So, lets spec out and set up a system of LiFePo4 batteries and inverter

[PNjunction]

The universal problem with any subforum about lithium is that it is too easy to stray off topic - mixing and matching lithium chemistries, voltages, applications and so forth making it nearly impossible to focus.

My suggestion to the board is to create SIX lithium sub-forums since each thread seems to grow into biblical proportions.

Lifepo4 / 3.2v nominal technical
Lifepo4 solar storage applications
Lifepo4 cost analysis

ALL OTHER / 3.7v nominal technical
ALL OTHER solar storage, EV, and miscellaneous applications
ALL OTHER cost analysis

This would immediately alert the DIY'er that there is a difference in chemistries, voltages, and also in application, where the needs are different in each case. The invevitable financial aspect that gets analyzed for decades could be quickly skimmed, and then those wishing to continue on can safely ignore that aspect.

Dunno' - sticking to the original thread subject when it comes to lithium seems to be very hard - and "containers" like this might help.

[Sunking]

Inmichagan I did not say you cannot use Charge Controllers made for FLA. You are correct some can be made to work with Lithium. As you pointed out you can set set voltage anywhere from 2.2 to 2.6 vpc for a Lead Acid battery. It just so happens that LFP is a good match for FLA voltages. Not any of the other lithium fall in the right range as they are too high of a voltage. For each 12 volts of LFP lithium (4S) you can adjust anywhere from 13.2 to 14.8 volts. I will give you that. However very few give you that latitude. Most are just switch selectable for Gel, AGM, and FLA. Outback and th enew Outback group called Midnite Solar can do it. I assume there maybe more.

However there is still one issue to work around, 3-Stag Algorithm they all use is a FLA algorithm. For Solar a 3-stage charge is useless, not enough hours in a day to complete. Again with some controlers you can work around that by setting all the 3 stages to the same voltage or close enough to work. But the one stage you cannot get around is FLOAT without some sort or external control you have to come up with. That is beyond most peoples capability. Floa ti srequired for lead acid, but is a killer on Lithium if you use Balance Boards

To use Balance Boards correctly, they must communicate with the charger to do two functions.

1. When the First Board turns on, charge current must be reduced to a level the Balance Boards can safely bypass so no more current flows in the fully charged cell. No Solar Charge Controller has that ability. Solution is real simple. Get rid of the Balance Boards, Bottom Balance, and DO NOT CHARGE TO 100% and you can do it.

2. The other thing is when the last Balance Board finally turns or, you must TERMINATE the charge. No Solar Charger can do that except Genasun but that charger is for a single 140 watt panel. Again if your controller has a programmable input you can program to turn the controller off.

Today there is no Charge Controllers to work with Lithium for Joe Homeowner. In the EV market is common practice. They use a CC/CV charger. Very simple algorithm which can also be used for Lead Acid as CC/CV is the bets for both, just slower than 3-stage. Difference is th eEV chargers are made for any battery type, I do mean any and fully programmable and use a Protocol called Cann Buss a proprietary communications protocol developed for the Automotive industry developed so all the parts of an EV speak the same language and can be controlled.

I have no doubt in the near future there will be Solar Charge Controllers made specifically for DIY solar. Right now econoimics are not there, and there is no lithium battery available to the public that will work as expected. Right now all we have is short lived Chi-Coms that in th eend cannot compete with lead acid. Lead Acid last much longer, more reliable, and most important much less expensive both short and long term.

Now you can find a couple of commercial solar battery systems using Lithium batteries. But it is not DIY, and you cannot buy the batteries. They operate at much higher voltages which means you Joe Homeowner cannot DIY. If the public tried we would have a lot of dead people working with 300 to 600 volt batteries.

I applaud folks like you being pioneers who sail the unknown waters. Sounds like you have done your homework, open minded, and know what is going on. Keep it up. Also come back and share what you learn about cell drift. I think you will find exactly what myself and hundreds before you have discovered, the cells do not drift to any significant degree. It is the cell monitors and Balance Boards that cause them to drift which are a necessary evil for a consumer product. But you will never hear a battery manufacture admit that if you know the technology, Balance Boards are not needed and root cause of most premature failures. What company in their right mind is going to say you do not need their product. You would have to be a Willy and Karrak to believe that crock of BS.

[karrak]

Anyway, I have no doubt that my +200 mV flier (near death) cell was my own fault for it not having a BMS board like the others. As for the variation of the others, in the coming months, I'll have the data for the same cells, being used in the same manner, charged in the same way, without the boards. I'll post what I find. Unfortunate most forums are filled with opinion without the underlying data, for me, raw data is supreme.

I totally agree with you that raw data is supreme. I, and I am sure many others are very pleased that you are taking the time to post the information so we can all learn from it. I am also impressed that you are prepared to share your mistakes. We all make mistakes, by sharing information it hopefully means we won't all make the same mistakes.

