LFP Operating Voltage

[sailorboy1]

I am still trying to learn the "normal" of my fairly new 400ah LFP bank. I just charged to 95% and was about 14.15V under charge at the time. I stopped and now with -7amp discharge going at 95% SOC am reading 13.3V

Is this normal as I thought it would be higher?

[ThereAndBack]

Sounds normal to me. "Resting voltage" is different than voltage curve under load.

[wholybee]

Originally Posted by sailorboy1

I am still trying to learn the "normal" of my fairly new 400ah LFP bank. I just charged to 95% and was about 14.15V under charge at the time. I stopped and now with -7amp discharge going at 95% SOC am reading 13.3V

Is this normal as I thought it would be higher?

That is normal in my experience. The largest voltage increase is that last 5%. Even if you did charge all the way to 100%, only a couple minutes of discharge would bring it to 13.3V ish.

[sailorboy1]

OK thanks. It seems under normal house operating loads my battery monitor (that I believe) gives a voltage around 13.1. In the old lead acid days I could add 0.1 to whatever my BM said and that voltage would match the SOC. I just looked at the chart for my LFP and finally noticed it goes 13.1-13.2 from 40% to 70% so in this range the same still looks to apply.

Mainly these are new inexpensive LFP drop ins and I am watching closely still and getting their feel.

[Captain Bill]

All battery monitors are not created equal. Depending on which one you have you could be looking at rounding errors. Point one volts means a lot in the LFP world, not so much in the LA world. One monitor I have will report 13.2V while another device in the charging system reports 13.29v and another reports 13.3V. It simply reflects the rounding algorithm of each device. Your battery monitor may simply be rounding down or more simply be reporting the tenths digit of the actual number. 13.3 may be actually 13.399, which can be a significant difference in the LFP world. Your numbers seem to be in line with my experience. While others may claim that voltage in LFPs doesn't sag. This has not been my experience. The difference is that LFP sags at the hundredths of a volt level, not the volt level so a small load can make gauging SOC under load when a few hundredths of a volt makes a difference between 40 and 70% and you're displaying at only the tenth of a volt level.

[sailorboy1]

Yeah. The thing is 0.1V change doesn't = a 30% state of charge change in the lead acid world like it does in lithium.

[sailorboy1]

I have a related question now after 7 weeks and since this is still a short thread I will add it here:

What voltage under "normal" house loads (being around 5-10a) are people seeing, or are comfortable seeing, on their LFP batteries?

When I first installed mine I would get up in morning and see about 13.1, this morning I am 12.9 at about 70% SOC according to my battery monitor. I feel the batteries got a pretty good charge yesterday as the solar says it got to 14.3.

Still working on learning to just be happy with them.

[sailorboy1]

never mind

It appears to be load related. When I wrote that I was -122ah and drawing about 10a at aroun12.95. Later I looked again and was -133ah and drawing only 1.5a and was 13.08

[fxykty]

You have noticed that your charging voltage is not you battery’s resting voltage, right? So even though you charge at 14.3V, the resting voltage at high SOC will likely be around 13.4V? Then take off 0.1-0.2V for voltage drop for the load. So, totally normal.

[Scubaseas]

Originally Posted by fxykty

You have noticed that your charging voltage is not you battery’s resting voltage, right? So even though you charge at 14.3V, the resting voltage at high SOC will likely be around 13.4V? Then take off 0.1-0.2V for voltage drop for the load. So, totally normal.

Don't think LiFePo4s work that way. mine don't. I can charge and the voltage is virtually the same next day if no loads were attached.

It's really hard to tell SOC via voltage alone with LFP unless they are isolated. You need to be reading in hundredths of volts and even then the charge curve is so flat it's almost pointless if measuring it live with a load on it. You can get results but only going out two digits and with zero load on the battery. Better to use a good , smart BMS or use a good battery monitor. The battery monitor make take a couple of WEEKS to learn what's going on with LFP batteries.

