Magical LiFePO4 batteries giving more than rated capacity!

[Archerite]

I have a self build battery bank with 11 8Ah LiFePO4 batteries which should be a total capacity of 88Ah. Right? I have been using this battery for a little over a month now and besides the small capacity I am really happy with it. It's charged from solar 98% of the time and if it get's to low I use a grid charger but that's really slow at 5A.

Last night I was testing something and forgot to disconnect it in time....and the victron smart shunt showed 0% already! Whoops. But to my bigger surprise the system was still on...at 11.8v so far to low. I don't have a batteryprotect installed yet...going to look into that really soon. But what's even more suprising is the deepest discharge shown by the shunt:


Where did that 7Ah come from??? Is it an error in the shunt's calculations? Last sync was only 4 days ago. I have been using high power 150-200 watts in that time, and some of it got recharged by solar up to 80-90%. Is this something that happens more often that a deep discharge gives more than total rated capacity in a parallel battery system?

This is the battery I am using (11 in parallel): https://www.eco-worthy.com/collecti...4-12v-8ah-24ah-lithium-iron-phosphate-battery
I had expected the BMS inside each battery to be shutoff already but it seems they had more capacity than they are rated for. It was 03:40AM that I noticed it...so no sun keeping the system on either. They are deep cycle batteries so I think this one time going below 0% is not to damaging....at least I hope not.

 

[sunshine_eggo]

No magic, just goodness.

Ratings are minimums. Many cells outperform their ratings because they need to survive their cycle life. If the guarantee is 4000 cycles when discharged to 20% SoC with 80% capacity remaining after those 4000 cycles, there's a better chance of meeting spec if you're starting from 105% rated rather than 100% rated.

While LFP doesn't suffer significantly from Peukert, if the draw down was notably less than 16A, you might get an extra % or two.

There are multiple accounts of 100Ah battleborns testing to 107-110Ah.

 

[Archerite]

The part about magic was a bit of a joke offcourse ...but thanks for explaining it. I knew there had to be some kind of reason behind the extra capacity. I did test a single one of these 8Ah batteries to give about the rated capacity...but not with any testing software though. Just pull power until the BMS shutsdown and that was close to over 7.8Ah I think. Not sure how much current I was drawing for that test anymore.

While LFP doesn't suffer significantly from Peukert, if the draw down was notably less than 16A, you might get an extra % or two.

The draw from the batterybank was about 12A average over a period of 5-6 hours. And divided over 11 (cells) that's just a bit over 1A for each of them I think...in ultimate conditions. I have seen some deliver 2A while others gave nothing while building the battery, not sure what that was about. Probably a slight voltage difference or something but I made sure each (cell) had less than 0.02V difference before it was all wired together on the busbar's.

There are multiple accounts of 100Ah battleborns testing to 107-110Ah.

From a big and expensive brand I would expect this. But I would not immediately put eco-worthy in the same leage as Battleborn based purely on their price and rated capacity. But if in this case each 8Ah battery actually has 8.7Ah to meet their cylce rating...than it makes sense I got so much out of it. Also the low draw per battery as you explained makes sense that they gave a little more.

Thanks for the clearing that up. ?

 

[newbostonconst]

If you hook your battery charger directly to the batteries the charging power will not go through the shunt and will show up as excess power coming from your batteries when it is really coming from your charger.

 

[Archerite]

If you hook your battery charger directly to the batteries the charging power will not go through the shunt and will show up as excess power coming from your batteries when it is really coming from your charger.

I am a bit confused where you read that I have connected my charger before the shunt. (no offense )

My DIY battery is currently inside a wooden crate that can be stacked and this is the inside of the busbar box. It also holds the individual fuses for each of the 11 batteries before they join the busbar's in the center. The main positive goes into the disconnect switch and then through an 80A mega fuse before it exits at the bottom. The main negative goes from the opposite side of the busbar directly to the shunt...before it even leaves the battery! Even the BMV is inside of it and you can see it in the lower right corner.?

All the loads and solar chargers are connected to the outgoing wires so there is no possibility a charger was connected before the shunt. Besides the grid charger was not even connected at all and it was 03:40AM and completly dark outside. So not even the solar chargers were supplying any current..and if they did it would not show up in the shunt's statistics like this.

The idea of the batteries being rated slightly above what it says on the label to meet the 3000+ charge cycles makes the most sense to me. I just did not expect that it was such a high extra reserved capacity in there. I do think that they were about to shutdown because the voltage curve in VRM was dropping hard already.

Now that I know they perform well...it's really tempting to buy more of them to expand the capacity. But I am also thinking of building a new one from EVE cells and put in a better BMS with bluetooth. If money was no object I would have placed an order already for a 1000Ah battery bank, hahahah.

