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

[Maine Sail]

Originally Posted by ebaugh

Balmar's 614 or ARS regulator will work with LFP since they have custom charging profiles. But so will many others.

But either way, the chargers need to charge to your target voltage of 14.0-14.4V, then stop charging by regulating at 13.6 or less. Held long enough, anything over 13.6, will overcharge.

The Balmar MC-614 works very well with LiFePO4. You can change the voltages such as bulk 1, absorption and float (we currently charge to 13.9V bulk, then drop back to 13.8V absorption, then 13.6V float), turn off temp sensing, control field % if you want to manually protect the alt or belts (we are set on belt manager 3 which keeps the alt cool & out of temp limiting range), you can adjust alternator temp sensing temps, you can shorten or lengthen the time it stays at voltage, you can reduce voltage when the bank is near full and even use a voltage below resting voltage to essentially turn the alt off. You can also interrupt the B+ wire with HVC relays or a manual alt cut switch.

There are not many voltage regulators on the market that are as controllable/adjustable as the Balmar MC-614. Just be 100% sure your voltage sense wire is actually sensing the physical + battery post, & only the battery post, and that you do not use the battery temp sensors.. The - wire for the reg should ideally also run directly back to the battery.

If you drop back to 13.6V +/-, at about 96-98% SOC, it takes a very long time before the bank is actually full. We recalibrate our battery monitor at 5A acceptance/13.8V.

We also have a manual alt cut switch. Because these banks charge so fast we often don't need any charging when motoring. I find it is easier to just flip the alt off manually and leave the solar panel going. Our solar controller charges to 13.7V, in absorption, but it rarely gets there unless we've recently charged to full with the motor. We do not leave solar, shore charging charging or any loads on when we leave the boat.

[pz5r9n]

A bit of topic but this price seems to put these batteries within range?
Rechargeable Batteries & Battery Chargers - 3.2V 100Ah LiFePO4 (Lithium Iron Phosphate) Rechargeable Battery

[caipirinha]

Hello,

I've done it ! (Read the whole thread, that is). I still have a few questions.

Current status :
Both house banks (Exide gel, 24 V 230Ah) are dead
Charging systems are Tecsup charger (to be replaced, see below) and Leece-Neville alternators, shaft driven, switch-activated field, external HeartInterface regulators on both engines (boat is a catamaran) and 24V Fisher-Panda generator (will probably be removed) connected to banks via Surepower 1202
Monitoring with one Xantrex Link10 for each bank

Planned change :
Replace Tecsup charger by Victron Multiplus 24-3000-70-16 (done but not yet connected)
Replace house batteries by LFP (questions below), cell logger & alarm only (no BMS)
Install solar panels

Questions on batteries
1) was planning to get Winston due to perceived better hi-temperature resistance, but likely supplier (ev power europe) has CALB & Sinopoly as well, cost moreor less same, any argumented suggestion ?
2) Seems most people here would recommend single bank, I have seen / heard other folks recommend twin banks to alternate and thus cycle rather than float. Any argumented comment
3) If single bank, is it really a plus to have 2P8S (for example each 3.2 V cell being parallel 100 Ah rather than single 200 Ah) ? Note availability is not similar

Questions on hookup
1) If single bank, how to best connect both alternators : anything needed or will Incharge regulators manage ?
2) Same questions if twin banks, additionnally how to switch from bank 1 to bank 2 ? (Probably have to go through disconnect, hence restart all electronics...)

Questions on operating parameters
1) recommended voltage settings for Multiplus, Incharge alt regulators, solar panel
2) I am not (yet) a liveaboardb, and boat may remain unvisited for up to 6 months. Is it best to disconnect all loads and charging means (as suggested by Mainesail) or is it possible to leave some charging means connected (solar and/or Multiplus) allowing batteries to float and allowing to leave on some potential loads (bilge pumps, etc...)