[fxykty]

Originally Posted by Scubaseas

Don't think LiFePo4s work that way. mine don't. I can charge and the voltage is virtually the same next day if no loads were attached.

It's really hard to tell SOC via voltage alone with LFP unless they are isolated. You need to be reading in hundredths of volts and even then the charge curve is so flat it's almost pointless if measuring it live with a load on it. You can get results but only going out two digits and with zero load on the battery. Better to use a good , smart BMS or use a good battery monitor. The battery monitor make take a couple of WEEKS to learn what's going on with LFP batteries.

2 years of full time live aboard with our LFP battery, so I guess I don’t know what I’m talking about. I simply describe what I see with our system. However, our charge voltage is only 13.8V, so maybe that changes things versus a higher charge voltage. Our BMS (a Tao) calculates 100% SOC when charge current decreases to 2.5%C for 30 seconds at the charge voltage. After charging is stopped when that tail current is reached, the battery immediately drops to around 13.4V. YMMV

Our BMS displays battery voltage to two places of decimal and cell voltage to three places. Plenty of precision. Doesn’t relate to an accurate SOC, and certainly our BMS doesn’t use voltage to calculate SOC. Nor the battery monitor (Victron), though we trust the BMS’ SOC calculation ahead of the monitor’s.

[Scubaseas]

Originally Posted by fxykty

2 years of full time live aboard with our LFP battery, so I guess I don’t know what I’m talking about. I simply describe what I see with our system. However, our charge voltage is only 13.8V, so maybe that changes things versus a higher charge voltage. Our BMS (a Tao) calculates 100% SOC when charge current decreases to 2.5%C for 30 seconds at the charge voltage. After charging is stopped when that tail current is reached, the battery immediately drops to around 13.4V. YMMV

Our BMS displays battery voltage to two places of decimal and cell voltage to three places. Plenty of precision. Doesn’t relate to an accurate SOC, and certainly our BMS doesn’t use voltage to calculate SOC. Nor the battery monitor (Victron), though we trust the BMS’ SOC calculation ahead of the monitor’s.

I've only had my LFP bank for a little over a year. Liveaboard this boat 1.5 years, last boat 6 years. My battery manufacturer (Renogy) specs 14.4V as max charge. I have a Victron Skylla charger that has a SOC display that differs from my BMS SOC that differs from my Balmar battery monitor. The Skylla will under charge (like 80-85%) SOC at around 13.85V but once every three or four weeks will put out 14.2 to 14.4V and put the SOC up to 99% or so. It's all magic to me really. I have no idea what the algorithm is for either the BMS or the battery monitor.

I guess my point is if you just slap a voltmeter on an LFP you are not going to determine the SOC accurately unless you go out two places right of zero and then it's probably not terribly accurate the more cells you have or the older the bank.

Then there's the overcoming muscle memory trying to keep the bank at 100% all the time which if I understand it is not great for an LFP bank. 60 to 80% SOC seems to be where manufacturers want you to store the batteries any thing long term.

[fxykty]

Originally Posted by Scubaseas

I've only had my LFP bank for a little over a year. Liveaboard this boat 1.5 years, last boat 6 years. My battery manufacturer (Renogy) specs 14.4V as max charge. I have a Victron Skylla charger that has a SOC display that differs from my BMS SOC that differs from my Balmar battery monitor. The Skylla will under charge (like 80-85%) SOC at around 13.85V but once every three or four weeks will put out 14.2 to 14.4V and put the SOC up to 99% or so. It's all magic to me really. I have no idea what the algorithm is for either the BMS or the battery monitor.



I guess my point is if you just slap a voltmeter on an LFP you are not going to determine the SOC accurately unless you go out two places right of zero and then it's probably not terribly accurate the more cells you have or the older the bank.



Then there's the overcoming muscle memory trying to keep the bank at 100% all the time which if I understand it is not great for an LFP bank. 60 to 80% SOC seems to be where manufacturers want you to store the batteries any thing long term.