I am not sure you can make a valid comparison between what has happened in the past with the battery balance and what will happen in the future as the battery characteristics will change with time. To have any scientific validity you would have to have two groups of cells, one with boards and one without, then you would have to have enough cells to make it statistically relevant.

Simon

[karrak]

As for me and my racing golf cart is a product of evolving. I did not wake up one day and say I am going to build a racing golf cart. Firs tone in around 2008 was just a plain ole cart using DC motors and FLA batteries with. FLA batteries were too expensive with a horrible track record. Everyone for the EV community knew about Thundersky Turds. Being an engineer and studying at first it appeared Thundersky batteries problem was lack of battery management. I was one of the Forum Engineers typing away one MUST USE BMS or ELSE. Bu tunlike you and Karrak I did not stick my head into the sand. I actually kept an open mind and researched.

I heard of Jack on EVTV going aginst the grain when he discovered it was the Shunt Boards that were causing a great deal of the failures and fires. I dismissed it at first as BS because I was an engineer and knoew you MUST USE BMS. Well as it evolves BMS is meaningless. BMS has no definition. In the early days a BMS was the Shunt Boards aka Vampire Boards.
...
So I played with the Vampire boards and chargers. Guess what happened Willy. I got up one morning and found a dead cell out of the clear blue sky. A vampire board failed shorted and destroyed the cell. Guess what? I remembered seeing Jack's video a couple of years earlier. Called him and got more up to date information. Hit up the EV community and found there were hundreds of people with the same problem. Not only did I learn it was the Shunt Regulators causing the destruction, it was also the source of DRIFT Karrak and you are clueless about.

So wanting to learn more, I did a lot more research, talked to battery manufacture I knew personally for 35 years and worked with, Custom EV builders. I evolved from the darkness and learned I was part of the problem being an Engineer that you MUST USE Automation and the public is not smart emough. I was right. You are not smart enough to operate lithium without a lot of technology.

Well, you as an engineer should have realised that battery balancing boards are not a BMS, but are just there to keep a battery in balance. You should also have checked whether the failure of a balancing board could damage your battery. If you have had some sort of cell monitoring and alarm system, which is the main component of any BMS system it would have picked up the fault and saved your battery from damage.

Now me, I also did my research and decided that monitoring cell voltages was the most important thing to do, and that balancing was a secondary issue that could be done either manually or with balancing boards. I decided to do the balancing manually and maybe change to automated balancing at a later date if need be, I am still doing it manually. I also came to the conclusion that it was very important to have the battery balanced before it was commissioned.

This setup has served me very well for over two and a half years. My $50 cell monitoring system has saved my battery from damage due to a stuff-up by me and the battery supplier and I have had no issues balancing the battery manually and have had no issues with my cell monitoring system pushing my battery out of balance.

But Coulometric Efficiency has nothing to do with Drift. It is charge efficiency and the heat generated which breaks down the electrolyte and contaminates the SEI barrier. That is the Flaw with Chi-Com cells as their CE is on th eorder or 96 to 975 that causes th elow cycle life and rapid capacity losses.

This is not true, obviously putting links to other post so you can do some more research doesn't work so here it the text from one of my previous posts. Please tell me if there are any flaws in my logic.

What if we have two identical cells connected in series with a capacity of say 100Ah, one has a cell leakage of 5% per year, the other 2.5% per year. We charge them both to 100%SOC, We then leave them for one year. At the end of the year one has 97.5Ah of charge stored, the other has 95.0Ah of charge stored. If we try to draw 97.5Ah from this battery one cell will be at -2.5Ah. Are they still balanced at the bottom end?

I have been measuring the current coming and going from battery for around a year now. My figures are 23981.33 Ah has gone into my Winston "Chi-Com" battery and 23726.71 Ah have come out of the battery. This give an efficiency of around 99% or Coulometric Efficiency of around 0.99. The efficiency of the battery is even better than this as this doesn't take into account the power used by my battery monitoring system. I will eventually get around to factoring this in to my calculations.

Simon

[inetdog]

Wow! A few more improvements and you will have a greater than unity system.

I find that .99 charging efficiency hard to believe, but I guess as long as you stay well below 100% charge and have batteries with low self discharge it is possible.
Some battery AH meters have internal algorithms to divide the measured charging AH by a preset efficiency factor. I assume yours has been verified to be plain vanilla.
We are used to far worse charging efficiencies with lead acid.

[karrak]

Wow! A few more improvements and you will have a greater than unity system.

I find that .99 charging efficiency hard to believe, but I guess as long as you stay well below 100% charge and have batteries with low self discharge it is possible.
Some battery AH meters have internal algorithms to divide the measured charging AH by a preset efficiency factor. I assume yours has been verified to be plain vanilla.
We are used to far worse charging efficiencies with lead acid.

