diy solar

diy solar

How to maintain 50% SOC when away from the boat?

Sabre602

Putterer/Tinkerer
Joined
Sep 2, 2021
Messages
5
Location
Northwest Washington State
I am preparing to assemble a lithium house battery bank on my boat and have questions about how to maintain the bank at the correct state of charge to optimize battery life. We will have a 12v, 560 ah bank using 280 ah cells in a 4s2p configuration. 600 watts of solar cells and a large frame 80 amp alternator will both supply a Kisae DMT1250 50 amp DC-DC charger that incorporates a MPPT solar controller. A Sterling alternator protection device will guard against BMS cutoff. Shore power charging will come through a Samlex Evo 1212F-HW 1200 watt inverter with 60 amp charger. We virtually never use AC power aboard...I can't recall the last time I switched on our present inverter...so we don't need more inverter power than that. The reefer is AC/DC so doesn't need the inverter. Everything will be appropriately fused, of course.

I haven't selected the BMS system yet, but am favoring a 100 to 200 amp BMS with a plan to bypass the BMS with a contacter for occasional high current draw with the anchor windlass.

Like most boats, it spends most of the time at the dock. I know that some folks will entirely disconnect the battery at 50% SOC, while others have mentioned setting their electronics to cycle the house bank between 30 to 50% SOC. I would love to get some ideas about how to automate limiting the state of charge. I don't want to have to keep track of when to manually switch the bank on and off. My wife enjoys being aboard even when I'm at work or out of town, and she needs it automated.

Thanks in advance for your experience and suggestions!
 
Two ways:
1) Set the BMS to turn the charger off at a specific voltage. Cheap Chinese BMS's might not allow this...

2) Set the charge controller to stop charging at a specific voltage.
 
How to know SOC, when somewhere between the knees?

The only sure way I can come up with is to occasionally charge or discharge until a knee is reached, get cells re-equalized again, then move back to estimated 50% point. It's not like you can use a hygrometer.
 
Two ways:
1) Set the BMS to turn the charger off at a specific voltage. Cheap Chinese BMS's might not allow this...

2) Set the charge controller to stop charging at a specific voltage.
Yeah, so this is what would make sense for lead acid, where voltage can be used to show SOC, but how does this work with the wide, almost flat voltage curve "between the knees?"
 
Yeah, so this is what would make sense for lead acid, where voltage can be used to show SOC, but how does this work with the wide, almost flat voltage curve "between the knees?"
Voltage between the knees is not flat.. just flatter than most other chemistries.

A lithium ion battery in the 3.65v range is at around 50%
A LiFePO4 cell at 3.2 is at around 50%

Any decent multimeter will give you three digits.

Lithium Ion, Lithium Phosphate, and lead acid, are all very predictable.. voltage = a good approximation of SOC.. This doesn't always apply to the top and bottom ends near those knees, but it very much does in the middle.

If you get into a NiCad or a NIMH battery, then the problem becomes far more apparent.. They'll start out at 1.4, deliver a bit of juice, drop to 1.3, deliver a bit more juice, then drop to 1.2 and sit there until you think you're meter is stuck.. Then they drop to 1.1 and and sink fast. Its as if they have three knees and a hump...

But for lithium, just put the battery at its nominal voltage (3.2 or 3.65) and leave it.
 
Voltage between the knees is not flat.. just flatter than most other chemistries.

A lithium ion battery in the 3.65v range is at around 50%
A LiFePO4 cell at 3.2 is at around 50%

Any decent multimeter will give you three digits.

Lithium Ion, Lithium Phosphate, and lead acid, are all very predictable.. voltage = a good approximation of SOC.. This doesn't always apply to the top and bottom ends near those knees, but it very much does in the middle.

If you get into a NiCad or a NIMH battery, then the problem becomes far more apparent.. They'll start out at 1.4, deliver a bit of juice, drop to 1.3, deliver a bit more juice, then drop to 1.2 and sit there until you think you're meter is stuck.. Then they drop to 1.1 and and sink fast. Its as if they have three knees and a hump...

But for lithium, just put the battery at its nominal voltage (3.2 or 3.65) and leave it.
Sorry but state of charge is only accurately measured by a current monitor with a shunt. Like a Victron BMV or similar.

Using voltage doesn't work. Are you just getting off a charge cycle? or were the batteries idle for days? or were you just discharging? or are they colder then a day ago? or are they currently charging? or are you currently discharging?....this all causes the voltage to move giving you a false state of charge if using voltage.

Voltage does not work.....doesn't even get you close.
 
I would be interested in how people have stored at 50% also. I think you’d need a shunt to measure down to 50%, but would need to totally disconnect the batteries after that.

THat’s one of those things that is in all the data sheets that seems a bit hard to judge.

A shunt can accurately measure a battery down to the 50% point, but parasitic drain can take it down in a way the shunt may not measure. When storing my 12 volt 458 ah lead acid battery bank for two months, I left the shunt on and removed everything else from the system. After two months, the SOC read 100%, but the batteries charged about 40% of the battery. The milli amps the shunt uses adds up.
 
