in your system as well as wire length attributes to a certain amount of resistance and which results in voltage loss. in other words If you set the charge controller to 3.5 volts the battery may only see 3.2 due to this voltage loss You have to adjust the controller to compensate for the loss. you may also have different losses for different components in your system depending on where there located for example inverter is on a longer length of cable than where the charge controller is located.
HOW TO RESOLVE use a good quality meter . . .measure the terminals at the battery pack , measure the terminals at the controller or inverter, subtract the difference and add this difference to your charge controller or inverters settings. [when testing for charger controller you want to be under charge, for inverter you would want to be under discharge]. measure and test again to ensure this was sufficient compensation
STORAGE, batteries should be stored at about 40% SOC if going to be in storage for long periods of time. . high SOC storage causes premature cell degradation
DON'T BE AFRAID to use the battery at 100% if needed, calendar aging [definition, loss of capacity due to time alone] [thanks will for this input] also plays a role in long term degradation so if even if you baby these cells to the max, with old age you will still loose capacity. .so these numbers above are to baby the cells, but it is perfectly okay to use full 100% soc when you need it , when taken into consideration that the effect that calendar aging has on the battery. Its is a balance between the battery is loosing capacity on its own due to aging and you babying or not babying the cells , because of calendar aging some say who cares and just use it at 100% at all times , however that defeats the purpose of this post of attempting achieve maximum service life. . but it definitely doesn't hurt to use it when you really need it. .
Balance--(run batteries to a high SOC to trigger a good balance on your BMS 4 months or so, -as a maintenance. think of as a equalization ]
A great read by Joey on taking care of your cells https://diysolarforum.com/resources/how-charging-works-in-the-context-of-lfp-batteries.233/download
END POST
______________________________________________________________________________
ORIGINAL OPENING POST
i have spent about 6 hours reading thru google and and almost every relevant search result this website has as well as youtube. . and and my brain is on information over load. . I have a simple question, but i can only seem to find complicated answers and debates between people. I have several charge controllers and all in one units i will be setting up soon for family members and im looking to find what are the best settings to achieve a usage of between 10%-90% SOC on lifepo4 batteries. . . from everything i read it seems this is the safest for long term reliabilty [10-15 years or more if possible] . . but cant find info how to set the settings to achieve this .. . . . as we all know almost all charge controllers are designed for lead acid, so we are stuck with programming such parameters and have to make do with float and absorption settings. . what im looking for is the approximate voltages settings on a per cell basis for the following
bulk
absorption
float
low voltage cut off
also any other settings you recommended changing for lifepo4. .
should these numbers be different when under load?
any other considerations when using mpp units ?
the most straight forward answer i found was wills on his site, [ https://www.mobile-solarpower.com/diy-lifepo4-solar-battery.html ]
however his recommendations were based on 100% SOC which i prefer not to use since were looking for extra long life with this batteries ,so i prefer to stick to 80% however his post was the most straight forward answer of the 6 hours i have been searching online . . essentially looking for same answer but with 80% in mind. .
Only thing I would've add is that voltage drop is proportional to the current. No current = no voltage drop.
Any imbalance with Parallel cells so far that you have experienced?
It is impossible to get imbalance with parallel cells. What you can get is different characteristics between cells that reduce the capacity of the stronger cell in the pair.
Cell balancing is surely one of the most misunderstood aspects of LiFePO4 battery construction. The amount of misinformation i see on this forum is staggering (especially with respect to parallel cells)
I do understand that but how often do you check parallel cells for integrity especially if they are going to be used over long-term. At 500+ Ah i usually run separate resistance meters that track and watch overitime for voltage drift or resistance imbalance between multiple parallel cells. So far my highest parallel creation was 1000 Ah per parallel setup ( usually 200 Ah 5 in row).And they get out every 2 years for benchtop observation and if passed going back to the battery battery bank. Every battery manufacturer and battery chemistry is different so starting at the beginning takes some learning how to read and find cells in parallel that need attention. 50+kwh 48V setup is easy to manage . Just need to use common sense and experience when playing with this types of energy at one bank setup.
I don’t have anything capable of separating out the busbar resistance from the IR of individual cells. I’d like to see the seperate resistance meters you use.
If a parallel pair show up a discrepancy, i will remove them both and check them. Until then i treat them as one cell.
At one stage my battery went over 6 years without me even looking at it - no way i’m taking it apart every 2 years ?
squowse
Solar Enthusiast
- Joined
- Jan 4, 2021
- Messages
- 605
As they are in parallel, the cell voltages should balance out and both cells should discharge fully?
squowse
Solar Enthusiast
- Joined
- Jan 4, 2021
- Messages
- 605
You can put a voltage meter over the item and measure the voltage drop. then R = V / I. Will not work for voltage sources like cells - you would need to use the dV/dI method for those.
If you have 2 cells in parallel connected by a very low resistance busbar, please explain what sort of voltage meter will seperate the IR of the two connected cells?
For sure measuring the IR of the combined cells is trivial, samcat was saying he is able to measure individual cell IR of cells in a parallel - that’s what i’m interested in.
squowse
Solar Enthusiast
- Joined
- Jan 4, 2021
- Messages
- 605
Sorry. I read "busbar resistance".
Sorry if someone didn’t understand my post. I do remove bus bars for individual cell internal resistance.....for internal resistance under load or charge i use all cells in parallel with known values for each cells. It is just necessary to know where leads are applied on the bus bars and need to be permanent for any future checking to eliminate any difference in measurement. Usually if voltage is higher than single cells in parallel that means that my resistance meter is reading something else than set of cells in parallel. I will post later on how this is done with minimal errors when in the service. One thing we all know that when checking resistance between two meter leads, meter will try to measure closest paths between the leads and resistance will try to find path that is easiest to connect two leads when measuring for resistance.
Similar threads
