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

[sytaniwha]

Hi all,

It's great that this thread has kept on discussing the issues of using LiFePO4 for house banks.

We've just returned from sailing for the last 6 months, and we're about 18 months into using our 400Ahr CALB cells (4 cells in series) on our 46ft cruising monohull. I've posted quite a bit of info previously about our set-up, so if any of you guys relatively new to the thread want to know about it, search through my previous posts.

Here's an update on what we've found regarding our battery:

• we hit full charge termination at 3.4v/cell, and we get more than the rated 400Ahr from the cells,
• We're using all SLA charge equipment, with the charge profile set to our 13.6v cutoff, and it works perfectly,
• We have an over-voltage alarm, but no cutoff apart from the charge devices, and an under-voltage cutoff (just like we did with our SLA battery),
• We manually balanced the cells right at installation using full charge voltage techniques (a grunty power resistor). It took a couple of charges to get them perfectly balanced. We haven't had to re-balance the cells since (check them every couple of months now),
• We've been living aboard in cruise mode for the last 6 months, so I'd call that about as much use as we'll ever have,
• When in the marina we leave the shore charger connected (50A, so C/8), as we did with the SLA's, and it's perfect (solar cells would be good too). The nice thing about good shore chargers, MPPT solar controllers, and marine alternator regulators is that the most common failures appear to result in zero charge current, which is the safest failure mode,
• We've found that voltage definitely isn't a good predictor of capacity during discharge because the curve is so flat. Even with 3 fridge/freezer compressors running, and the windlass turning, we haven't seen voltages below 12V unless we have deliberately discharged the cells to find the low end,
• But the voltage cut-off on charge is very reliable,
• This 400 Ahr battery is giving us way more useful capacity than our old SLA Trojan battery (of 720Ahr I think), which suggests that the calculations regarding usable capacity for LiFePO4 which suggest that they can replace a SLA of about double the capacity is roughly correct,

Don't ask how many cycles we've had out of the battery - it's an impossible figure to compare outside of a lab because we have no consistency in anything (DOD, disharge rate, charge time, etc etc).

I agree with everything Dave (Go Boating Now) said in a post a few days ago regarding there being little need for a "BMS" however you interpret that. I'd add that the EV applications that LiFePO4 is mostly used in (and which drove BMS development) also used many series/parallel connections (In some cases literally 100 parallel strings). That would be a nightmare for keeping cell balance regardless of what chemistry you use, so would necessitate some sort of cell level monitoring and control - hence the invention of BMS. In our case (no parallel cells), the cells have remained perfectly balanced thus far.

Just some practical comments for anyone interested.

Cheers,
Paul.

[Bumperdick]

Hello to all,

Being very interested in new battery technology I read this from start to finish. A very helpful thread indeed! After reading the thread I have a few observations/questions on the use of LiFePO4 as the main bank on a boat.

a) I don't like the use of a LVC. I am not sure I understand how the LiFePO4 owners on this thread use it, though. Is it an alarm, or does it actually cut off the power? The latter would obviously not be acceptable on a boat, at least not while cruising with the nav equipment operating.

b) Cost is an issue. I feel this issue is not correctly addressed on the thread. Frankly I feel there are 'fanboys' trying to paint a slightly too optimistic picture, this is not real criticism because I realize it is quite normal to approve of your own decision to purchase but for those still pondering if this technology is right for them a more level comparison is in order. First of all, I have been using forklift lead-acid batteries for ever on my boat, and these can be discharged to 20% SOC or even 0% if you charge them soon after, so I believe the comparison between lead-acid and LiFePO4 should be between these type of LA bats and not between golf cart or other even more substandard design. (Somebody even insinuated that they left their LA battery alone (unconnected) for four months and on return it had died. I just left my car battery disconnected for 5 (five) months and my car started 1st try, and you can't get much more substandard than car batteries for house banks)
To get back to cost. On this thread I read that the current best price for LiFePO4 is $1.25/Ah@3.2V or something like $3600 for a complete 600Ah battery @12v (advertised prices are hard to come by!) I can buy 6x2V LA traction cells C/10 1000Ah for some $1900. This battery would likely slightly outperform the LiFePO4. If I were to spend $3600 I could get close to double the capacity. Such a LA system would definitely outperform the LI in terms of capacity, but what IMO is more important in a boat, it would be much more fool-proof/easy to handle in terms of overcharging or discharging. I will be the first to admit that it needs maintenance that some people may dislike that lot. Its weight and size will be an issue for a smaller boat or a racing boat. On my boat (steel 75ft) weight and size does not matter much and the battery compartment is a large ventilated (to outside) zinc-lined steel box so safety is also guaranteed. (Still, LiFePO4 would be A LOT safer if a boat were to roll over. Batteries or not, I would not survive such an event anyway...)