Provided the chargers have provision for user setting of charge voltage, using voltage will work. On my boat and the RV setting the charge voltages between 13.2 and 13.3 for a 12v system will hold a SOC in the 50% to 60% region. Depending on the calibration accuracy of the chargers the actual voltage setting may have to be determined experimentally. Charge the battery pack to full, set the charger to 13.2 volts. Set the battery monitor to 100%. Sit back and see how the system performs.
A charge to full capacity will be needed periodically to remove and possible memory effect caused.

Mike
 
Is it actually important for shelf life to have lithium at a particular SoC? Or is it voltage which is important, in terms of degradation? If voltage of all cells was kept between the knees would that be good enough? We understand degradation is much faster if held at voltage extremes.
 
For me, the spec sheet says store at 50%. So I’m planning on how to do this for several months of storage in sunny AZ.
 
Sorry but state of charge is only accurately measured by a current monitor with a shunt. Like a Victron BMV or similar.

Using voltage doesn't work. Are you just getting off a charge cycle? or were the batteries idle for days? or were you just discharging? or are they colder then a day ago? or are they currently charging? or are you currently discharging?....this all causes the voltage to move giving you a false state of charge if using voltage.

Voltage does not work.....doesn't even get you close.
Is it actually important for shelf life to have lithium at a particular SoC? Or is it voltage which is important, in terms of degradation? If voltage of all cells was kept between the knees would that be good enough? We understand degradation is much faster if held at voltage extremes.

You two should watch this video.. this is where I learned.
 
Sorry but state of charge is only accurately measured by a current monitor with a shunt. Like a Victron BMV or similar.

Using voltage doesn't work. Are you just getting off a charge cycle? or were the batteries idle for days? or were you just discharging? or are they colder then a day ago? or are they currently charging? or are you currently discharging?....this all causes the voltage to move giving you a false state of charge if using voltage.

Voltage does not work.....doesn't even get you close.
Check out the video I posted..
Addressing your comments: 1) For storage purposes, we don't need "accurately measured", we just need to set the battery's state of charge somewhere in the middle near its nominal voltage for storage.

Using voltage does work, in fact, in the application we are discussing, it works extremely well. This wouldn't be true if you were charging or discharging a cell at or near its maximum C rate.. as C rate increases, so does charge/discharge hysteresis.. Many experience and notice this hysteresis when discharging a battery near its limits.... and we refer to it as "voltage sag" under load. (Talk to the lead acid people! LOL) That same thing happens during charging only it reverses.. Charge a battery at a high C rate and the voltage will settle and drop after the charge is removed.

Hysteresis is a big thing for small batteries working hard.. its a big thing for an EV trying to accelerate a 4000 lb car to 70mph, but not so much for solar storage batteries in general. A 120 amp hour battery being discharged to storage voltage over the course of 10-20 hours isn't going to have much voltage sag.. could probably be measured in the <50mV range.

The problem in your statement is that the poster is looking for storage voltage and you're giving him an answer related to determining an accurate state of charge.. When a battery goes into storage, it doesn't care if its SOC is 40% or 60%.. what it cares about is that its not at high voltage, or such a low voltage that self-discharge will drain it dry.

The fact is, the lower the storage voltage, the less damage to the battery.. the problem is that if you go too low, the cell's own self discharge could drop it below the chemistry's floor and cause damage.

So the trick is to store the battery at the lowest voltage possible, but no so low it falls off the cliff.

Again, watch that video.. its like a mini-college education on lithium batteries.. and I'll also say its the most informative lithium related hour of learning I have ever had.
 
First, are there any loads involved? The best way is probably to turn everything off, and then disconnect everything, including the BMS, from the cells.
Second, understand that keeping at exactly 50% is difficult, and not necessary desirable. There is some research that shows that micro cycling, keeping the battery very near a specific voltage/SOC, while small loads and small charge currents maintain that, can create a memory effect in LifePO4.
If the battery is kept anywhere between 20%-80% you basically eliminate all of the concerns with storing the battery fully charged. There isn't likely any difference between letting the battery be at any voltage within that range. The life of well cared for LifePO4 batteries is so long, I don't think there is any evidence that lower voltages really result in a longer life-only that keeping it fully charged can reduce it slightly. Case in point, there are many LifePo4 batteries in boats that are 10+ years old, and still test at full capacity as new. They were very well cared for, but not stored at 50% The big thing is they were not charged at held at 3.65V per cell while not used.

