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

[OceanSeaSpray]

Originally Posted by SV THIRD DAY

How long have you been running your LiFePO4 bank?
And share with us your set-up for comparison....that would be great to see.

Long enough to demonstrate that every scenario that trashed other simple-hopeful lithium installations around here can't happen to mine and my setup consists of a small aluminium box 120x70x20mm with one LED on it living near the cells, two DC disconnectors, one small relay and a little optional piezo buzzer.

One disconnector separates the load bus and the other one does the charge bus. The small relay is there to safely disable the alternator early, and other things if you feel like it.
Standby power consumption is virtually zero.


I am sorry to hear about your deep discharge experiment, because below 2V/cell the copper starts going into solution into the electrolyte. When you recharge, of course it doesn't go back where it had come from, but drops out wherever it is, creating a short-circuit risk and increasing self-discharge rates. Too late now.

[SV THIRD DAY]

Originally Posted by OceanSeaSpray

I am sorry to hear about your deep discharge experiment, because below 2V/cell the copper starts going into solution into the electrolyte. When you recharge, of course it doesn't go back where it had come from, but drops out wherever it is, creating a short-circuit risk and increasing self-discharge rates. Too late now.

Oh sure...that's how the chemistry works...but what does that mean in the real world? I can put my chemistry degree to work and sound educated all day long. But what does that mean to a cruiser on a boat that has his fancy dancer BMS controller fail on him and drained his LiFePO4 battery bank down? Or what about your system (which sounds exactly like the system I have on my 400AH test bank No 2 by the way...easy and clean) what happens if it fails? Because for every piece of electronic equipment being used on a boat and in the world today, there is a product failure statistic rate saying there will be a measurable failure rate. So what then? A new bank? Your boat blows up? Your mom cuts you out of the family fortune for being the black sheep son that went out cruising in the first place? What...no one wants to really know?

That's the problem no one wants to do the tests and gather the data that matters to know if the scary sounding chemistry matters in how we are using these banks on a boat. Is it one strike and you're out or can you have a screw-up and still not as some suggest have to put your bank at the dumpster for a cruiser hobo to come along and pick up.

Does it matter how long ago that Dead Drain Test was conducted along with how long I've not see any measurable negatives from it? Sure it does, it might also wash away a little of your smugness to know it was longer than my last bank of T105 batteries lasted...but hey...I'm just a bozo, so what do I know....

This Gigantic Thread isn't a pissing contest OceanSeaSpray, you win...I lose, but everyone knows anyway that "When at Sea, Gentlemen sit to Pee". This thread/topic is about sharing what people have done with their LiFePO4 banks. What worked, what didn't etc. Some will only be comfortably with a fully automatic BMS that talks to the internet via a WiFi connection so they can read the status of their batteries on their smart phone while on vacation in Vegas. Me...If I want to know what's going on with my bank, **** I look up from the desk where I live and work full time and look at the Individual Cell Volt Meters, the Bank Volt Meter and the AH Counter just like I have done for the last 7yrs of living aboard. Did that with my over-priced and over-rated T105s and doing it now with the fabulous LiFePO4 bank. The only difference is that now I see numbers that make me happy rather than pissed off.

[deckofficer]

Rich,

I haven't posted to this thread for awhile but I'm still a happy camper with my cells. One bank has over 400 cycles and still has capacity that is about 15% over advertised. Another bank hasn't been charged or discharged for 18 months and shows 13.28 volts with every cell showing 3.32 volts. Self discharge is much lower than advertised, capacity higher than advertised and no BMS. Best battery purchase I've ever made and looking forward to doing a new project based on these excellent cells.

[transmitterdan]

Originally Posted by SV THIRD DAY

Oh sure...that's how the chemistry works...but what does that mean in the real world? I can put my chemistry degree to work and sound educated all day long. But what does that mean to a cruiser on a boat that has his fancy dancer BMS controller fail on him and drained his LiFePO4 battery bank down? Or what about your system (which sounds exactly like the system I have on my 400AH test bank No 2 by the way...easy and clean) what happens if it fails? Because for every piece of electronic equipment being used on a boat and in the world today, there is a product failure statistic rate saying there will be a measurable failure rate. So what then? A new bank? Your boat blows up? Your mom cuts you out of the family fortune for being the black sheep son that went out cruising in the first place? What...no one wants to really know?

That's the problem no one wants to do the tests and gather the data that matters to know if the scary sounding chemistry matters in how we are using these banks on a boat.

But they have done the tests. The problems are real and the concerns you raise are exactly why there are no commercially available affordable Li large format batteries and charging units. The liability is just too high for the big guys (in the US West Marine, Defender, fisheries, et. al.) for this technology to be sold to the everyman boater. For many (most?) cruisers we can handle the issues. But if even 1% of retail customers have problems it's too great a risk for big box retailers.

