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

[Colin A]

I'm curious about this quote. And weather sitting at 3.4V would actually result in over charging. Anyone have any documentation on it. I found this link seeming to say charging would end at 98% capacity at 3.4

"It is a misconception to believe that SoC can be controlled using voltage. 3.4V will still charge a LFP cell to 100% given enough time and 3.3V will struggle to achieve 30% no matter what. The transition is so steep that it makes voltage unusable."

Charge voltage experiments with lithium iron phosphate batteries showing how capacity varies with charge voltage and higher cycle live with lower charge voltage

[SV THIRD DAY]

Originally Posted by billysurf

We are doing research on lithium batteries for a converted bus (MCI or GM) and can't get the answer to a very basic (although loaded) question. How long can you be without a power hookup. I know the boat community could answer this much faster than the RVers!

For example, running one A/C unit and typical RV needs how much battery power would you need to run a fan or 2, washing machine.dryer once every 2 days, laptop, 40" tv, directv, coffee pot for a week. Just looking for ballparks. No solar at this point. I appreciate your time!!!!

The "until it runs down" answer may seem a bit glib...but with the data you gave us it's about the best that can be said.
We/you need to know at minimum the following to answer that:
1. How big of a LiFePO4 bank in terms of Amp Hour Rating
2. How big is your power draw also in terms of Amp Hours.

Once you know that it's an easy calculation. But otherwise it's just a WAG (wild ass guess).

Now as a SWAG (Scientific Wild Ass Guess) from the sounds of the equipment you want to run, if you have a 1000AH LiFePO4 Battery bank I will give you 1/2 to 1 Day as a reasonable guess without plugging in. LiFePO4 batteries are not magic...they are just more efficient than old school Lead Acid, but a power hog is a power hog either a Boat or RV and I know if NO ONE that would plan to run a AC Cooler unit off of Batteries...no matter Lead Acid of LiFePO4...it's just not practical due to the high Amp Draw well not unless you have unlimited $$$ to throw at a battery bank. LiFePO4 batteries do not take the place of a Generator, which is what you will need for the type of loads you are talking about.

[billysurf]

Thanks for the info! The truth is I have no idea what will be running inside of the bus as we're still in the research phase but I really appreciate your educated guess! I am a voice actor and the reason I thought batteries may be my answer for longer periods of time was because I need completely silent power operation. I've never heard of a generator that's completely silent, so I just may have to except that I'm limited to locations with hook ups.

[darylat8750]

You need to make a list of the things you expect to run and number of hours each day they will run Then you may get real answers. I have a refrigerator that runs about one third of every hour when it's cool out and as much as one half the time when it's hot. TV and video player and a few fans and lights. I have an 800 amp hour bank at 12 volts made up of 8 six volt golf cart wet cell batteries. I run a diesel generator at least an hour a day. I could run these load with solar panels. Air conditioning and washing clothes would require way more batteries and solar panels than I have room or weight budget for. If you can't make a power budget you should consider getting some professional help.

[billysurf]

Thanks!

[marujo.sortudo]

If you're willing to give up A/C, TV, and W/D when you're off hookup you could go much longer, it that's a good trade off to you. Also, solar is the silent generator and could get you more time and quiet away from hookups. Just some things to weigh. Figure out the minimum you need and what's most important to you. Do the math to figure out a power budget.

[thomasow]

To KISS or not to KISS .. that is A question.

This is indeed a LONG thread, with lots of good information – providing one digs it out. I wanted to offer some thoughts on the conversation between Rich and OceanSeaSpray regarding the KISS approach (full disclosure here: I and a few others are working on open sourced DC charging devices, including a BMS, with one key feature being better full system integration and coordination. Eliminating issues brought up here and more – like the King-of-the-mountain battle between multiple charging sources, but I degrees)

KISS – Maybe there is a bit of confusion here in it that setting battery terminal voltage a little below the max (ala 3.4vpc) will protect batteries from overcharging. This is not true, as OceanSeaSpray has pointed out. No matter what the voltage is, if energy flowing into a battery exceeds that dissipated by internal consumption of the battery (Almost nill in LiFeP04 technology), the SOC will continue to rise. And with Li based batteries this will eventually cause undesirable consequences – the least of which would be shortening battery life. So in this way, OceanSeaSpray is correct, the KISS approach has some risk.



But what seems to be missing is stepping back to see how these batteries are actually being used. In Rich’s case (and I suspect the others he has helped with the KISS approach??), batteries are not camped at the charge voltage indefinitely – slowly gaining SOC. Instead they see a daily charge/discharge cycle – discharging over the night and hence interrupting the slow SOC creep up.