I find it hard to believe as well. I usually charge to around 90%-95% SOC, unlike Lead Acid you don't get a decrease in coulomb efficiency charging at high SOC

This is not charging efficiency, it is overall current efficiency of the battery. Looking at overall power efficiency I only have figures for the last few months, these are 203.259983 kWh going into the battery and 190.222493 kWh going out, or around 94% overall efficiency.

Simon

[Sunking]

Well I have been nice up until now. It is now time to put Kaarak out of my misery and finish him off. So will the real Karrak please step forward.

I agree with everyone these LFP threads get out of hand. I am tired of it too. However IJ am not the guilty party and did not start it. That honor goes directly to Karrak and it is time for Karrak to depart and leave for good. I got the knock out punch that will clearly show who the real Karrak really is.

Karrak came to this forum for one purpose and one purpose only. To discredit and attack me and I will prove it beyond any reasonable doubt. All anyone has to do is pay attention and follow the event time line.

Karrak first appearance came on May 11th of this year in this thread. Started nice enough, and then just got ugly. Every thread Karrak has been involved with on this forum since that day is to discredit me. Those of you that have been around know I speak the truth.

Karrak got so desperate and beat up so badly went to , Endless-Sphere, and Energy Matters recruiting muscle to come here and try to beat me up. He finally found some poor guy username WB9K a fellow ham operator who works for A123 Systems a Lithium battery manufacture. Poor guy had no idea of what he was stepping into. I had no real problem with him as he can only have one POV, the one he is paid to say and promte. If he said anything different would get him fired. I agreed with him if you are a manufacture of equipment using Lithium batteries, or a manufacture of the batteries, Thy Shall Use a BMS and Top Balance. Anything else would get you sued into history and bankruptcy. Poor guy got his arse handed to him and left. This is DIY forum, not a manufacture forum promoting their product.

Unfortunately Karrak just cannot give up and know when he has been beat. To this day comes here to discredit me, and follows me around except on a DIY forum where he knows to stray clear of because he knows he will get crushed. Everything that is used on Solar LFP batteries comes from the EV sector. Not the commercial side, the do it yourself side, and custom EV makers and EV racers. Again a manufacture can only take 1 POV, the one that butters their bread, and CYA from lawsuits and liability. However on a DIY EV forum you find some very tech savy folks, both pros, and folks like myself who have a great deal of electrical background, but no ties to the EV industry or battery manufactures. If you go there the same debates brake out and evenly divided. The manufactures claim you must use BMS and Top Balance, and the other side of the pros who say BS. Both are correct with respect to their interest. Manufactures insist knowing full well the Balance Boards are the root cause of failures. They will not admit publicly, but the data and history is there. It is far less risky for the manufactures to take their posistion and makees perfect sense. The public is ignorant. Much less risk and lees expensive to make warranty replacement when their equipment fails and kill a cell. All it takes is one injury fire or death and they are out of biz. But for those of us who know what is going on do not buy into that.

Raceers and DIY EV guys know the weakness of LFP batteries. The biggie is over discharge. It is instant death. There are two ways to over discharge a cell. Leave a load connected and bleed it to death. That is what the vampire boards do. The other is a unbalanced pack and the adjacent cells eating the lower capacity cells by reverse polarity. You see when a LFP cell goes to ZERO becomes esentially a peice of wire and able to pass current. You will not know until it is too late.

How do you stop that? There are two ways. Depend on automation that monitors the cells so if any cell drops below 2.5 to 2.9 volts to disconnect the pack. If that fails they will eat each other as the higher capacity cells drive them to death.

The other is passive using a strategy and common sense. Bottom Balance the pack. All the cells will arrive at 0 capacity at the same time, thus eliminating the adjacent cells to pick up the slack and kill the lower capacity cells. You can only truly Balance thee cells at the bottom where both voltage and capacity or equal. The second line of defense is already there in you LVD. You set the LVD to 3 vpc. So for a 16s or 48 volt system set th eLVD to 48 volts. The pack voltage would have to drop to 40 volts to reach the danger zone of 2.5 volts. FWIW th ereal cliff is 2.0 volts but that is another topic. With BB pack and proper settin gof the LVD is all you need to protect yourself from over discharge. On the charge side get rid of the killer Balance Boards and only charge to 90% SOC and stay away form the upper limit. You have no biz going to 100%. It only cuts cycle life.

OK enough it is time to put the final nail in Karrak. He will never tell you this on this form, but he agrees with me and has said so unknowingly. Don't deny Karak members have heard you say it. Those two members are PNJunction and Inetdog. They know everything I have said is true about you. They are members of the same forums you go try to discredit me and recruit muscle.


