LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[mbianka]

I'm about to start my seventh year with 8A4D AGM's setup up in a 48 volt string for my electric propulsion system. Batteries are holding up very well. So far I don't see much reason to change to LiFPO4. Lith Ion was fairly new in 2008 when I converted to electric propulsion and $$$. But, my major reason for sticking with AGM was my plans at some point for ICW cruising and the ability to replace a battery (if I needed too) in a relatively short time (like 24 hours) when underway. Lead times for Lith Ion was months back in 2008. Has delivery times for the Lith Ion changed? How about price over the years? Just curious.

[mcarling]

Originally Posted by T1 Terry

If you have a 36v series string battery then connecting 2 cell loggers across the 12 cells, watch the cell votage as you approach the 42v mark, are all the cells at exactly 3.5v at the same time the 42v mark is reached?

Exactly? Of course not! The same is true for lead-acid.

Originally Posted by T1 Terry

Keep a log over multiple discharge/recharge cycles ad watch what happens just at the end of the 42v charging cycle, then you will understand the need for individual cell voltage monitoring.

I understand what happens, but see no "need" to monitor it. The practical effects are negligible and not, as far as I can see, in any way relevant to practical use of the battery bank. I check mine about once a year and have yet to find any need to rebalance.

[T1 Terry]

Originally Posted by mcarling

Exactly? Of course not! The same is true for lead-acid.


I understand what happens, but see no "need" to monitor it. The practical effects are negligible and not, as far as I can see, in any way relevant to practical use of the battery bank. I check mine about once a year and have yet to find any need to rebalance.

It's my turn not to understand what you are saying. The way I read it you are saying the cells are ot damaged by high voltages, as in voltage higher than 3.6v, is that what you are saying?

T1 Terry

[georgec]

I look at the BMS as an insurance policy on your bank. It's your decision. It is not needed on a one or two hundred dollar bank on a bicycle, but on a three to five thousand dollar house bank, I definitely want to know what is going on. If you cook a cell in the middle of nowhere, not so easy to get a replacement.

[Hkalan]

Monitoring... Some love it, some care less...

To each his own !

But if you have all the bells and whistles, it sure is nice to know all the details.

My two cents... Knowing the charge, and seeing how I use the current can never hurt the situation ! Showed me that I needed to upgrade the watts of solar I had onboard, and even showed me when to de sulfate the old lead acid.

Wish I had a system in place to help me understand the wife... Intoxication is working for the time being LOL

Alan

[goboatingnow]

Originally Posted by georgec

I look at the BMS as an insurance policy on your bank. It's your decision. It is not needed on a one or two hundred dollar bank on a bicycle, but on a three to five thousand dollar house bank, I definitely want to know what is going on. If you cook a cell in the middle of nowhere, not so easy to get a replacement.

Absolutely . What a nonsense idea thAt you'd spend 1000s on the batteries and then not bother spending a few 100s on monitoring it

Dave

[mcarling]

Originally Posted by T1 Terry

It's my turn not to understand what you are saying. The way I read it you are saying the cells are ot damaged by high voltages, as in voltage higher than 3.6v, is that what you are saying?

My cells have never reached 3.6V and the voltage at which damage occurs depends on the duration (or current, depending on how you want to look at it) and (according to some of the academic literature) on the temperature. An impulse of 10V will not harm a LiFePO4 cell but 3.8V sustained over a few hours probably will.

Originally Posted by Hkalan

seeing how I use the current can never hurt the situation !

... except that a BMS failure could cause an imbalance.

Originally Posted by goboatingnow

What a nonsense idea thAt you'd spend 1000s on the batteries and then not bother spending a few 100s on monitoring it

House banks generally don't cost thousands of dollars because few boats need more than about 250AH at 12V (nominal) in a house bank. Propulsion banks do cost thousands of dollars. How many people do cell-level monitoring of thousands of dollars worth of lead-acid batteries in a propulsion bank?