c) Charge current. For me this is the (only real) strong attraction point to LiFePO4. I would truly love to charge my 24v/1500Ah with C/1 @ 1500 amps and be done 1/2 to max 1 hour !!! The disappointing thing is that nobody on this thread seems to take advantage of this strong point? Frankly I don't even know if the technology is there. That you need to charge an EV for many hours is probably due to mains current limitation and charging technology may be cost-effective/available only for this 'mass market'. Using a 100A alternator on a 400Ah bank feels like losing out on the very best feature of this battery type. It would really be worth installing a heavy generator to enable C/1 (or better!) charging. Since nobody seems to be doing it, I am not sure how it would work out on the lifespan of the batteries...?

d) It is definitely not true that the same mishandling that damages LiFePO4 will damage a LA traction battery. You can discharge a traction to zero and recharge it a day or two later to full capacity. You can put 16 volts on a traction battery and it will actually be perfect (equalization) even if you leave it on for 20 hours - given it has enough water or a watering system. That robustness is an interesting feature, not just for boats but also for a stationary application including solar panels. I believe LA traction might have the upper hand over LiFePO4 there. After all with all this talk about faling BMSes, it seems HVC and LVC systems could fail just as well. And then, if there is nobody watching the dials....what?

e) SOC. Given the very flat LiFePO4 curve I assume it is almost a necessity to have a battery monitor that you can really rely on. Including a correctly programmed Peukert coefficient. This is assuming that you don't want to just rely on your LVC to protect your batteries?

I can imagine that space/weight-constrained people (RV, smaller sailboats, racing/speedboats) may want to pay today's large premium for LiFePO4. But I am not yet convinced: the damage done by discharging completely worries me. The use (if I get it right) of LVC systems worries me. The difficult equalization also worries me. That is hasn't been an issue on this thread may be because of the very low charge currents everybody seems to use. It seems that you would definitely need an active BMS when charging at high currents. But then you can have a BMS failure, and then? Not in the middle of the ocean, please!

Is there anybody out there on a large powerboat that is doing long distance, fast-charging a large LiFePO4 bank? If so what equipment are you using and what are you're experiences???

Greetings,

Bangaburt

[ebaugh]

Originally Posted by Bumperdick

Hello to all,

Being very interested in new battery technology I read this from start to finish. A very helpful thread indeed! After reading the thread I have a few observations/questions on the use of LiFePO4 as the main bank on a boat.

a) I don't like the use of a LVC. I am not sure I understand how the LiFePO4 owners on this thread use it, though. Is it an alarm, or does it actually cut off the power? The latter would obviously not be acceptable on a boat, at least not while cruising with the nav equipment operating.

b) Cost is an issue. I feel this issue is not correctly addressed on the thread. Frankly I feel there are 'fanboys' trying to paint a slightly too optimistic picture, this is not real criticism because I realize it is quite normal to approve of your own decision to purchase but for those still pondering if this technology is right for them a more level comparison is in order. First of all, I have been using forklift lead-acid batteries for ever on my boat, and these can be discharged to 20% SOC or even 0% if you charge them soon after, so I believe the comparison between lead-acid and LiFePO4 should be between these type of LA bats and not between golf cart or other even more substandard design. (Somebody even insinuated that they left their LA battery alone (unconnected) for four months and on return it had died. I just left my car battery disconnected for 5 (five) months and my car started 1st try, and you can't get much more substandard than car batteries for house banks)
To get back to cost. On this thread I read that the current best price for LiFePO4 is $1.25/Ah@3.2V or something like $3600 for a complete 600Ah battery @12v (advertised prices are hard to come by!) I can buy 6x2V LA traction cells C/10 1000Ah for some $1900. This battery would likely slightly outperform the LiFePO4. If I were to spend $3600 I could get close to double the capacity. Such a LA system would definitely outperform the LI in terms of capacity, but what IMO is more important in a boat, it would be much more fool-proof/easy to handle in terms of overcharging or discharging. I will be the first to admit that it needs maintenance that some people may dislike that lot. Its weight and size will be an issue for a smaller boat or a racing boat. On my boat (steel 75ft) weight and size does not matter much and the battery compartment is a large ventilated (to outside) zinc-lined steel box so safety is also guaranteed. (Still, LiFePO4 would be A LOT safer if a boat were to roll over. Batteries or not, I would not survive such an event anyway...)