So let's drive the last coffin nails in coffin. I quote you directly right here in your own hand.

You stated on this Forum:

There is evidence and I agree that it is not a good idea to have battery balancing circuits across the cells. Automatic battery balancing is probably unnecessary for a system that you or I might build and maintain

I agree with you, it is not needed. Then in another thread you say:

I don't have any balance boards

WTF. What in the heck is your Major Malfunction? Well I will tell you what your major malfunction is. On this Forum you say:

I find this rather amusing as I have never used Pb batteries

Last one and you will hate this when you said on this Forum

From my experience over many years with lead acid

Your major Malfunction is you are a LIAR.

So would the real Karrak please step forward.

[Mike90250]

When does the chest beating stop?

[dax]

Unfortunately there are no LFP chargers for DIY solar as LFP is not ready for solar. Economics do not work. With DIY solar you are forced to use Chi-Coms like Calb. None of the Chi-coms will last as long as Lead Acid. LFP cost roughly twice as much and last half as long, and in some cases only 1/5th as long as lead acid. Your CALBS came with a worthless 1 year warranty or 1000 cycles. Try making a warranty claim.

Only came back to let those who asked about our lifepo4 solar controllers, that we have decided on a manufacturer and will have them available in Aus in a month or so. So our new website is parked until stocks arrive. For those interested, as won't be back here again, expect after what I'm about to say they will chuck me out. The website when operational is southernocean-industries.

As for warranties, what planet are you on. Our cell manufacturer gives us 5 years warranty and we give 5 years full warranty on components and 10 year parts warranty. Now we can repair cells and offer our customers a 50% trade in for their old cells, when needed, simply because lifepo4 cells are easily rejuvenated. Our mass produced controllers, will come with a 2 year replacement warranty and 5 year parts warranty, panels come with 10 years capacity warranty and I have panels from back in the 1980's which are still producing good energy.

Having a read through the threads, revealed some of the biggest myths I've seen regarding lifepo4 and they are life span and controllers. As for them being uneconomical, hilarious rubbish only those without any knowledge of them and unable to understand them would claim. Our customers get back their outlay within 5 years and some commercial, shorter times. Our longest outback off grid install of 2000amp x 5kw in a farm stay, got their money back within 2 years because the cost of diesel generators was exorbitant and they suffered blackouts constantly, or no fuel during the wet season. They haven't used the genny since the installation and want to upgrade their system because their business has expanded, which we intend doing in the new year and with our new integrated controllers. They will be our on site test bed for them, so they get them pretty cheap and may have to put up with teething problems.

<deleted sentence. Mod>
no idea about anything to do with lifepo4 used in off grid, RV/marine and is spreading lies about them for some insane reason. Lifepo4 is very much ready for off grid and being installed around this country successfully every day.

We have had been using lifepo4 systems for 6 years, in the form of 2 x 120amp 12v packs and using our own dedicated controllers, both solar and mains on job sites daily. My house system has been in over 5 years and we have 2000 installs so far, none have ever had a problem with their systems to speak of. Testing the 120amp packs after 6 years of use and abuse, capacity sits at 126amp, which is 1 amp above when I first got them. How that happens I don't know and our techies are trying to understand it, the same with voltages, some days the work packs will come home at 3v a cell and the next morning they have 3.27. We don't use a BMS or balancers on these work packs, they are connected to a 40ah controller and switch off at 14v. Cells will go over fractionally if there is no load, but with a load, they balance beautifully and my house pack has been without a BMS and balancers for the last year. When not in use on the job, they sit on portable panels and have yet to go down to 12v in all the years of use. They still amaze me as to how much you get out of them, when first started keep checking voltages many times a day, now don't even bother. At lunch time they go on the panels, plus 40amp charger and then are used until knockoff.

We would not install a customer system without BMS balancers, until next year when our new charge controllers are proven and they will control cell lines individually, switching off each line as it reaches 3.5v. They will be available in February we hope. The design has been proven and we have overcome pack temperature control, having tested it in situations of temps over 50degC and in the next month our test pack will returned from way down south where it spent a winter and it's temp control kept it at 18-20degC. The fix was so simple and cheap, should extend life spans much more.

Our solar controllers have been so successful we have not been able to keep up with the demand from our customers and other installers, so we are having them manufactured to our specifications in bulk and have orders for over 50 already and there has been no advertising. In fact the only ones aware outside the industry is this forum.

As for bottom balancing, good laugh and may work on some hobby stuff, but on off grid, absolutely ridiculous and if you have any experience you would know why. We balance our system packs before installation at 3.2v and with our parameters, the active balancers they stay well balanced. Customer records show on average, a cell line will go over between .01-.03v. If they are connected properly they rarely ever go over and to connect them properly with lifepo4, means you have to connect to all 4 points of the pack, otherwise they will be really hard to keep in balance. Charge into lifepo4 doesn't work like lead acid in any way and that's where many make their mistakes.