However, place those batteries on shore power over the winter and the story changes – one is back to the slowly raising SOC situation. Further, without any BMS to actively protect from over discharging, any unnoticed failure (say, leaving the boat unattended and the shore power cord gets unplugged) could lead to damaged batteries as well.

So yes, I can fully see how the KISS approach works – in a real world balanced system, where one is also able to monitor manually for issues. Nice and simply.



It has been education and at times entertaining to read all the postings here, including the side tracks. But I think it is useful to consider the WHOLE picture when making assessments of different approaches. Both for validation, but also to gain insight into the weaknesses as well.

[SV THIRD DAY]

Originally Posted by thomasow

To KISS or not to KISS .. that is A question.

This is indeed a LONG thread, with lots of good information – providing one digs it out. I wanted to offer some thoughts on the conversation between Rich and OceanSeaSpray regarding the KISS approach (full disclosure here: I and a few others are working on open sourced DC charging devices, including a BMS, with one key feature being better full system integration and coordination. Eliminating issues brought up here and more – like the King-of-the-mountain battle between multiple charging sources, but I degrees)

KISS – Maybe there is a bit of confusion here in it that setting battery terminal voltage a little below the max (ala 3.4vpc) will protect batteries from overcharging. This is not true, as OceanSeaSpray has pointed out. No matter what the voltage is, if energy flowing into a battery exceeds that dissipated by internal consumption of the battery (Almost nill in LiFeP04 technology), the SOC will continue to rise. And with Li based batteries this will eventually cause undesirable consequences – the least of which would be shortening battery life. So in this way, OceanSeaSpray is correct, the KISS approach has some risk.



But what seems to be missing is stepping back to see how these batteries are actually being used. In Rich’s case (and I suspect the others he has helped with the KISS approach??), batteries are not camped at the charge voltage indefinitely – slowly gaining SOC. Instead they see a daily charge/discharge cycle – discharging over the night and hence interrupting the slow SOC creep up.

However, place those batteries on shore power over the winter and the story changes – one is back to the slowly raising SOC situation. Further, without any BMS to actively protect from over discharging, any unnoticed failure (say, leaving the boat unattended and the shore power cord gets unplugged) could lead to damaged batteries as well.

So yes, I can fully see how the KISS approach works – in a real world balanced system, where one is also able to monitor manually for issues. Nice and simply.



It has been education and at times entertaining to read all the postings here, including the side tracks. But I think it is useful to consider the WHOLE picture when making assessments of different approaches. Both for validation, but also to gain insight into the weaknesses as well.

Now that's not fair...you can't boil it all down in one post to let the others off the hook in reading all hundreds of pages...heck...that's where the "fun" comes into this LiFePO4 game isn't it...ha ha ha

Ah but....(isn't there always an "Ah...but") has anyone ever ran a test to tell us just how long it would take a LiFePO4 battery bank to "overcharge" when set at a lower cell voltage as the KISS approach does? Once you do that test....the worry/panic/hair-on-fire goes away, at least it did for me anyway. It's where not relating the scientific data to the real world forces you into designing for a Magnitude 15 earthquake on your outhouse...

The Cruisers I know with the KISS LiFePO4 approach just turn off their charging sources when they leave their boat for the summer for example in San Carlos, MX.

[ColdEH]

I agree with Rich, KISS.
I use one charging source, my purpose built generator for the batteries . No solar , no shore power charger . If i plug into shore power it would be just for a small heater, never to charge my batteries , I do that myself. If i have to leave the boat for a week, no problem fridge and freezer are the only thing left running and that will go for ten days without a charge . I only have a low voltage cutoff , no high voltage cutoff. I am hoping to get many many years out of my batteries.

KISS

Regards

[Colin A]

Originally Posted by SV THIRD DAY

Now that's not fair...you can't boil it all down in one post to let the others off the hook in reading all hundreds of pages...heck...that's where the "fun" comes into this LiFePO4 game isn't it...ha ha ha

Ah but....(isn't there always an "Ah...but") has anyone ever ran a test to tell us just how long it would take a LiFePO4 battery bank to "overcharge" when set at a lower cell voltage as the KISS approach does? Once you do that test....the worry/panic/hair-on-fire goes away, at least it did for me anyway. It's where not relating the scientific data to the real world forces you into designing for a Magnitude 15 earthquake on your outhouse...