Safest storage is around 50% SOC, which is why quality cells like Winston are sent to customers charged to 3.28V.

If your only charge source is shore power then relying on the Skylla to control charging makes sense. But if you have multiple charge sources then it makes much more sense to have your BMS control charging and control all your chargers via CANbus. Of course, for back up configure each of your chargers to charge properly in case CANbus drops out.

Drop in battery manufacturers seem to specify relatively high charging voltages, though it is a fact that any voltage over about 13.6V will fully charge an LFP battery if it’s held long enough. Hence the use of tail current to determine when SOC = 100%.

Regarding the algorithms to calculate SOC, you hopefully have one that you trust the most to be a reflection a reality. For us that’s the BMS, as it is the only component that knows what’s going at a cell level. I’m not sure if that’s the case for a drop in battery.

Continuing charging to maintain an LFP battery at or near 100% is lead acid thinking. I’m continually surprised by posters who are determined to keep their LFP battery at a continuous high SOC. Definitely not a good thing for longevity.

[s/v Jedi]

A lot of people are killing their LFP batteries with incorrect charging, without knowing how it should be charged or even looking up manufacturers instructions.

Most don’t even know terms like CC, CV, “termination voltage” etc. and don’t know how to program a battery monitor to accurately show SOC% but then still look at that readout and use it.

Also, talking about voltages while a battery is being charged or discharged is meaningless, as is the measurement itself when it isn’t known what this was measured with and exactly where it was measured.

The only recommendation for these cases is to hire a professional to analyze your installation, fix it where needed and explain it’s use as a completed system. There’s just too much to learn to know internal workings; one would need to study electrical principles first, then how Coulomb counting battery monitors work, how LFP cells charge and the effects of charge rate in relation to termination voltage etc.

[Scubaseas]

Originally Posted by fxykty

Safest storage is around 50% SOC, which is why quality cells like Winston are sent to customers charged to 3.28V.

If your only charge source is shore power then relying on the Skylla to control charging makes sense. But if you have multiple charge sources then it makes much more sense to have your BMS control charging and control all your chargers via CANbus. Of course, for back up configure each of your chargers to charge properly in case CANbus drops out.

Drop in battery manufacturers seem to specify relatively high charging voltages, though it is a fact that any voltage over about 13.6V will fully charge an LFP battery if it’s held long enough. Hence the use of tail current to determine when SOC = 100%.

Regarding the algorithms to calculate SOC, you hopefully have one that you trust the most to be a reflection a reality. For us that’s the BMS, as it is the only component that knows what’s going at a cell level. I’m not sure if that’s the case for a drop in battery.

Continuing charging to maintain an LFP battery at or near 100% is lead acid thinking. I’m continually surprised by posters who are determined to keep their LFP battery at a continuous high SOC. Definitely not a good thing for longevity.

The Renogy batteries I have do not have cells, they use multiple pouches. I have multiple charge sources, 5.8KW genset, HO alternator, Solar and a wind genny. A CAN bus would be lovely if you want to send me one. Along with compatible chargers of course. Thanks. Big expense to fix what's not broken. I'd love an entire integrated Victron system but my portfolio and sensibility won't allow it. I went with Renogy because they were reasonably priced & on sale, available at the time I needed them, easy to install with off the shelf battery holders and had Bluetooth monitoring of the cells packs to meet the then upcoming ABYC standards for LiFePo4 batteries. It works for me. Plus I already had Renogy solar controllers.

I don't particularly trust any of the algorithms or their displays but I don't obsess about my SOC either. Somewhere between 20% and 80% works for me. If the house is getting below 20% I will run the genset or the motor if I'm away from the dock and the other charge sources aren't working . I look at the SOC displays often more out of habit than necessity. Hard to break years of worrying about the 50% SOC on a LA battery bank.

I run my autopilot and sailing instruments off a separate 200AH LiFePo4 bank.