It's the same with cell and pack sizes, we use 40-50amp cells in 500amp packs paralleled to make one bank. This way you can control cell temps, charge regimes and easily access any problems. As an example, we recently installed a 5000amp x 10000w panel system, which is set up at 750w per 60amp controller and 500amp packs and it runs dual 24v 3000-6000w inverters.

I'm sure there will be those who will try to pick it to bits, but that's because they don't have a clue about these systems and how to make them disaster proof. The aim of off grid system is to ensure an uninterrupted supply, you can only do that if you have backup and our systems provide backup. Normally only one part of system breaks down, if you listen to the experts, that would mean you have no power. But with our system, you can isolate the faulty component and still have power, even if it is a bit less.

It's perfectly clear there is no understanding of energy supply system in this 21 century by the supposed experts and engineers. We went through a number of engineers when starting out, but they are so stuck in their programming they can't change their thinking to suit this new and very different technology. So I with no technical knowledge or understanding of electronics and the electricians in our company took it all on and now we have it pretty much together.

When you boil it all down, lifepo4 is very simple to use, keep them between 3v-3.5v and you will get to use 80% of the stored energy. Above and below those parameters, your cells with struggle after a short time of 1-3 years. Keep the cells at temps between 15-22degC, 18degC seems to be the best. If possible steer clear of large cell sizes, they fail because of temperatures differences within the cells and pack, for some reason people put them together so lightly they have no air circulation at all. Yet to have a small cell swell up from abuse, or ignition, but large cells swell after short periods when charged and discharged hard. Our techies say it's a very simple thing to understand when you understand how the chemical constituents act within the cell and how they are made up. Because there is such a thin membrane between the active components, any repetitive increase internally in heat in a large cell, restricts that heat from being dissipated and that goes for plastic encased cells. They can't release heat in any way, expect to melt the membrane and plastic casing, buckling the cell. Metal small capacity cells, with good air circulation through the pack, can dissipate heat rapidly, it's just pure logic when you have spent years actually working with these wonderful energy systems.

I noted on one thread someone said we are claiming to have a superior battery, but that is not the case, we are claiming we have superior control systems. In my opinion and having tested many different brands and sizes of cells over the years, the Chinese ones we install have been in use for more than 5 years. Not just one system, but now 2000 systems and no failures.

I think that qualifies us as having some experience with lifepo4, compared to those who only use their last century understanding to confuse people with past its use by date technical jargon, into believing they are not what they are claimed to be and inferior, when they are better than anything on the planet in every way.

[PNjunction]

Hi Dax, welcome back!

Testing the 120amp packs after 6 years of use and abuse, capacity sits at 126amp, which is 1 amp above when I first got them. How that happens I don't know and our techies are trying to understand it, the same with voltages, some days the work packs will come home at 3v a cell and the next morning they have 3.27.

Yes, they will do that. Diffusion is part of the answer. Guys that do bottom balance properly, will take days bringing each cell down to say 2.7v, and wait a day or two and do it again - and possibly again to get it to finally rest at an ocv of 2.7v. Those that don't wait 12-24 hours aren't *truly* bottom balancing, but I know you don't do that - but just a point of interest that this is a common phenomena.

As for bottom balancing, good laugh and may work on some hobby stuff, but on off grid, absolutely ridiculous and if you have any experience you would know why.

Actually it does work, but one has to consider the application. Now you definitely need individual cell monitoring for triggering on the weakest cell. Perfect for EV's or other applications, but top-balance (a misnomer really due to each cells differing capacity, internal resistance, and voltage diffusion response) is probably the most convenient for our application.

We balance our system packs before installation at 3.2v and with our parameters, the active balancers they stay well balanced.

That has to be a typo, otherwise you are balancing in the middle of the charge curve, which represents a very wide state of soc. Do it either at the top or the bottom, but never in the middle. Some newbies who receive cells and see them all sitting at 3.2v may mistakenly think that their cells are all well balanced - far from it.

Also, perform this test like I have: Marry a cell that is half or anything lower in capacity than the others in a bank. Hook up the top-balancers. You can easily top-balance them, but an unusable battery is the result. Ie, I have taken one cell out of my 40ah / 4S bank, and replaced it with a 20ah cell from another. Sure I can top balance it, but in the end, that 20ah cell is the limiting factor. This is an extreme example of course, but I like to actually do stuff like this when I get tired of typing myself smart.

Customer records show on average, a cell line will go over between .01-.03v. If they are connected properly they rarely ever go over and to connect them properly with lifepo4, means you have to connect to all 4 points of the pack, otherwise they will be really hard to keep in balance. Charge into lifepo4 doesn't work like lead acid in any way and that's where many make their mistakes.