Of all the things that I've ever spend thousands on, I've never spent hundreds monitoring any of them. If it were normal to spend so much on monitoring, 10% of the population would be employed as private investigators.

[simat]

Originally Posted by mcarling

My cells have never reached 3.6V and the voltage at which damage occurs depends on the duration (or current, depending on how you want to look at it) and (according to some of the academic literature) on the temperature. An impulse of 10V will not harm a LiFePO4 cell but 3.8V sustained over a few hours probably will.

I would be surprised if the internal current leakage of all the cells would be identical so over what might be a long period of time the cells will get unbalanced.

On tests I have run the difference between 3.50 volts/cell and 3.76 volts/cell is a SOC percentage difference of only 4.5%, and between 3.50 volts/cell and 3.60 volts/cell is only 1.8%.

When you stop charging them, the further out of balance your cells become the greater the voltage difference between them. From the literature I have read higher voltage will reduce the life of the cells so you will end up with uneven ageing of the cells.

... except that a BMS failure could cause an imbalance.

Yes a failure in the balancing circuitry could cause imbalance, that is why you also should have some sort of individual cell monitoring with an alarm as well. You could do the balancing manually but IMHO should always have some sort of monitoring and alarm.

The other reason for monitoring and an alarm is to guard against human stuff-ups and unforeseen events. When I ordered my cells I paid extra to have them fully charged and balanced. This was not done and my cells were out of balance by ~30%. It was only my cell monitoring and alarm that saved potential damage to some of my cells.

[Maine Sail]

Guys,

Thanks for the link to the schematics I can see now that with four cells it draws balanced.. I measured the draw on the #2, #3 & #4 wire and it was well below what I would call a "reliable" resolution level but with it booting up on #1 & #2 I incorrectly assumed it drew power only from #1 and #2... Good to see the schematic so I can see how it actually works..

[mcarling]

Originally Posted by simat

I would be surprised if the internal current leakage of all the cells would be identical so over what might be a long period of time the cells will get unbalanced.

On tests I have run the difference between 3.50 volts/cell and 3.76 volts/cell is a SOC percentage difference of only 4.5%, and between 3.50 volts/cell and 3.60 volts/cell is only 1.8%.

When you stop charging them, the further out of balance your cells become the greater the voltage difference between them. From the literature I have read higher voltage will reduce the life of the cells so you will end up with uneven ageing of the cells.

Which literature? I read Applied Energy, which the peer-review journal where the professors who do battery research publish. The consensus there is that cell voltages do not give an accurate or particularly useful view of what is going on inside the cell. There is very nearly but not quite a consensus there that monitoring overall battery bank voltage is just as good as monitoring cell-level voltage.

Originally Posted by simat

Yes a failure in the balancing circuitry could cause imbalance, that is why you also should have some sort of individual cell monitoring with an alarm as well. You could do the balancing manually but IMHO should always have some sort of monitoring and alarm.

The other reason for monitoring and an alarm is to guard against human stuff-ups and unforeseen events.

Battery bank voltage monitoring will do just as well as cell-level monitoring against human stuff-ups and failure of external equipment such as battery chargers -- particularly if charging voltage is conservative. If there is a need to squeeze every last iota of capacity out of the batteries, then cell-level monitoring can help, but not as much as charging each cell individually. I prefer to keep it simple and reliable rather than try to squeeze out the last several percent of possible capacity.

Originally Posted by simat

When I ordered my cells I paid extra to have them fully charged and balanced. This was not done and my cells were out of balance by ~30%. It was only my cell monitoring and alarm that saved potential damage to some of my cells.

As above, I balanced mine in the beginning. I didn't need any cell-level monitoring to tell me that balancing in the beginning would be a good idea.

[Maine Sail]

Originally Posted by simat

On tests I have run the difference between 3.50 volts/cell and 3.76 volts/cell is a SOC percentage difference of only 4.5%, and between 3.50 volts/cell and 3.60 volts/cell is only 1.8%.