Last question on battery suppliers : anyone having feedback on Winstońs Italian affiliate / representative ? Are they selling to individuaĺ customers ?

Thanks in advance for sharing your know-how and experiences

[ebaugh]

Originally Posted by caipirinha

Hello,

I've done it ! (Read the whole thread, that is). I still have a few questions.

Current status :
Both house banks (Exide gel, 24 V 230Ah) are dead
Charging systems are Tecsup charger (to be replaced, see below) and Leece-Neville alternators, shaft driven, switch-activated field, external HeartInterface regulators on both engines (boat is a catamaran) and 24V Fisher-Panda generator (will probably be removed) connected to banks via Surepower 1202
Monitoring with one Xantrex Link10 for each bank

Planned change :
Replace Tecsup charger by Victron Multiplus 24-3000-70-16 (done but not yet connected)
Replace house batteries by LFP (questions below), cell logger & alarm only (no BMS)
Install solar panels

Questions on batteries
1) was planning to get Winston due to perceived better hi-temperature resistance, but likely supplier (ev power europe) has CALB & Sinopoly as well, cost moreor less same, any argumented suggestion ?
2) Seems most people here would recommend single bank, I have seen / heard other folks recommend twin banks to alternate and thus cycle rather than float. Any argumented comment
3) If single bank, is it really a plus to have 2P8S (for example each 3.2 V cell being parallel 100 Ah rather than single 200 Ah) ? Note availability is not similar

Questions on hookup
1) If single bank, how to best connect both alternators : anything needed or will Incharge regulators manage ?
2) Same questions if twin banks, additionnally how to switch from bank 1 to bank 2 ? (Probably have to go through disconnect, hence restart all electronics...)

Questions on operating parameters
1) recommended voltage settings for Multiplus, Incharge alt regulators, solar panel
2) I am not (yet) a liveaboardb, and boat may remain unvisited for up to 6 months. Is it best to disconnect all loads and charging means (as suggested by Mainesail) or is it possible to leave some charging means connected (solar and/or Multiplus) allowing batteries to float and allowing to leave on some potential loads (bilge pumps, etc...)

Last question on battery suppliers : anyone having feedback on Winstońs Italian affiliate / representative ? Are they selling to individuaĺ customers ?

Thanks in advance for sharing your know-how and experiences

1) Of the three I like CALB
2) Single bank provides easier to use capacity than managing multiple banks. Alternators can both charge a single bank. One probably will end up working harder.
3) Any cell size connected first in parallel then in series is fine. Biggest consideration here is fit in boat. And availability as mentioned....

1) Bulk 13.8, Acceptance 13.6, Float 13.2 using custom settings and measured at the battery after install
2) If in the water you will have too? Pick your most reliable charging system and turn the others off. As is a good practice for any boat left unattended, hire someone to visit every week or two. Give them specific instructions for the batteries. Be sure all non essential loads are off, especially the inverter.

[hellosailor]

My friend swears that balqon [owned by Winston] told him that even if he uses one of their "twelve volt" batteries that are packaged and sold with just the two 12-v terminal posts, like a regular 12v battery, he =must= use a bms with it and not just connect an alternator or regulator.

How you would use a bms when the individual battery cells are sealed up inside the case is a good puzzle but he said they told him he =must= use one. Maybe that's a way to dodge the warranty, maybe it is a mistake. Maybe in five more years all the lithium battery people will tell the same story and stop quarreling.

[diugo]

Originally Posted by hellosailor

How you would use a bms when the individual battery cells are sealed up inside the case is a good puzzle

Well technically the B in BMS stands for Battery, not cell. Although it's far wiser to manage LiFePO4 batteries at the cell level, you still could have basic battery-level management that opens the circuit below 12.1V or above 14.5V. Especially if the cells are well matched both for capacity and internal resistance, as is presumably the case with 12V batteries.