Sure enough - one of the biggest mistakes is not cleaning the oxidation off of terminals, which in the case of prismatics are usually a mix of dissimilar metals. Ie, copper and aluminum, PLUS the end-user's own connections, which are frequently steel, nickel-plate, stainless etc. These MUST be cleaned first, and a light coating of anti-oxidant like No-Alox, PeneTrox, and the like should be applied - and connected together relatively quickly to prevent oxidation from happening again. Some assume that if it *looks* clean, it isn't oxidized. Not so and the result can be high or varied resistance contacts.

Those poor house-owners that had aluminum wiring instead of copper can relate.

If possible steer clear of large cell sizes, they fail because of temperatures differences within the cells and pack, for some reason people put them together so lightly they have no air circulation at all. Yet to have a small cell swell up from abuse, or ignition, but large cells swell after short periods when charged and discharged hard.

That's a sweeping generalization which sets off flags. Was the temperature REALLY caused by poor installation techniques, ie no cleaning of terminal contacts prior to assembly, which resulted in a high-resistance contact which burned up the cell? That's one possibility. Aside from the assumed full charge and discharge at the factory, cells need to be fully charged at least once, preferably upon receipt to help prevent hot-spots clumping of material. (Typically 3.6v until .05C is reached). Those of us who run conservatively at a lower voltages routinely may not realize this. Was anyone at your site trying to charge down to 0C absorb? If so that hurts the cells too. Were they charging at less than .05C, and trying to drive cells above 3.45v? Since that is *already* just at or under absorb, any voltages higher than that are actualy coming from electrolyte heating / parasitic phenomena, and not any real charging which was far over already.

There are ways to hurt cells other than just abiding by a voltage spec.

Try it - it's fun. Take a spare cell, and charge at less than .05C. Watch it rise to 3.45v (which at this point is fully charged). It will plateau there, but keep watching. A while later - voltage zoooooom from electrolyte heating / parasitics. Be safe of course. Aside from being slower than watching paint dry, this is one reason I either set my upper limit to 3.45v, or use a higher charge current >.05C and change the upper limit to something like 3.55v for absorb.

Another commonly overlooked problem is applying more than C/50 charge current when the cells are below 3.2v ocv. THIS is a common internal separation cause - too high a charge current when the cell is in the deep part of the discharge knee. Cells need to be brought back gently to 3.2v, and THEN the normal higher-current charge can be applied. EV'ers discovered this the hard way. Come home after a severe deep discharge, put on the mondo-charger, and vent / swell time - or just severe cycle life reduction. Heck, even small lifepo4 powersport chargers like Tecmate-Optimate TM-291 chargers know this.

This is something often overlooked in a solar storage application, and another reason to go conservative - just as you reach say 90% SOC with solar, clouds pass and you are now LESS than .05C. Uh oh, the controller is trying to allow charging to 3.55v, which at .05C is overcharge! May want to investigate that, and possibly incorporate some very low charge-current smarts of .05C or less into the controllers which would knock down the high trigger to 3.45v after some sort of time period algo.

Our techies say it's a very simple thing to understand when you understand how the chemical constituents act within the cell and how they are made up. Because there is such a thin membrane between the active components, any repetitive increase internally in heat in a large cell, restricts that heat from being dissipated and that goes for plastic encased cells. They can't release heat in any way, expect to melt the membrane and plastic casing, buckling the cell. Metal small capacity cells, with good air circulation through the pack, can dissipate heat rapidly, it's just pure logic when you have spent years actually working with these wonderful energy systems.

Sounds like they only have EV / RC backgrounds and are now applying it to your product. Just be careful of tunnel-vision. Large prismatics, even the lowest-grade ones, can easily withstand 0.5C to 0.1C without heating up, PROVIDED the end-user infrastructure is sound. Fortunately, I'm not interested in selling to the guy next door, but those of us in a technical forum that have the ability to do it right. Going beyond 0.1C for long amounts of time (manufacturer specs point the time and current out) can result in heating. Thing is, WE ARE NOT EV, nor are we running a high-current application - that is for anyone designing a system with an adequate capacity in the first place. Typical currents of 0.1 to 0.2C in a solar storage application with decent autonomy are common - unless one is trying to stuff them into a small space, like EV nooks and crannies. Prismatics, initially designed for amateur / niche commercial EV's, basically laugh at our low-current (aka "Sub-C") application.

I think that qualifies us as having some experience with lifepo4, compared to those who only use their last century understanding to confuse people with past its use by date technical jargon, into believing they are not what they are claimed to be and inferior, when they are better than anything on the planet in every way.