Do we think if it were possible to make lots of $ selling these systems the big retailers would just leave that market alone?

[SV THIRD DAY]

Originally Posted by transmitterdan

Do we think if it were possible to make lots of $ selling these systems the big retailers would just leave that market alone?

There's no real money in selling LeFePO4 battery systems to the Cruising Community/Boat world...not on a Risk/Reward basis anyway.

Take Mainsail for example...there probably isn't a person here on the forum with as much real life test data and understanding of the systems and will he sell you a package drop in system? Nope...Why?

[Jd1]

Minor thread drift warning

I have previously posted that I am using the BMV 700 SOC meter to monitor what my battery is doing. It is a brilliant meter with all kinds of functionality including history data and it can be used to control the charge bus relay on a lithium setup based on SOC data. It is really the neatest thing since sliced bread.
Unfortunately, every silver lining has a black cloud to go with it
Should you loose power to the meter, although it keeps history data, it does NOT keep your current SOC data. Instead, you are greated with a flashing '-----' for everything that relies on SOC data. At this point your only choice is to manually re-sync the display which brings the SOC to 100% no matter what the actual state is. To get the meter to read accurately, you have to drive the battery SOC to your reset point or to where your real 100% point is. The meter will never display more than 100% (which is understandable) so you can't tell how close you are to the real 100% mark if you happen to know the SOC figure at the instant the meter power was lost.
IMHO this is a MAJOR faux pas because '------' tells you absolutely nothing other than power was interrupted. A flashing figure of the SOC before power was lost would tell you that power was lost while still letting you track where you were before. Once you are at a known 100% you can resync but the meter is perfectly usable until you have a chance to top up the batteries to 100%.
I have suggested this change to Victron but have no idea if they would contemplate this change (which likely would not involve a hardware change but only a firmware update)
FYI, on an experimental/homebrew install, it is amazing how often power is lost for minor reasons such as a rewire, moving something or what-have-you and it is extremely frustrating to have to re-sync the batteries on a large bank. My resync point is 14.8V pack voltage and involves a top balance. Sure I could sync to a lower figure but I prefer the 100% mark to be at the top balance point.

[transmitterdan]

Originally Posted by SV THIRD DAY

There's no real money in selling LeFePO4 battery systems to the Cruising Community/Boat world...not on a Risk/Reward basis anyway.

Take Mainsail for example...there probably isn't a person here on the forum with as much real life test data and understanding of the systems and will he sell you a package drop in system? Nope...Why?

I can't speak for Mainsail. I'm talking about vendors with tens to hundreds of millions of $ or Euros in annual revenue. They can afford to do the R&D and certification testing required to bring a product to market. And there would be huge demand for a product that cost only 50% more than lead acid. It would be pretty big if it only cost twice as much as LA.

A mass market product has to be "idiot" proof. Mastervolt has brought such a product to market. On West Marine's web site they list a 360AH 12V Li battery for just over $10,000.00.

MASTERVOLT Lithium Ion Batteries | West Marine

That's why this thread exists.

I don't know what West Marine's gross margin is on products but in specialized retail it is typically between 50-75%. Upscale shoe stores make about 50% gross margin so why should a marine store do worse than that?

So let's say WM pays Mastervolt about $5,000.00 for the battery. It probably has 8 180AH cells in series parallel that cost $250.00 each from the battery factory in China. Then they have to put all the electronics and a box around it. That probably costs a couple hundred plus that much in labor. So all in Mastervolt builds it for about $2,500.00 in material and labor and makes $2,500.00 gross profit (50% gross margin). Then WM sells it for $10K and makes $5K gross profit. Out of all that then the marketing costs, transportation costs, warranty costs, accounting, taxes, etc. come out. In the end nobody is getting rich selling a few hundred of these batteries a year.

For drop-in LiFePo to be commercially viable there has to be another application besides boating that can drive the demand up and cost down. They would have to sell several hundred thousand units a year in that other consumer oriented market. Then someone could "marinize" it and sell it to boaters at twice the mass market price. But there is no consumer market like that. The EV OEMs all want to integrate their batteries into the vehicle and don't want a "drop-in" battery because it costs too much. The power grid OEMs want huge battery banks but can't use drop-in packaged 12V consumer batteries either.

I believe in the technology for boaters. It is great to get the lead out. But it isn't a good business to be in.