[diugo]

Originally Posted by pz5r9n

A bit off topic but this price seems to put these batteries within range?
Rechargeable Batteries & Battery Chargers - 3.2V 100Ah LiFePO4 (Lithium Iron Phosphate) Rechargeable Battery

$1.40/Ah is not really a bargain. I think at that price I would choose CALB over Tenergy: Battery - CALB / Sky Energy 3.2V, 100AH LiFePO4 cell

[hellosailor]

"Well technically the B in BMS stands for Battery, not cell"
Yes, but do you know of any that do it that way? That manage the battery as one item, rather than managing the battery cell-by-cell ?
If you just want to keep the battery in bounds, all you need is an alternator with regulator to prevent overcharging, and keep an eye on your "charge" light to not run it down. Or use a battery buddy that cuts out when the voltage drops too low. Hardly what gets sold as a BMS, is it?

[diugo]

If Balqon sold me an ostensibly sealed 12V LiFePO4 battery then insisted I get a BMS for it, then a battery buddy would fit the bill quite nicely.

OTOH, if I bought individual cells, then it would be incumbent on me to protect each one individually.

Simple as that.

I might point out that the "12V LiFePO4 battery" is a relatively recent invention, whereas the classic BMS has existed for a decade or more. If the 12V concept successfully catches on, then the need for expensive, complex cell-based management may well need re-evaluation.

[ebaugh]

One year benchmark.....

I just pulled 4 of my 48 100 Ah GBS LFP cells and ran a capacity test. First I charged to 3.58V, and top balanced the 4 cells at .3C. Then at 30A/.3C discharge rate to 2.85V, the tool I used calculated a capacity of 77.4 Ah, for the weakest cell. I was hoping to see a bit more than this and wondered what I lost by not going to 2.5V, so I immediately restarted the test to 2.5V and saw another 3 Ah, for a total of 80.4 Ah.

Attached to the post is the graph from my test setup. The RC logging regenerative charger/discharger (Powerlab 6) did a great job at regulating the discharge rate. However, my clamp meter showed just under 2A more than the charger saw and used in its calculations. So that would get you to 85 Ah. I estimate the "stronger" cells might show another 3 Ah, so the most optimistic case appears about 88 Ah.

Unfortunately I have no baseline to compare with, but I feel pretty sure they would have tested to capacity on delivery based on EV users doing out of the box testing. In Grenada, I just did not have the equipment to perform this test.

These cells have seen 320-340 partial cycles, the average cycle 85-50-85% SOC twice daily over 160-170 days off grid at anchor, the remaining time floated at 3.32V per cell. The EMS/BMS has never tripped due to voltage extremes. I expected to see more like 95-96 Ah Capacity. Assuming some linearity, that would get me to 2000 cycles at 80% capacity remaining.

As It stands at the moment, subject perhaps to some more testing, the implied cycle life is more like 800-1000. Still better than AGM or FLA golf carts, especially in my engine room. But not nirvana.....this is about 2X the life of golf carts for us.

Questions:

1) I may have given up something by not charging to 3.65V for the test. If so how much?

2) Is there any reason to run a couple full cycles on the test bench? Is there any reason to suspect the figures to improve since these are really the first full cycles (100-0-100% SOC), these cells have seen.

3) Any other thoughts and/or suggestions?

[hellosailor]

"OTOH, if I bought individual cells, "
From what I gather, you actually can't buy individual cells from Balqon. Apparently what they sell as a "cell" really is a parallel battery of two cells in one case. Every time.

Which begs to ask then, why isn't there a BMS inside that battery either? Wait, I'll have to ask the emperor's tailor. <VBG>

You know, a hundred years ago, having tobacco smoke blown up the (ahem) was actually a medically recommended procedure. These days...I don't care what their song is, I just wish they'd all sing the same one.

[goboatingnow]

Originally Posted by hellosailor

"OTOH, if I bought individual cells, "
From what I gather, you actually can't buy individual cells from Balqon. Apparently what they sell as a "cell" really is a parallel battery of two cells in one case. Every time.