 

[sunshine_eggo]

Now that I know they perform well...it's really tempting to buy more of them to expand the capacity. But I am also thinking of building a new one from EVE cells and put in a better BMS with bluetooth. If money was no object I would have placed an order already for a 1000Ah battery bank, hahahah.

The more batteries you parallel makes it more difficult to balance loads between them. You've already seen disparity with your 11 batteries. More will make it worse.

If batteries are operating within spec, then they will make rated capacity; however, those that are bearing more of the load will wear faster. It may never matter for your application.

 

[newbostonconst]

I am a bit confused where you read that I have connected my charger before the shunt. (no offense )

The shunt is meant to measure everything that goes into and out of the battery. So the shunt should be the first thing that is connected to your batteries. If say you connect your wall charger directly to a battery terminal the shunt will not see that you are adding charge to the battery and will see the energy you add to charge the battery as battery power going out of the battery as more capacity. All chargers should be connected to the opposite side of the shunt as the battery connection.

I am not saying you did this, I am just providing this as a potential reason for showing extra capacity over the battery manufactures ratings.

 

[newbostonconst]

The more batteries you parallel makes it more difficult to balance loads between them. You've already seen disparity with your 11 batteries. More will make it worse.

If batteries are operating within spec, then they will make rated capacity; however, those that are bearing more of the load will wear faster. It may never matter for your application.

Not to argue but learn, I don't see how a weak battery in parallel will fail faster the more batteries in parallel you have.

Parallel batteries are balanced batteries.

Providing all connections are good and no cells are shorted and they all have the same chemistry..... then the voltage is going to be the same across all batteries and you should never put one battery at a lower voltage thus lower state of charge if all the batteries.

More batteries in parallel will increase your chances of having a bad battery but wont cause a battery to go bad IMHO.

 

[sunshine_eggo]

Not to argue but learn, I don't see how a weak battery in parallel will fail faster the more batteries in parallel you have.

Batteries that experience higher charge and discharge currents will wear/fail faster than an identical battery subjected to lower currents.

Parallel batteries are balanced batteries.

Parallel batteries are voltage equalized batteries, not balanced.

Providing all connections are good

"good" - do you also mean equal resistance?

and no cells are shorted and they all have the same chemistry..... then the voltage is going to be the same across all batteries and you should never put one battery at a lower voltage thus lower state of charge if all the batteries.

Agree.

More batteries in parallel will increase your chances of having a bad battery

Don't agree beyond the statistical probability that owning more batteries means a greater chance of having a battery failure - even if not connected to one another.

but wont cause a battery to go bad IMHO.

I never said that. I said:

The more batteries you parallel makes it more difficult to balance loads between them. You've already seen disparity with your 11 batteries. More will make it worse.

If batteries are operating within spec, then they will make rated capacity; however, those that are bearing more of the load will wear faster. It may never matter for your application.

 

[mikefitz]

self build battery bank with 11 8Ah LiFePO4 batteries

Would it not be easier and have a similar resulting 100 Ah battery using 4 off 100Ah cells or a ready built 100 Ah battery.
11 of 8Ah cells cost $528, a 100Ah battery is $359.
The disadvantage of using multiple 8Ah cells is the difficulty in current sharing. With multiple spade connetions and buss bar connections there is a greater risk of unequal volt drops and long term reliability . It would seem from your comments there may be issues in this area.

Having said that , compliments on a very neat and well constructed unit.

Mike

 

[newbostonconst]

Batteries that experience higher charge and discharge currents will wear/fail faster than an identical battery subjected to lower currents.

The more batteries in parallel the less load/lower current each battery is going to see so in your logic the more batteries is better.

Parallel batteries are voltage equalized batteries, not balanced.

If they are charged all the way they are then balanced at each charge cycle.

 

[Archerite]

The more batteries you parallel makes it more difficult to balance loads between them. You've already seen disparity with your 11 batteries. More will make it worse.

If batteries are operating within spec, then they will make rated capacity; however, those that are bearing more of the load will wear faster. It may never matter for your application.

I know I am stretching the claim of "unlimited parallel connections" made for this battery with 11 of them. I had 12 originally but during top balance charging and final measurements I noticed one pack did not fully charge to the same voltage as the others. I even discharged and recharged it on it's own but got no higher than 13.20v...while all others charge to 13.40v on average.

While I would like to have more capacity I will wait until I can afford bigger 280Ah cells and put in a better BMS with bluetooth like I said. I am currently looking what's available and the costs are high....but if done right they should last over 10 years if they are good cells right? I watched a lot of youtube video's and follow a few like OffGridGarage to see how other people build their battery.