The Cruisers I know with the KISS LiFePO4 approach just turn off their charging sources when they leave their boat for the summer for example in San Carlos, MX.

Rich, I'm curious how long you think they will go. I ask because I'm on a different use model my boat and the one's I crew on/use, are all weekend (or week long vacation) boats that use solar to charge between uses. I wonder if it would be a problem in that situation to have it sit a few weeks with solar on at 3.4 per cell. Currently one boat has gel the other golf cart batteries. The gel boat uses only solar (outboard powered no alternator) and that would be the one I would like to experiment on.

It's funny how much conflicting info is out there.

[Maine Sail]

Originally Posted by SV THIRD DAY

Ah but....(isn't there always an "Ah...but") has anyone ever ran a test to tell us just how long it would take a LiFePO4 battery bank to "overcharge" when set at a lower cell voltage as the KISS approach does?

I have done that test.

4S X 100AH CALB SE Cells (one of my testing packs)

- Pack consistently testing at 101.2 - 101.3Ah. Capacity test proceeding this test was 101.3Ah

- Pack was charged to 13.8V until current dropped to 15A

-Voltage then dropped to 13.6V until current went to 0.00A

-The 13.6V & 0.00A had settled in on a brief split second "blip" of current about every 30 minutes when I turned it off.

-The time to 0.00A, from when voltage was dropped to 13.6V, at 15A & 13.8V, took about 16-17 hours. (Reduced voltage at 3:03PM, measured amperage at 9:17pm and then again at 3:24 am then again at 7:00 am. By 7:00 am it was 13.6V & 0.00A but I don't know exactly when it the power supply began its on/off between 0.1A and 0.00A....

-Ran initial capacity test on pack and it delivered 101.2Ah this is a 100% full charge battery for this battery no matter whether I charge it to 14.4V or 13.8V. At 13.8V and 15A of tail current this 100Ah pack is not full. At 13.8V the current needs to tail off towards 0.02C or about 2-3A for it to deliver the AH capacity. The duration of you hold constant voltage is what drives up the Ah capacity. Lower voltages just require longer CV periods to attain 100% SOC. Once an LFP is "full", at a constant voltage, there is really nowhere else for excess charge energy to go.

-I then ran the same test only this time charged to 14.4V and 15A then dropped to 13.6V to 0.00A. Capacity test results were identical.

-I have also charged this pack only to 13.6V (no higher voltage then a drop back to 13.6V) and let current go to 0.00A then run a capacity test. It still delivers 100% it just takes longer to get to 0.00A.

That said at 13.8V or 13.6V or even a 14.0V charge none of the cells have strayed voltage wise. Once I push to 14.2V to 14.4V is where I start to see some cell divergence.

My observation is that yes you can get 100% of the Ah capacity at 13.6V - 14.0V but the cells don't tend to diverge in voltage, even at 100% SOC. However once you start to push beyond 14.0V you can start to see upper knee voltage spreads. What I can't really say is how much damage floating these cells at 13.6V (3.4 VPC) does in terms of cycle life. We know from all the literature that floating and storage at higher voltages does degrade the cells but quantifying this is hard because cell temp also plays into the level & speed of cell degradation..

I have a lot of data I have recorded with my test packs, in chock full notebooks, and someday, when I get more time, I hope to write about these tests. This test is an n=1 but should be able to be extrapolated fairly well to other LFP prismatic cells. I conducted this to see how long a shore charger at 13.6V it would take for the batteries to become fully charged.

This is all a result of duration at CV. Charge to 13.8V and stop, not full. Charge to 13.8V and 0.02C full etc.... The higher the voltage the less time it needs to be held constant, the lower the voltage the longer it needs to be held but even 13.6V seems to get there on this n=1 LiFePo4 pack.

[ColdEH]

Hey MaineSail great data , as usual . My charging is done with constant voltage only , my system dosas not allow me to tail off the current at all . I charge to 14v at full 200 amp current and then shut down . The alternator gives full capacity to whatever set point I choose . I have not noticed any drift in cell balance or loss of capacity. It would be nice to have data about what happens to these cells if they are kept at a full state of charge and the impact of that on there longevity.

Regards

[SV THIRD DAY]

Originally Posted by Maine Sail

I have done that test.

4S X 100AH CALB SE Cells (one of my testing packs)

- Pack consistently testing at 101.2 - 101.3Ah. Capacity test proceeding this test was 101.3Ah

- Pack was charged to 13.8V until current dropped to 15A

-Voltage then dropped to 13.6V until current went to 0.00A

-The 13.6V & 0.00A had settled in on a brief split second "blip" of current about every 30 minutes when I turned it off.