c) Charge current. For me this is the (only real) strong attraction point to LiFePO4. I would truly love to charge my 24v/1500Ah with C/1 @ 1500 amps and be done 1/2 to max 1 hour !!! The disappointing thing is that nobody on this thread seems to take advantage of this strong point? Frankly I don't even know if the technology is there. That you need to charge an EV for many hours is probably due to mains current limitation and charging technology may be cost-effective/available only for this 'mass market'. Using a 100A alternator on a 400Ah bank feels like losing out on the very best feature of this battery type. It would really be worth installing a heavy generator to enable C/1 (or better!) charging. Since nobody seems to be doing it, I am not sure how it would work out on the lifespan of the batteries...?

d) It is definitely not true that the same mishandling that damages LiFePO4 will damage a LA traction battery. You can discharge a traction to zero and recharge it a day or two later to full capacity. You can put 16 volts on a traction battery and it will actually be perfect (equalization) even if you leave it on for 20 hours - given it has enough water or a watering system. That robustness is an interesting feature, not just for boats but also for a stationary application including solar panels. I believe LA traction might have the upper hand over LiFePO4 there. After all with all this talk about faling BMSes, it seems HVC and LVC systems could fail just as well. And then, if there is nobody watching the dials....what?

e) SOC. Given the very flat LiFePO4 curve I assume it is almost a necessity to have a battery monitor that you can really rely on. Including a correctly programmed Peukert coefficient. This is assuming that you don't want to just rely on your LVC to protect your batteries?

I can imagine that space/weight-constrained people (RV, smaller sailboats, racing/speedboats) may want to pay today's large premium for LiFePO4. But I am not yet convinced: the damage done by discharging completely worries me. The use (if I get it right) of LVC systems worries me. The difficult equalization also worries me. That is hasn't been an issue on this thread may be because of the very low charge currents everybody seems to use. It seems that you would definitely need an active BMS when charging at high currents. But then you can have a BMS failure, and then? Not in the middle of the ocean, please!

Is there anybody out there on a large powerboat that is doing long distance, fast-charging a large LiFePO4 bank? If so what equipment are you using and what are you're experiences???

Greetings,

Bangaburt

Interesting analysis. The main thing I'm not sure about is:

1) The cost of large 2V traction cells. The ones I've seen are more than you quoted.

2) I have seen data that indicates these cells have a life of up 2000 cycles, but nothing that would support discharging them to 20-30 % SOC every cycle and still see that sort of life.

Can you provide any further info on this?

The packaged LiFePO4 systems try to solve the HVC/LVC cutoff issue with multiple banks, although I assume a LVC scenario would soon hit both banks. I solve it by leaving my AGM start batteries paralleled when underway. Ideally I would set LVC at around 12.6V, the LiFePO4 would be essentially empty and the AGMs almost fully charged. But I can't set my current BMS to that value.... I do have an audible alarm I can set on my engine monitoring system to that voltage. I'm the only one I know doing this but it works so far for me.

The EV guys are doing some C/1 charging, but I don't think it's very widespread yet. It's much easier to charge 120V at 100A then 12V at 1000A. Maybe others have better luck, but even with new connectors I've had heat problems with the standard "stud" type connections at 200A. The cables are fine, but the connections heat up if the current is continuous.