I am not saying you did this, I am just providing this as a potential reason for showing extra capacity over the battery manufactures ratings.

Alright, then that's where the confusion came from. The way you said it sounded to me like I stated that my charger was in front of the shunt. English is not my native language so maybe that's why I understood it like this. While the instructions for the shunt make it very clear where the loads and chargers should go...it might help others. Thanks for explaining.

Would it not be easier and have a similar resulting 100 Ah battery using 4 off 100Ah cells or a ready built 100 Ah battery.
11 of 8Ah cells cost $528, a 100Ah battery is $359.
The disadvantage of using multiple 8Ah cells is the difficulty in current sharing. With multiple spade connetions and buss bar connections there is a greater risk of unequal volt drops and long term reliability . It would seem from your comments there may be issues in this area.

You are absolutely right it would be cheaper to have gone with a bigger one at the start. But I did not plan on building the battery like this with eleven 8Ah batteries. At first I upgraded my 2 x 7Ah SLA to 2 x 8Ah LiFePO4. then I got 4 extra a few weeks later as a self build power station. Then I bought two more for another experiment....and decided to put 6 in parallel in February. Since that worked well...I just bought a few more to have 12 in total! But like I said above there was one that did not charge to the same voltage before it's BMS shutdown so I only use 11 now.

I know it's not optimal this way and to prevent voltage drops and difference in resistance as much as possible I cut all wires from battery to busbar to be exactly the same length. After having made most of the wires I forgot to include the extra length on the positive side from the fuse holder and the extra leads....about 15cm I think it was. But it's the same for every battery so it's not different. I think the negatives are 30cm each so positives 45cm. It's not great but this battery was not build to stay in service like this for that long anyway.

The issues in current draw I mentioned were only measured while building the battery 6 weeks ago. I have not retested current draw per battery after that initial build and testing fase. I was looking for some kind of BMS that could work with 12V per cell but I don't think it exists right?

Having said that , compliments on a very neat and well constructed unit.

Thanks!? It took me over 4 days to rebuild the battery and crimp all those lugs on the wires. My fingers hurt for days after Here are some more pictures of it while testing and building




I just realized I never took a picture of the entire thing. It's in a cabinet now with a lot of...ehm...wire mess on top of it that I don't dare putting up here.

I know it got out of hand a little...and I won't be expanding this battery with more cells. I will go on a hunt to find EVE cells for a good price somewhere from a reliable seller.

 

[newbostonconst]

I know I am stretching the claim of "unlimited parallel connections" made for this battery with 11 of them. I had 12 originally but during top balance charging and final measurements I noticed one pack did not fully charge to the same voltage as the others. I even discharged and recharged it on it's own but got no higher than 13.20v...while all others charge to 13.40v on average.

On parallel batteries if there is different voltages at some batteries then there is a bad connection somewhere.
Nice setup....

 

[Ampster]

The more batteries you parallel makes it more difficult to balance loads between them.

I am not sure I understand the physics of what you are saying. I do agree with @newbostonconst that if there is a voltage difference it most likely is a bad connection. More connections increase the probability of one being bad.

 

[sunshine_eggo]

The more batteries in parallel the less load/lower current each battery is going to see so in your logic the more batteries is better.

again:

If batteries are operating within spec, then they will make rated capacity; however, those that are bearing more of the load will wear faster. It may never matter for your application.

If they are charged all the way they are then balanced at each charge cycle.

The above, which I can agree with, is different than your prior statement:

Parallel batteries are balanced batteries.

 

[sunshine_eggo]

I am not sure I understand the physics of what you are saying. I do agree with @newbostonconst that if there is a voltage difference it most likely is a bad connection. More connections increase the probability of one being bad.

How you parallel batteries matters, even with perfect connections:

Calculation of parallel string battery currents

Hi, everybody! I'm a retired EE. I didn't want my EE skills to fade away after retirment, so I exercise them when I can. Whenever I see a non-trivial circuit analysis problem on one of various forums I frequent, I challenge myself to solve it. Recently I watched Will's video about current...

diysolarforum.com

 

[Archerite]

On parallel batteries if there is different voltages at some batteries then there is a bad connection somewhere.
Nice setup....

It's actually the BMS inside of these 8Ah batteries that cut of charging at some point. And this one single battery cut's off too soon and drops to 13.2V after an hour with nothing connected. Like I said I did a discharge and recharge cycle on it and it was the same.
However....it was part of an earlier 4-6 battery bank and maybe it had a bad connection at the time that "damaged" it somehow. maybe.

Thanks ?

I do agree with @newbostonconst that if there is a voltage difference it most likely is a bad connection. More connections increase the probability of one being bad.