The fact of the matter is that until you see them being sold in your local gas-station, lifepo4 is a niche product and will remain so no matter how much cheerleading from those that love them do, like myself.

For instance, lifepo4 makes LOUSY standby/backup batteries, *for a solar storage system*, unless you have very very deep pockets to over-spec the capacity. Huh? For a standby system that WE would use, that means storing the batteries at much less than a full charge. Thus, to get the capacity we need for a standby op, it has to be much larger and of course more expensive. Or, one could just store them at 100% SOC and replace them sooner than later.

So as it is, for ME, lead-acid is still the most cost-effective solution for standby / float and the reason I still have those around. For my daily cycling - lifepo4 of course. But because I don't worry about sulfation and the like, they can be found quite deeply discharged and not suitable for emergency standby because of that. Thus, my lead-acid's performing that role, which can be kept at 100% SOC with the proper maintenance.

Anyway, courses for horses I guess. I wish you all the success in the world, but we have to admit that lead-acid isn't going away tomorrow. Or the next decade. When it appears at the corner gas-station is when it will have truly arrived.

[inMichigan]

I totally agree with you that raw data is supreme. I, and I am sure many others are very pleased that you are taking the time to post the information so we can all learn from it. I am also impressed that you are prepared to share your mistakes. We all make mistakes, by sharing information it hopefully means we won't all make the same mistakes.

My years in China highlighted that to communicate technical issues between two people of different skill levels &/or language difficulties, the key was not to repeat the same argument with the same logic (this is a engineering habit), but instead, try from very different angles. Once the other person has the ah-ha moment, then, the original, more literally correct explanation will make sense.

I am not sure you can make a valid comparison between what has happened in the past with the battery balance and what will happen in the future as the battery characteristics will change with time. To have any scientific validity you would have to have two groups of cells, one with boards and one without, then you would have to have enough cells to make it statistically relevant.

Yeah, I knew somebody might bring that point up. I have four ideas along:
a) A compromise would be after 90 days of 'no boards', put them all back on long enough to see the variation across the cells is greater than the non-board data for the same point in time.
b) Randomly assign 8 with and 8 without (and ignore #13)
c) If the drift after 360 days (4 times the original duration) is still statistically less the 'with board, I'd have a strong opinion about it, but not conclusive data. I am glad that I played/tested the new batteries in May/June, and then rebalanced before the July 3rd initial grid tie event. So at least I didn't complicate the experiment with "new out of the box cells"...
d) I did wire the battery cabinet with a second Gigavac DC switch that could be used with a second battery bank. In the spirit of scientific knowledge, as a cost savings, I'll use a truck to pick up any donated cells at a nearby shipping terminal to save on the lift gate charge. Go ahead and send additional brands for benchmarking, and let's answer the AC ripple question with a nice 800 AH set as well.

(d) is my preferred choice and will keep me busy as well. The holiday season is just around the corner.

I might consider (b), its correct follow-up would be then to rebalance & repeat a few times with increasingly higher charging voltage. I would expect my lower charging voltage would have the no-board tight, somewhere in the middle they might both have the same variation, and at a higher voltage the 'with board' set has less variation. If that is what happens, then, many conflicting forum arguments might all be true, depending on the charging strategy. And I'm certain my charging strategy (being on-grid hybrid) certainly does not need to be the same as an off-grid cabin or marine, or racing golf cart, or hobby RC.

inMichigan

[karrak]

My years in China highlighted that to communicate technical issues between two people of different skill levels &/or language difficulties, the key was not to repeat the same argument with the same logic (this is a engineering habit), but instead, try from very different angles. Once the other person has the ah-ha moment, then, the original, more literally correct explanation will make sense.

Hard to do for someone with an engineering mind. In a previous life my company sold equipment we made to the Chinese. I enjoyed the experience of visiting China and working with the people there. Quite a different mindset and way of doing business but with many good points.

d) I did wire the battery cabinet with a second Gigavac DC switch that could be used with a second battery bank. In the spirit of scientific knowledge, as a cost savings, I'll use a truck to pick up any donated cells at a nearby shipping terminal to save on the lift gate charge. Go ahead and send additional brands for benchmarking, and let's answer the AC ripple question with a nice 800 AH set as well.

(d) is my preferred choice and will keep me busy as well. The holiday season is just around the corner.

I might consider (b), its correct follow-up would be then to rebalance & repeat a few times with increasingly higher charging voltage. I would expect my lower charging voltage would have the no-board tight, somewhere in the middle they might both have the same variation, and at a higher voltage the 'with board' set has less variation. If that is what happens, then, many conflicting forum arguments might all be true, depending on the charging strategy. And I'm certain my charging strategy (being on-grid hybrid) certainly does not need to be the same as an off-grid cabin or marine, or racing golf cart, or hobby RC.