What you will see different is the Li will take your maximum rate until fully charged. So no waiting around to get to 100% SOC if you want a full charge. But most significant no sulphation, no need to go to 100% unless you desire. This saves me about 20% on generator run time at anchor.

[Bumperdick]

Hello Ebaugh, others;

For LA traction cycling you can check the brochure for the EnerSys 'Solar-One with HuP technology', obviously a product positioned as a 'premium product' - mostly marketing no doubt - that guarantees 2100 x 80% cycling. But let's face it, how many cycles is my battery going to get? I am not a forklift user. Here is my take.
I prefer a quiet boat and have sized my house bank so that I have to run a generator a maximum of once every three days for four hours to charge. I believe that if you need to run an auxiliary much more often than that you might actually prefer a Whispergen (I am limiting myself to boats now).
In the latter case you obviously don't need all that much battery capacity and probably not LiFePO4 either. In the first case you cycle (worst case) from 80% to 20% every three days so max 10x per month. Subtract shorepower and I get <100 cycles per year. Do I care if the battery gives 1000 or 5000 cycles? Not really. It will be interesting reading from the LiFePO4 owners who complete their 20 years of cycling - if they exist. But no doubt today's LiFePO4 adopter will long since have moved on to nanotube supercaps ;-)

I checked and you can buy a 24v 1000AH 12x2Vcells LA forklift traction bat for €2587=$3350 today, look for the 10PzS900H (900Ah C/5 1160Ah C/20) in combination with that price. It is good for >1500 cycles@80%. The price is actually less than I quoted earlier at $1700 for 12V, and that from the first random battery seller I found that had a price on them. I suppose you could get them a LOT cheaper direct from China to make for a real LiFePO4 comparison. For $3600 you could then get a bit more than 2000Ah@12v, and of those 2000 you could already safely cycle 1200Ah (20->80%), which is twice the amount of the entire LiFePO4 capacity you could get at that price and so definitely more than twice the cycling capacity of that battery.

Today I charge at 300A@24v, so Ebaugh I suppose your connectors are not the very best...corrosion? Aluminium? But here I cannot go much higher in charging current for my LA batteries. Now, for LiFePO4 on the other hand, how would I get to 1000A without adding another seven 100A chargers to the collection? Given enough smaller cells paralleled I imagine getting the power in should be doable... But even if I actually installed another 7 chargers (OK, think space might be an issue after all...) then seriously, do I want to be the guinea pig testing it out on my new $12000 1kAh 24v LiFePO4 bank?

About sulphation. Of course I only bulk charge most of the time. Once every ten cycles or so after the main charging is done I switch on my quiet, small variable speed DC 'backup' generator and let it run for the day to get through equalization. I think that an adequately sized solar panel could do much the same, perhaps even more effectively, instead of letting some HVC shunt away any excess current... So it is not that big of a hassle - imho.

What *I* want is more battery capacity and less generator time. My boat is not suited for solar panels so if LiFePO4 can give me that at a reasonable price then I am a taker. So far the costs and risks appear to be high.

Later,

Bangaburt

[ebaugh]

Bangaburt,

I don't have any experience with tubular traction batteries. But they do have more impressive specs than the floor sweeper or golf cart types I'm used to working with. But I still was unable to find anything that indicated a standard of 80% DOD for 2000 cycles. What I did find was this:

http://www.zvei.org/Verband/Fachverb...%202001-09.pdf

If I read the charts right, it looks like under optimal conditions you can get 1300-1500 cycles at 80% with flooded or 60% with VRLA. Then it looks like you might do better at lower rates of discharge, but I don't know what the I5 rate referred to is. That may be where you quote the extra cycles. But the cost of non standard temps looks pretty high.

I'd have to call around to find pricing in the US, it is elusive on the Internet. The lowest for LiFePO4 is $3450 for 1000Ah at 12.8V through Balqon.

In any case, if you only charge every 3 days, I agree there isn't a cost benefit case for Li other than size, weight and perhaps temperature.

In my case, I charged twice per day. First to try and stay within the 50-80% SOC range, and because we have an electric galley, to cook. So I burned through 6 at least partial cycles for every one of yours. And my batteries are in the engine room in the tropics. So they are always above standard temps. Not too bad with just the genset, but much higher with the mains running.