Which begs to ask then, why isn't there a BMS inside that battery either? Wait, I'll have to ask the emperor's tailor. <VBG>

You know, a hundred years ago, having tobacco smoke blown up the (ahem) was actually a medically recommended procedure. These days...I don't care what their song is, I just wish they'd all sing the same one.

cell level BMS are only useful on series connected cells. not much use on a parallel only cells , if you think about it.

All Li cells/batteries are composed of a set of parallel cells

Dave

[ebaugh]

Originally Posted by hellosailor

"OTOH, if I bought individual cells, "
From what I gather, you actually can't buy individual cells from Balqon. Apparently what they sell as a "cell" really is a parallel battery of two cells in one case. Every time.

Which begs to ask then, why isn't there a BMS inside that battery either? Wait, I'll have to ask the emperor's tailor. <VBG>

You know, a hundred years ago, having tobacco smoke blown up the (ahem) was actually a medically recommended procedure. These days...I don't care what their song is, I just wish they'd all sing the same one.

Actually a lot more than 2....see attached picture.

[diugo]

Originally Posted by ebaugh

Actually a lot more than 2....see attached picture.

Okay wait. First, Mainesail puts his cells under some fancy lexan cover. Now you've taken it a step further and put one of your cells in a lexan enclosure. Next it'll be the whole boat made out of lexan!

[rfadler]

Originally Posted by ebaugh

One year benchmark.....

I just pulled 4 of my 48 100 Ah GBS LFP cells and ran a capacity test. First I charged to 3.58V, and top balanced the 4 cells at .3C. Then at 30A/.3C discharge rate to 2.85V, the tool I used calculated a capacity of 77.4 Ah, for the weakest cell. I was hoping to see a bit more than this and wondered what I lost by not going to 2.5V, so I immediately restarted the test to 2.5V and saw another 3 Ah, for a total of 80.4 Ah.

Attached to the post is the graph from my test setup. The RC logging regenerative charger/discharger (Powerlab 6) did a great job at regulating the discharge rate. However, my clamp meter showed just under 2A more than the charger saw and used in its calculations. So that would get you to 85 Ah. I estimate the "stronger" cells might show another 3 Ah, so the most optimistic case appears about 88 Ah.

Unfortunately I have no baseline to compare with, but I feel pretty sure they would have tested to capacity on delivery based on EV users doing out of the box testing. In Grenada, I just did not have the equipment to perform this test.

These cells have seen 320-340 partial cycles, the average cycle 85-50-85% SOC twice daily over 160-170 days off grid at anchor, the remaining time floated at 3.32V per cell. The EMS/BMS has never tripped due to voltage extremes. I expected to see more like 95-96 Ah Capacity. Assuming some linearity, that would get me to 2000 cycles at 80% capacity remaining.

As It stands at the moment, subject perhaps to some more testing, the implied cycle life is more like 800-1000. Still better than AGM or FLA golf carts, especially in my engine room. But not nirvana.....this is about 2X the life of golf carts for us.

Questions:

1) I may have given up something by not charging to 3.65V for the test. If so how much?

2) Is there any reason to run a couple full cycles on the test bench? Is there any reason to suspect the figures to improve since these are really the first full cycles (100-0-100% SOC), these cells have seen.

3) Any other thoughts and/or suggestions?

First I have to confess I have a huge interest in the outcome of your testing. I've spent a considerable amount of time thinking about going the LifePo4 route when I upgrade my house bank. That said I have spent a career performing software and electronics (hardware) testing and would recommend that you;
1) run the test again but go 100-0-100% SOC,
2) select four different cells and run the test again,
3) select four additional cells and run the test again

Using just four cells and one test cycle does not constitute a valid test. There are so many variables that you need multiple tests to confirm the result. That said, your time is not free and any and all information you've provided is very much appreciated.

I really appreciate the information you've shared on this forum and on your web site.