[toby]

Hi, you will find all sorts of people with different ideas about this!!
2.5 years ago I had my boat in Langkawi and had Winston Batteries send me out 4x400 amp batteries by sea freight as well as a 90ah 12v, all up it cost about $2700 including freight.
I hooked the 400 amp bank up as my house bank with no BMS. I installed a steca 3030 standard solar regulator and set it to gel batteries setting. I have just over 400watt of solar. I use the standard engine alternator.
They are the best ever! Couldn't be happier to be honest. We are liveaboards on our 43 foot cruiser with fridge/freezer ect going all the time plus all the gear when doing passages. You see a lot more amps go into the batteries from solar and engine than a standard lead acid battery because they don't seem to be affected by the heat and just keep sucking up the power, they're great. I isolate each one from time to time to check they all are the same voltage and they are, I'm happy. I spoke with a guy in west Australia who has a company called EV works using these same batteries to convert standard cars to electric and he said with such a simple system no need for BMS and I'm happy with that. I have also read an article of a guy with Winston batteries and got struck by lightning and had no problems.
There are a lot of people with negative ideas about them including how expensive they are but I really don't think they are at all when you look at the useable amps. And they "say" if I discharge them to 80% will get 5000 cycles and 70% 7000 cycles hopefully I never find out and they just keep going! Who knows yet I don't drain them down too much anyway.
I could be a very good sales rep for these batteries I'm that happy with them!
Keep it simple and chuck them in is my advise!



Sent from my iPad using Cruisers Sailing Forum

[T1 Terry]

Originally Posted by Jd1

Minor thread drift warning

I have previously posted that I am using the BMV 700 SOC meter to monitor what my battery is doing. It is a brilliant meter with all kinds of functionality including history data and it can be used to control the charge bus relay on a lithium setup based on SOC data. It is really the neatest thing since sliced bread.
Unfortunately, every silver lining has a black cloud to go with it
Should you loose power to the meter, although it keeps history data, it does NOT keep your current SOC data. Instead, you are greated with a flashing '-----' for everything that relies on SOC data. At this point your only choice is to manually re-sync the display which brings the SOC to 100% no matter what the actual state is. To get the meter to read accurately, you have to drive the battery SOC to your reset point or to where your real 100% point is. The meter will never display more than 100% (which is understandable) so you can't tell how close you are to the real 100% mark if you happen to know the SOC figure at the instant the meter power was lost.
IMHO this is a MAJOR faux pas because '------' tells you absolutely nothing other than power was interrupted. A flashing figure of the SOC before power was lost would tell you that power was lost while still letting you track where you were before. Once you are at a known 100% you can resync but the meter is perfectly usable until you have a chance to top up the batteries to 100%.
I have suggested this change to Victron but have no idea if they would contemplate this change (which likely would not involve a hardware change but only a firmware update)
FYI, on an experimental/homebrew install, it is amazing how often power is lost for minor reasons such as a rewire, moving something or what-have-you and it is extremely frustrating to have to re-sync the batteries on a large bank. My resync point is 14.8V pack voltage and involves a top balance. Sure I could sync to a lower figure but I prefer the 100% mark to be at the top balance point.

Simple solutions, don't sync, better to not know what the SOC and time to go is than falsely think you have a full battery when you don't. Set the auto sync for 13.9v, 1 amp and 5 mins, it might be only 99% full, but if that last 1% is mission critical you have a seriously undersized battery bank.
As far as "zero cell voltage and damage" all batteries including lead acid are damaged by dragging them down to 0v, reverse charging caused by current being dragged backward through a cell because the other cells still had some capacity remaining does even more damage, to the point of a case bursting or the cell internally short circuit and won't recover. The 0v is worse than the 4v and the longer it is held at either of these extremes the greater the damage will be. One bounce to 0v or 4v will do little damage, 12 hrs at 0v or 12v and look for a replacement cell.

No matter how technical the system build, nothing is fool proof, the world can generate better fools than it can generate genius, genius needs the right set of circumstances to be able to develop to full potential, the fool just need to touch things they don't understand and are not willing to take the time to learn about but rather cut corners because they just know
You can lead a head to knowledge, but you can't make it think

T1 Terry

[SV THIRD DAY]

Originally Posted by T1 Terry

One bounce to 0v or 4v will do little damage

Bingo...that's exactly what my test confirmed. 5.4v for 3.75hrs before starting the recharge and Nada problem that I can see yet. But if you would believe the doom and gloom panic from some I should be out $1700 for a new 400AH bank. The irony is that in all the years we have lived aboard I've never let my bank go below 12.1v. It was an interesting data point but not a real world event I actually lose sleep over happening. Its the top end that worries me but it would take a controller failing (which can and does happen) and me missing a siren alarm that would wake the dead for the high voltage alarm.