I know more connections increase the risks of one being bad and I think I made every attempt at minimizing them. The wires are all brand new 100% marine grade copper from a local store and not the cheap stuff from amazon. But with all this discussion of the issue I am really curious now how my individual batteries are doing. They are still charging up to 100% because it has been cloudy the last couple of days, so I connected the grid charger. Right now it's at 93% and the batterybank voltage is 13.68V meaning about 2 more hours with my slow charger.

I think I will keep it connected over night to let the batteries balance themself. After that I will disconnect everything and measure the voltage of the individual cells after also removing the fuses to see if any of them are lower than the rest of them. They should in theory be in sync now after daily use for 6 weeks I think.

 

[Ampster]

How you parallel batteries matters, even with perfect connections:

Okay, I get that part. I was only thinking of voltage and even that can vary slightly under load if internal resistance is different. At the end of the day the voltages will be close to equal subject to the resistance differences of the connections. Clearly a Tesla battery pack with over 60 parallel cells has mitigated most issue.

 

[sunshine_eggo]

Okay, I get that part. I was only thinking of voltage and even that can vary slightly under load if internal resistance is different. At the end of the day the voltages will be close to equal subject to the resistance differences of the connections. Clearly a Tesla battery pack with over 60 parallel cells has mitigated most issue.

when talking very small resistances, even very nearly equal voltages like 0.001V could see 5A difference between parallel cells.

I = V / R = 0.001V/.0002Ω = 5A

Even with great connection quality there are cell variations and wiring variations. The more batteries you put in parallel, the more variable the resistance... even with perfect connections. it's why Victron's wiring unlimited and most, if not all, battery manufacturers, set an upper limit to the number of recommended parallel strings.

To clarify, I'm not saying the OP is doing something wrong. It sounds like a non-issue for his application, but for others reading this thread, they need to know that 11P batteries are a wee bit outside best practices and have increased sensitivity to varying resistance. This has a potential impact on battery life if the batteries end up operating outside ratings.

 

[Archerite]

It may never matter for your application.

Even with great connection quality there are cell variations and wiring variations. The more batteries you put in parallel, the more variable the resistance... even with perfect connections. it's why Victron's wiring unlimited and most, if not all, battery manufacturers, set an upper limit to the number of recommended parallel strings.

To clarify, I'm not saying the OP is doing something wrong. It sounds like a non-issue for his application, but for others reading this thread, they need to know that 11P batteries are a wee bit outside best practices and have increased sensitivity to varying resistance. This has a potential impact on battery life if the batteries end up operating outside ratings.

I know that in nearly all examples and documentation it says to not put more than four batteries in parallel or series. Even Eco-Worthy says it for their bigger capacity batteries that have a yellow label. But these one's I got have a blue label and on their site, documentation and amazon (where I bought them) it clearly says the parallel connection is unlimited. I am not disagreeing with you on that resistance in the wires and connections or even the batteries themself is causing a difference. Just saying that the only reason I even thought it "might" be ok to do this is their specific claim of unlimited parallel connected batteries. I just thought it might be some kind of limitation on the BMS for other batteries or just because these are relatively small.

But to respond directly on my application for this battery, I indeed do not think it's a big issue. The batterybank needs to last about 6-12 months with daily usage to about 40% SoC. By that time I am hoping to have bigger 3.2V cells of 280-310Ah and build a bigger better battery. After that these will probably either go on a shelf or serve as experimental 24V or 48V setups.
It's an offgrid setup charged by solar (300watts peak) and I use it mostly to charge phone's, tablet's laptops and other small devices. Before my batterybank was big enough I had a stack of powerbanks that I charged from solar and then used that for charging the devices. I still keep them charged up from solar in case there are clouds for a few days, so I do not need to drain the main battery to charge my devices. Then I recharge them all at once when there is a lot of sun again to have a higher load.

I am also experimenting with inverters to 220V that I use to recharge my Bosch power tool batteries from. Since my "workbench/office" is 20 meters away from the batterybank I pulled a 4mm2 cable to that room. Near the battery is a Victron Orion Tr 12/24 so I can use more watts over the same wire...and also reduce voltage drop issues.
And then I have "car/truck power sockets" installed providing this 24V to USB chargers. I checked...and they can accept 32V in. And from the 24V I also run a Victron Phoenix inverter to run my TV from. I know it's not meant to be connected this way...but it works. And I did some nice Linux magic with socat to make the inverter show up in VRM. ?

In other words...it's mostly an experimental setup to learn from. And I have been doing that for the past 8 months and improving things here and there. Everything has either a fuse or breaker for protection of the wires. Already blew a few of them when I tried pulling more than the fuse's rating. Noting really bad happened so far...except for a cheap amp meter that went up in smoke.