Different situation, different solutions. But there is no doubt the traction solution is more widely proven since we have many more years of experience. Someone has to pioneer the technology to obtain real data for others. While I don't expect this...there is some data that suggests at our relatively low (for Li) rates, for DOD's kept in the 30 to 90% range, a LiFePO4 bank may last the life of the boat. If the chemistry can remain stable for 20 years, a complete unknown. But theoretically it doesn't "wear out" like flooded. Even for shorter periods than that, there is something to be said for being confident you don't need to replace batteries in Timbuktu, where costs, shipping and duties can make it more expensive.

Bob

[witzgall]

Originally Posted by ebaugh

In any case, if you only charge every 3 days, I agree there isn't a cost benefit case for Li other than size, weight and perhaps temperature.

I would add to that list:
Longer bulk charging - up to 90%
No Perkert effect - so running a high amp device will not effect you negatively ( Not such an issue for those with huge banks).
Safety - IMHO, with all things considered, LIFEPO4 are safer than wet cells.
Location - no venting needed, can lay on side (Same as AGM,Gel)

The big one, that is often skipped by most when discussing these cells is how they react to being left partially charged. LIFEPO4, as far as I can tell, are very happy to be left at 50% charge. So, when your AGM or wet bank is at 50%, you are compelled to charge it back up to preserve the cycle life. Not so with LIFEPO4.

If you are constantly running a daily negative AH draw, then this might not be a concern of yours. But if you are like me, trying to maintain a system where you do not need to run a generator or engine to charge you bank, then it is a BIG deal. In the bahamas last winter, our batteries were at times bouncing between 50-70% for days at a time. We did not feel compelled to start our engine to charge, waiting for either the wind to pick up (wind gen) or moving the boat (alternator charging) to bring the bank back up to capacity.


Chris

[Lagoon4us]

Originally Posted by witzgall

I would add to that list:
Longer bulk charging - up to 90%
No Perkert effect - so running a high amp device will not effect you negatively ( Not such an issue for those with huge banks).
Safety - IMHO, with all things considered, LIFEPO4 are safer than wet cells.
Location - no venting needed, can lay on side (Same as AGM,Gel)

The big one, that is often skipped by most when discussing these cells is how they react to being left partially charged. LIFEPO4, as far as I can tell, are very happy to be left at 50% charge. So, when your AGM or wet bank is at 50%, you are compelled to charge it back up to preserve the cycle life. Not so with LIFEPO4.

If you are constantly running a daily negative AH draw, then this might not be a concern of yours. But if you are like me, trying to maintain a system where you do not need to run a generator or engine to charge you bank, then it is a BIG deal. In the bahamas last winter, our batteries were at times bouncing between 50-70% for days at a time. We did not feel compelled to start our engine to charge, waiting for either the wind to pick up (wind gen) or moving the boat (alternator charging) to bring the bank back up to capacity.


Chris

Chris do you get sufficient windgen input to now say it's worthwhile fitting? I get very mixed responses on this point... Cheers Frank

[witzgall]

Yes, we did, with an old Air 403 and Spreco blades. The output was good, but not great. To the end we are installing a replacement, the Silentwind generator from the Spreo guys. This should increase the output something like 50-100% over the old unit. I think that wind+ solar is a great solution, as long as there is wind where you are sailing. The batteries are the final + in this scenario, as so much of the generated power ends up in the batteries (over 97%).

Chris

Originally Posted by Lagoon4us

Chris do you get sufficient windgen input to now say it's worthwhile fitting? I get very mixed responses on this point... Cheers Frank

[Bumperdick]

Well Bob you seem to be a huge power user... How many kWhs do you use per day??? I suppose it would be hard for you to size your banks so that you would need the amount generator time that I would like. If the claims about LiFePO4 cycling are right and the battery chemistry is truly temperature insensitive then I think you did the right thing switching.
Did you consider just running a small generator 24/7 instead of constantly cycling? Your energy budget seems to call our for that.

Originally Posted by ebaugh

<...> If the chemistry can remain stable for 20 years, a complete unknown. But theoretically it doesn't "wear out" like flooded. Even for shorter periods than that, there is something to be said for being confident you don't need to replace batteries in Timbuktu, where costs, shipping and duties can make it more expensive.