My momma always said, "Stupid is as stupid does"...ha ha ha

[Jd1]

Originally Posted by T1 Terry

Simple solutions, don't sync, better to not know what the SOC and time to go is than falsely think you have a full battery when you don't. Set the auto sync for 13.9v, 1 amp and 5 mins, it might be only 99% full, but if that last 1% is mission critical you have a seriously undersized battery bank.
T1 Terry

You didn't get the point .... we are not talking about the last one percent. If you loose power for some reason you will not know what your SOC is or was. It might have been 80% or it might have been 20% - you don't know. You can not rely on the battery again until you have re-established your reference point.
You just pulled into a nice quiet bay for the night, You know that you will need 150 AH for the night but your SOC shows as '-------'. You have no choice but to sit there with your engine running for at least as long as it takes to shove 150 Ah into the battery. It doesn't matter that you were expecting zero wind the next day and would be forced to motor and you could have charged then. It doesn't matter that the battery still had 350 Ah in it and you could have just dropped the anchor and gone to sleep. Instead, you have to endure 2 hrs (or whatever) of running the engine. It's a major oversight as far as the BMV700 series of SOC meters is concerned.

[toby]

Good on you Terry I'm glad your happy with your system. My point is I've kept it simple and it works for me. On my steca regulator I can see SOC and how many volts it is. It never gets to 100% SOC because it only charges to 14.2v and I am happy with that, a couple of extra % doesn't matter to me just as long as they don't over charge. My batteries have never gone below 13v and when I leave the boat for an extended amount of time with everything shut off I cover the panels. Every time I walk up the companionway steps I look and see the Steca readout with voltage. It's simple


Sent from my iPad using Cruisers Sailing Forum

[T1 Terry]

Originally Posted by Jd1

You didn't get the point .... we are not talking about the last one percent. If you loose power for some reason you will not know what your SOC is or was. It might have been 80% or it might have been 20% - you don't know. You can not rely on the battery again until you have re-established your reference point.
You just pulled into a nice quiet bay for the night, You know that you will need 150 AH for the night but your SOC shows as '-------'. You have no choice but to sit there with your engine running for at least as long as it takes to shove 150 Ah into the battery. It doesn't matter that you were expecting zero wind the next day and would be forced to motor and you could have charged then. It doesn't matter that the battery still had 350 Ah in it and you could have just dropped the anchor and gone to sleep. Instead, you have to endure 2 hrs (or whatever) of running the engine. It's a major oversight as far as the BMV700 series of SOC meters is concerned.

Do you monitor cell voltages? If you have 3.25v or better per cell you have 50% or better SOC, that does not mean 13v at the battery, it's cell voltage that is important.
If you connect the positive sensing lead from the Victron shunt to a positive cell terminal within the number 4 cell group that is not the terminal the main battery cable is connected to, yet another advantage of using multiple cells to build capacity, a mains power disconnect will not drop the Victron out so it may be an installation issue rather than a Victron short coming. After all, you rely on any instrumentation to be accurate or tell you that it's accuracy is impaired at the moment, if the Victron has not been monitoring what comes in or out of the battery constantly since it's last verified 100% SOC state then any reading would be inaccurate wouldn't it, therefore it could not be trusted.

T1 Terry

[Jd1]

<sigh> ..... I give up

[chemist]

JD,

I understand that losing the SOC indication on your BMV is a pain in ..., but I would not worry about it to much, because these numbers are inacurate anyway. Let me explain.

The best way to determine the TRUE SOC of a Li-battery is as Terry suggest to measure the voltage of the individual cells of a battery, and using a calibrated V vs SOC chart to determine the SOC of each cell. The weakest cell will give you the SOC of the whole battery.

If all cells are physically identical (same internal resistance, ... ) and have the same SOC, then you can also use the voltage of the battery as an indication of the SOC thereof. Again you would need a calibrated curve of volt vs SOC. This would be the second best way to determine the TRUE SOC of a battery.

The third and least preferred way of measuring the SOC of a battery is to use a Coulomb counter (ie integrating the potential difference over a resistance over time with a shunt). However this method relies on several assumptions: 1) the 100% SOC value is an arbitrary value set by the operator, 2) the Peukert constant is an estimate, 3) the total capacity of the battery is an estimate, 4) the amps used relies on a 200 Amp/50mV shunt, which will never be accurate over the whole range (eg. high input during charge, and low output during use), and finally 5) the assumption that none of this changes over time. The combination of all these assumptions leeds to the effect that the SOC indication seen on the BMV is inacurate, especially over longer periods of time, and should only be used with caution.

I suggest that you create a calibration curve for battery of Volt vs SOC using your BMV immediately after resetting it, and use this calibration curve to periodically check if the Volt measured by your BMV still matches the SOC calculation. If not the Volt reading is correct, the SOC calculation is not, and you should reset your SOC parameters.

With the calibration curve at hand it should also not be a problem to determine within limits, i know there is a rather flat discharge curve between 80-20 SOC, what your TRUE SOC is should your BMV loose its data again.

Ofcourse Mainsail already wrote a nice review on this.

Hope it helps.

Bas.