Options you have:
  • If this is a boat sitting at a dock that you will visit and take out on weekends just turn off the charger and all of the loads on the boat. The self discharge rate on LifePO4 is quite low. If you don't have it powering anything, it should be good a couple months that way. Most certainly, it will last a few weeks. For comparison, I live aboard and have a 300Ah bank. Even with small loads, 12V LED lights etc. I can go a couple weeks between charging. Most stuff is A/C power from shore, and I only charge when I need to. So unless you are putting it into long term (months to years) storage, you probably don't need to do anything other than turn everything off.
  • You can set your charger voltage, both absorb and float, to around 3.2V per cell. Small currents will go in and out, and SOC (if you have a meter and can monitor it) will wander around a small range. You can experiment to find the exact voltage will yield the SOC you want. I have read concerns about micro cycling causing a memory effect. I don't know how valid those are, or if they would be negated by your weekly use of the boat and full charge at that time. I did this when I first installed mine, until I read of the micro cycling memory effect. It works very well to keep the battery at ~ 50%. I don't know if the micro cycling concerns are valid.
  • You can use the relay outputs like on a Victron SOC meter to control a charger, to let it cycle throughout a range that you can set. That could be 20%-80%, or 49%-51%. However, without fully charging often, the meter will lose accuracy. So you could end up overcharging the battery despite the meter thinking it is at 50%. You could let it charge to 100%, to let the batteries balance and the SOC meter resync, and because you don't hold it there and let it deeply discharge, there shouldn't be a loss of battery life expectancy.
 
For me, the spec sheet says store at 50%. So I’m planning on how to do this for several months of storage in sunny AZ.
For long term storage its easy. Just disconnect everything, including the BMS. Tougher is where there is stuff powered on and working, and you will be away for a long time.
 
For long term storage its easy. Just disconnect everything, including the BMS. Tougher is where there is stuff powered on and working, and you will be away for a long time.
The other part I’m leaving out is besisea the SOC at 50%, this is an RV builds where storage temps get to 115. To install the battery I have a slide out battery tray and a ladder. Removing those batteries every summer is a lot of work. !00 LBS a battery up and down a ladder sounds like a lot of work. Even more if the pack needs to come apart.

It’ll be very tempting to leave them in for the summer.
 
I haven't selected the BMS system yet, but am favoring a 100 to 200 amp BMS with a plan to bypass the BMS with a contacter for occasional high current draw with the anchor windlass.
To address the windlass connection. If you bypass the BMS, then the BMS will be unable to track SOC. You will draw the battery down with windlass use, and the BMS will still report it as full (or whatever it was before windlass use) As long as you know that, it isn't a huge concern. A 200A BMS should support a typical windlass. I know of at least one boat using a 100A BMS with a windlass.
 
The other part I’m leaving out is besisea the SOC at 50%, this is an RV builds where storage temps get to 115. To install the battery I have a slide out battery tray and a ladder. Removing those batteries every summer is a lot of work. !00 LBS a battery up and down a ladder sounds like a lot of work. Even more if the pack needs to come apart.

It’ll be very tempting to leave them in for the summer.
I cruise in the tropics, where my batteries will be in use in 100+ temps. In the battery compartment it could be even warmer. It isn't good for them, but it is what it is. You can either do the hard work, or accept the storage isn't ideal.
Maybe find another use for the pack during the hot summer? Remove it, take it to cooler climate, and put it to use. Storage problem solved. :)
 
Ok, wow huge range of opinions on this one…
Here is mine!

This is for LiFePO4…
First, there is no memory effect…
Second, if you want an easy way to allow the system to stay connected, set the bulk to 13.2/12V set, and set the float to 12.8 done.
If you leave, the cells never hit 100%, solar can maintain phantom and self discharge, and the bank will float around 50% perfect for months of storage, and when you return, change the settings back where ya like.
 
Remove it, take it to cooler climate, and
That sounds like a great idea. Every work from the web gig I look at requires me back here too much or gets me less than minimum wage.

Honestly probably will just accept the risk of leaving them in. It’s not exactly risk free disassembling and moving them twice a year.

Only way I truly see me removing these is if I can find an ABS case for an 8S 280 ah bank, and I’ve got another thread on that with no results.

 
You two should watch this video.. this is where I learned.
This video is great. Thanks for posting it. Really appreciate it. Seriously.

Active Thermal Management all the way. It's more complicated, but if I insulate my LiFePO4 battery cell mass, and be careful about failure modes, my hope is that keeping them consistently below 25 deg C will reduce calendar aging, extending usable operating lifetime. Trying to include a per cell lifetime histogram of temperature vs time spent at that temperature, voltage vs time spent at that voltage, etc. in a logging module to additionally assist with cell life analysis going forward. Just have a loop every one second, add one second to the bin of the histogram that represents current temperature and voltage.

Counting cycles is also a topic. I would like to additionally be able to monitor how many Ah charged and discharged at a given voltage and given temperature. These are the critical aging bits that I want to see the data from my batteries in a year or more.
First, there is no memory effect…
From what I have read, there is a memory effect, but that the magnitude of the memory effect is so small (<2% if my hazy memory is operating nominally) that I feel sort of pedantic pointing it out, and feel that you're colloquially accurate in assessing it as no memory effect. 50+2 or 50-2 is still only 48-52% which does not meaningfully affect the critical mission outcome of maintaining middle SOC to encourage longer battery life.
 
Back
Top