If the chemical stability over 20 years is a complete unknown then personally I would be conservative and ASSUME that I would have to replace the batteries in Timbuktu ;-)

Greetings,

Bangaburt

[Bumperdick]

Originally Posted by witzgall

I would add to that list:
Longer bulk charging - up to 90%

Ok this is excellent news for me! You have my full attention. I would settle for 75%, too. Now, given the discussion, how do we fast charge my new 24v 1500 Ah LiFePO4?

Regards,

Bangabert

[ebaugh]

Originally Posted by Bumperdick

Ok this is excellent news for me! You have my full attention. I would settle for 75%, too. Now, given the discussion, how do we fast charge my new 24v 1500 Ah LiFePO4?

Regards,

Bangabert

Well...you have to give up your 24V notions and switch to LiFePO4 at a much higher voltage. Build 360 VDC LiFePO4 bank of 100 Ah capacity. Buy something like this charger:

http://blog.evtv.me/store/proddetail.php?prod=PulsaR1

It charges at 100A, up to 370V, from 240AC. Use the built in DC to DC converter to float a small 12V bank for your inverter. It's good to 50A, so if you average more than 50A it won't work. But if you need more, that's almost where you need a small genset and run 24x7.

How bad do you want to charge in an hour?

[ebaugh]

Originally Posted by Bumperdick

Well Bob you seem to be a huge power user... How many kWhs do you use per day??? I suppose it would be hard for you to size your banks so that you would need the amount generator time that I would like. If the claims about LiFePO4 cycling are right and the battery chemistry is truly temperature insensitive then I think you did the right thing switching.
Did you consider just running a small generator 24/7 instead of constantly cycling? Your energy budget seems to call our for that.

If the chemical stability over 20 years is a complete unknown then personally I would be conservative and ASSUME that I would have to replace the batteries in Timbuktu ;-)

Greetings,

Bangaburt

At the installation ...Timbuktu was only for another 18 months with golf cart batteries borderline to make it....I figured Li would work. I average 20-25A at 12VDC, it's a lot, but not enough to run a genset 24x7, unless we need air conditioning too. So far, we've done OK without it. But there is an argument to run it overnight for sleeping, both eliminating a battery cycle and with flooded batteries, getting back to 100% every day.

[Lagoon4us]

Originally Posted by ebaugh

I don't know exactly how the existing inverter is wired. Can you run (or at least attempt to) run every AC load on it? If so, it probably looks something like the Victron in this picture. If not, there may be a sub panel for those loads. The MSP2012 has a transfer switch, but the manual does not identify its rating that I could find. I would be skeptical running the whole house through the MSP2012.

But the left side of this diagram shows one way to incorporate the Cruise gen, and the right side is how you would wire in the Victron Multi if you go that route. Otherwise, all that side remains the same.

No we only run white goods and de-sal on the inverter, if we wish to use hotplates we run the genset this suits us. By this Lithium installation we are trying to bank the solar energy, main engine motoring and likely wind gen power..

We are not trying to have a floating power station just take advantage of amps we make by what we do....

[Lagoon4us]

Originally Posted by ebaugh

Almost has to be an oil pressure switch on the port engine that closes a solenoid. Or possibly something that senses charge voltage sensed near the port alternator.

I guess they do it that way to maximize the voltage at the windlass motor. I've heard of this being done on some boats. If you had lots of chain out and a single weak battery you might not get the engine started. It also smells of a bareboat charter design...

We are happy to start if we have lots out but would rather make that choice than rely on that port engine having to always be operational to retrieve the pick, a possibly dangerous scenario methinks...

[CharlieJ]

A simple solution for the Lagoon 440 windlass wiring dilemma would be to install a 1-2-Both-Off selector switch in the B+ to the windlass. Port engine start battery B+ to Position 1 on the selector switch, stbd engine start battery B+ to Position 2 on the selector switch and the windlass B+ to the Common terminal on the selector switch.

The operational rule would be very judicious about placing the selector switch in the Both position as that would parallel your two starting batteries.