Would love to know how you tested this? When I charge to 13.8V or 3.45VPC, with Winston cells, and let the current tail to about 10A @ 13.8V (400Ah bank) I get all the capacity in the bank. The key is to let the current tail off once 13.8V or 3.45VPC has been attained. If you simply stop charging at 13.8VPC you won't get all the capacity. With a larger current source you will attain 3.45VPC a tad earlier than you would going to 3.5VPC, and the current taper will be slightly longer, but still VERY, VERY short compared to lead acid.

At a .25C load I can harvest 425Ah's out of the bank when charged to 13.8V/3.45VPC... If I discontinue the .25C load at 2.8VPC and turn on a 10A load I can get another 15 Ah's out of the bank or 440Ah's out of a 400 Ah bank when charged to 3.45VPC..... At a .5C load I get about 398-402Ah's when charged to 3.45VPC and current allowed to tail off....

Our loads on boats, when using LFP for house banks, are nowhere near .25 or .5C.... My bank is now beyond 400 cycles, most to 80% DOD, and with 9 complete capacity tests, it is still putting up the same exact capacity it was 400 cycles ago.

Stay above 2.9VPC and below 3.5VPC and the banks will do just fine and will stay in balance for a long while. Begin regularly pushing above 3.5VPC, and holding it there for a period of time, and cell drift can occur. By staying below 3.5VPC I have seen no cell drift even when current is allowed to tail off to 0.1A.

Staying out of the knees is what I have found to be most important to keeping the cells in balance..

I personally prefer to stay below 3.475VPC or 13.9V for a 12V nominal bank. Safe, easy, bank stays balanced, and I get all the capacity I need out of the bank..

I still have a BMS for a LVE or HVE but do not use it for balancing, just insurance..

[mcarling]

Originally Posted by Maine Sail

Would love to know how you tested this? When I charge to 13.8V or 3.45VPC, with Winston cells, and let the current tail to about 10A @ 13.8V (400Ah bank) I get all the capacity in the bank. The key is to let the current tail off once 13.8V or 3.45VPC has been attained. If you simply stop charging at 13.8VPC you won't get all the capacity. With a larger current source you will attain 3.45VPC a tad earlier than you would going to 3.5VPC, and the current taper will be slightly longer, but still VERY, VERY short compared to lead acid.

At a .25C load I can harvest 425Ah's out of the bank when charged to 13.8V/3.45VPC... If I discontinue the .25C load at 2.8VPC and turn on a 10A load I can get another 15 Ah's out of the bank or 440Ah's out of a 400 Ah bank when charged to 3.45VPC..... At a .5C load I get about 398-402Ah's when charged to 3.45VPC and current allowed to tail off....

Our loads on boats, when using LFP for house banks, are nowhere near .25 or .5C.... My bank is now beyond 400 cycles, most to 80% DOD, and with 9 complete capacity tests, it is still putting up the same exact capacity it was 400 cycles ago.

Stay above 2.9VPC and below 3.5VPC and the banks will do just fine and will stay in balance for a long while. Begin regularly pushing above 3.5VPC, and holding it there for a period of time, and cell drift can occur. By staying below 3.5VPC I have seen no cell drift even when current is allowed to tail off to 0.1A.

Staying out of the knees is what I have found to be most important to keeping the cells in balance..

I personally prefer to stay below 3.475VPC or 13.9V for a 12V nominal bank. Safe, easy, bank stays balanced, and I get all the capacity I need out of the bank..

I still have a BMS for a LVE or HVE but do not use it for balancing, just insurance..

This is consistent with the 100+ peer-review academic articles I've read concerning LiFePO4.

[Maine Sail]

Originally Posted by mcarling

This is consistent with the 100+ peer-review academic articles I've read concerning LiFePO4.

Yet the "masses" still want "drop in" replacements to charge at 14.4V plus and even then still not use any sort of BMS......

I figured all that out on my own through actual physical testing, pretty easy stuff to figure out when you do it manually and watch what happens.