-The time to 0.00A, from when voltage was dropped to 13.6V, at 15A & 13.8V, took about 16-17 hours. (Reduced voltage at 3:03PM, measured amperage at 9:17pm and then again at 3:24 am then again at 7:00 am. By 7:00 am it was 13.6V & 0.00A but I don't know exactly when it the power supply began its on/off between 0.1A and 0.00A....

-Ran initial capacity test on pack and it delivered 101.2Ah this is a 100% full charge battery for this battery no matter whether I charge it to 14.4V or 13.8V. At 13.8V and 15A of tail current this 100Ah pack is not full. At 13.8V the current needs to tail off towards 0.02C or about 2-3A for it to deliver the AH capacity. The duration of you hold constant voltage is what drives up the Ah capacity. Lower voltages just require longer CV periods to attain 100% SOC. Once an LFP is "full", at a constant voltage, there is really nowhere else for excess charge energy to go.

-I then ran the same test only this time charged to 14.4V and 15A then dropped to 13.6V to 0.00A. Capacity test results were identical.

-I have also charged this pack only to 13.6V (no higher voltage then a drop back to 13.6V) and let current go to 0.00A then run a capacity test. It still delivers 100% it just takes longer to get to 0.00A.

That said at 13.8V or 13.6V or even a 14.0V charge none of the cells have strayed voltage wise. Once I push to 14.2V to 14.4V is where I start to see some cell divergence.

My observation is that yes you can get 100% of the Ah capacity at 13.6V - 14.0V but the cells don't tend to diverge in voltage, even at 100% SOC. However once you start to push beyond 14.0V you can start to see upper knee voltage spreads. What I can't really say is how much damage floating these cells at 13.6V (3.4 VPC) does in terms of cycle life. We know from all the literature that floating and storage at higher voltages does degrade the cells but quantifying this is hard because cell temp also plays into the level & speed of cell degradation..

I have a lot of data I have recorded with my test packs, in chock full notebooks, and someday, when I get more time, I hope to write about these tests. This test is an n=1 but should be able to be extrapolated fairly well to other LFP prismatic cells. I conducted this to see how long a shore charger at 13.6V it would take for the batteries to become fully charged.

This is all a result of duration at CV. Charge to 13.8V and stop, not full. Charge to 13.8V and 0.02C full etc.... The higher the voltage the less time it needs to be held constant, the lower the voltage the longer it needs to be held but even 13.6V seems to get there on this n=1 LiFePo4 pack.

Great info Amigo...
It always nice to get a reality check from someone else with similar data.
Living on the boat full time I'm always going into a discharge cycle when the sun sets and my solar output stops. I have left the boat for several periods of 5 days with no loads on and when I returned my Amp counters showed no overcharge condition. For both of those times, I turned my solar controller down from the normal voltage setpoint of 13.6 to 13.4v. Have I left the bank for 120 Days at 13.4v on my solar controller? No...but then again I never will also so for my application it's just not something I worry about. If I did have to do that...I would just turn off the solar.

Again....great DATA.

[OceanSeaSpray]

There is data showing capacity fade as a result of storage at 100% SOC (and lower)and it is slow even at 30degC. It accelerates greatly with temperature at any SOC.

https://hal.archives-ouvertes.fr/UNI...E/hal-00876555

However, this relates to cells that were charged and then stored, not kept at a charging voltage indefinitely (i.e. continuous overcharging).

I had run similar tests as Maine Sail using solar charging down to 13.6V, but without quantifying capacity each time. The battery went to zero current acceptance each time (i.e. full). Charging only got more and more inefficient as voltage dropped. It might not have mattered as much if the sky was blue and the sun shining every day, but we have a winter here.

Another test was varying absorption time with alternator charging (say 0.35-0.50C current). No absorption was an unmitigated disaster as the bank couldn't be charged past about 75% SOC. We were heading into winter then and that installation needed to make the most of engine running time, because solar wasn't enough to offset consumption.

The only strategy that delivered solid, consistent results throughout the year and seasons has been Charge/Absorb/Terminate and then just make up for consumption to use any energy available and extend discharge cycles until it is time to recharge again.

[roetter]

Hurricane Matthew and some other things together killed my 2P4S 700Ah Winston cells. Still have some capacity, but had to remove some badly bulged cells.

Where can I get prismatic cells - preferably 400ah - at a decent price and quick at the moment?