I agree, (d) is the best choice.

I don't want to appear like a broken record but by comparing the current consumption of all the balancing boards you have removed you will get a good idea if it is them that was causing the imbalance. If you don't have the equipment to measure the current consumption directly using a bench power supply to simulate a LFP cell, there are ways to do this using your battery, fluke multimeter and a resistor.

I think you are right that some of the conflicting arguments are right depending on how you charge and discharge your battery. It would be good to get as much data as possible to find out what factors effect battery operation and then to have a rational discussion about different strategies on managing LFP batteries to get the best possible safe utilisation from them depending on ones application.

Simon

[karrak]

Well I have been nice up until now. It is now time to put Kaarak out of my misery and finish him off. So will the real Karrak please step forward.

I agree with everyone these LFP threads get out of hand. I am tired of it too. However IJ am not the guilty party and did not start it. That honor goes directly to Karrak and it is time for Karrak to depart and leave for good. I got the knock out punch that will clearly show who the real Karrak really is.

Karrak came to this forum for one purpose and one purpose only. To discredit and attack me and I will prove it beyond any reasonable doubt. All anyone has to do is pay attention and follow the event time line.

If you have been nice up till now I would hate to see you when you turn nasty.

So let's drive the last coffin nails in coffin. I quote you directly right here in your own hand.

There is evidence and I agree that it is not a good idea to have battery balancing circuits across the cells. Automatic battery balancing is probably unnecessary for a system that you or I might build and maintain

Only problem is that you have deliberately altered the text I wrote by deleting the words "by themselves" before the full stop which changes the whole meaning of the quote. Below is the full paragraph to put this in context

"I can't see how a properly designed Battery Monitoring system could possibly damage a large battery bank that is being charged on a regular basis in an off-grid system. There is evidence and I agree that it is not a good idea to have battery balancing circuits across the cells by themselves. I think a well designed battery balance circuit that takes account of the failure modes that might occur in the circuit/software coupled with some sort of battery monitoring that would pick up that a balancing circuit had failed would be OK. Automatic battery balancing is probably unnecessary for a system that you or I might build and maintain, however I think it might be useful in cases like my friend's 48 volt system, I am still undecided about this."

Then this one

I find this rather amusing as I have never used Pb batteries

The full sentence was
"I find this rather amusing as I have never used Pb batteries in my off-grid system, NiCd and LFP yes, Pb no."

It seems obvious to me, that you see this as a war, where you are right and that those that don't agree with you don't know what they are talking about, are only worthy of contempt, and that you will go to any lengths including dishonesty to try to show you are right.

I am afraid that unless I have some heavenly apparitions I will not necessarily be believing the gospel according to Sunking. There are some things I agree with you on and you do make some good points.

I would like to think that I could easily accept arguments that you put forward that have merit but find it difficult to look at anything you write with any objectively because of the way I think you have tried to intimidate and bully me and others in the past.

Simon

[inMichigan]

I don't want to appear like a broken record but by comparing the current consumption of all the balancing boards you have removed you will get a good idea if it is them that was causing the imbalance. If you don't have the equipment to measure the current consumption directly using a bench power supply to simulate a LFP cell, there are ways to do this using your battery, fluke multimeter and a resistor.

I checked the manual of my Fluke 179, seems I can do mA measurements. I'll do some testing in the coming weeks.

Decided to do one cycle of deep discharge and refill to see if all the drama has had any serious damage... with a great sigh of relief, 3.35 v/c drained to 3.14 v/cell put out 95AH and the recharging took 100.5AH.

inMichigan

[inMichigan]

Feeling the need to plot something this morning...



The is a plot of 3 variables:
Horizontal--Days since the time when I did a rebalancing
Vertical Position--At this time of measuring each cell's voltage, how much different was the average voltage from the time when it was balanced.
One would expect that if measuring at a state of charge that is very different from the balancing level, the variation in each cell voltage ought to be bigger
Sphere towards the bottom tend to be measured in the morning after using Grid-Zero mode to drain the batteries by feeding the house.
Sphere Size--I used the standard deviation of the cell voltages.
If you aren't familiar with standard deviation, roughly, think of this as a kind of range (biggest - smallest).
Think of each sphere as a cloud representing the 16 or 17 cells

As mentioned one of the threads, real data needs a lot of footnotes...

"1st" was during my experimental phase. Mostly charged by the inverter & grid. One CC was setup for some minor testing. BMS boards installed on all cells
"2nd" was after installing the full system on the wall. Cell #13 BMS had failed and was removed. I did not include Cell #13 in the standard deviation.
"3rd" was after removing all the BMS mini boards

I've been measuring more often, so I have a lot more different conditions in the 3rd set. Day/night, different periods of rest, should make for interesting analysis in a few months.

inMichigan