I then began discovering the white papers but none of them really addressed fractional "C" or off grid type use or does it well. (please send me any that do) While a lot can be transferred over a lot of it can't because they are physically using them differently than we do on boats. I wanted simple, safe and reliable and without a BMS that "needs" to balance cells. Staying out of the knees achieves this.

I found charging voltages over 14.0V/3.5VPC to be not as safe or reliable and that they did lead to cells drifting.

I also found all the capacity was already in the bank at below 3.5VPC. Why go any higher, push into the danger/drift zone, and cause/create a need for continual cell balancing? Why do that when keeping the voltage below 14.0V/3.5VPC works and does everything you need it to....?

The nice thing about the House Power BMS is that it won't shunt balance until AFTER the HVC relay has tripped. This means HVC gets in the way of shunt balancing EVERY TIME the charge source tries to exceed 14.2V. It can not shunt balance automatically unless it is wired wrong.

With the HPBMS this means balancing would be done manually, attended (like equalizing an LA batt) and can be done safely at the lowest possible current to maintain "shunt balancing" and no more... Automatically letting a 200A alternator push to shunting levels is simply crazy IMHO..... The HPBMS does not allow this to happen...

I am shocked that the cell makers still put charging voltages in manuals that put you well into the knees needlessly.. I guess like the early GEL batteries if you want to sell more batteries they "must" be compatible with lead acid charging voltages.. DUMB IMHO......

I have watched one "lithium packager" FINALLY drop their voltage recommendation from 14.6V to 14.0V on a 12V nominal pack. This was a company selling a branded product for at least three to four years that had it wrong because they simply chose to not do the testing themselves and to trust the very poorly written Chinese manuals...... I am sure that 14.6V recommendation hit them right in the wallet and they finally had to get smart about it....

A 14.6V charge voltage recommendation dropping to 14.0V is a HUGE change, but now they at least "get it"....

I still will use a BMS as an insurance policy. Seen far to many LA banks drained by a bilge pump or over charged due to a failed regulator or charger malfunction.......

[mcarling]

Originally Posted by Maine Sail

Yet the "masses" still want to charge at 14.2V plus and even then still not use any sort of BMS.

With or without a BMS, I would not charge at 14.2V plus.

Originally Posted by Maine Sail

I still will use a BMS as an insurance policy. Seen far to many LA banks drained by a bilge pump or over charged due to a failed regulator or charger malfunction.......

Any reason why you wouldn't do HVE and LVE at the simpler battery bank level? It would protect against the bank being drained by a bilge pump or overcharged due to a charger malfunction just as well as cell-level HVE and LVE.

[T1 Terry]

Originally Posted by mcarling

My cells have never reached 3.6V and the voltage at which damage occurs depends on the duration (or current, depending on how you want to look at it) and (according to some of the academic literature) on the temperature. An impulse of 10V will not harm a LiFePO4 cell but 3.8V sustained over a few hours probably will.

But you quite emphatically tell us you do not do cell level monitoring, so how would you know? It's the bury the head in the sand theory that what I don't know can't hurt me?
Think of the number of combinations 12 numbers added together could equal 42v, can't you see that cell level monitoring tells you what each number is, battery level monitoring only tells you the sum 12 of these voltages? Without an alarm warning you cell has reached 3.6v you simply would ot realise it had happened.

Originally Posted by mcarling

... except that a BMS failure could cause an imbalance.

Battery Monitoring Systems do not cause an imbalance if they fail, automated cell balancing systems that use a lod to discharge high cell can and often do, I agree with your loathing of such system, I feel they are far more likely to damage a battery than any other issue, besides over voltage charging of course. The way I see it, only by monitoring cell level voltages can this be achieved.
let me assure you, brand new cells charged to 4.6v I think it was, causes a one time fatallity, I have 2 cells to prove it, neither of them burst their cases but both of them slowly vented their electrolyte resulting in no ability to accept charge current of ability to hold a charge. As soon as current was applied, the voltage rose the point the current fell to zero, all te way back up to 4.6v, I stopped the test at this point as the heat build up within the cell was becoming obvious.

T1 Terry