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

[200MPH]

SoC readings taken from BEP battery monitor. It's not perfect, but it's repeatable. My test methodology can be thought of as pretty reasonable for an 'absolute' test, but perfectly acceptable as a 'relative' test. I could have used the DataTaker to also measure voltage drop across the main boat shunt resistor and used it to calculate amps drawn and Ah used, but just decided to keep it simple as time was limited and I had other things to do (like get wind controller set correctly).

Why did I use about double the typical Amp draw? Time constraints, that's why. In any case, it's still a fractional draw on a battery type that has minimal Puekerts effect, so not hugely relevant. A C20 rate would be more relevant if performing this test on a LA battery, sure.

Test took 7:54. I'm aware this is a little less than my estimated 8:15 based on current draw.

[john61ct]

And in fact at 12.1V you still had significantly more AH capacity left at the end, your actual capacity may be a lot higher than rated.

Or the BEP meter is off?

[nebster]

Originally Posted by 200MPH

It also verifies the 12.8V low voltage buzzer alarm I've setup so that everyone who uses the yacht gets a notification at around 80% discharged.

One of the challenges with a single low voltage buzzer is that, if your system has typical wiring and internal cell resistance, you'll find that the voltage drop due to load can really throw off the effective SOC at the threshold.

For example, on my LFP pack with a 500W load, 51.2V also means I have somewhere around 20% SOC. But with a 10kW load, a reading of 51.2V means there is more than twice the capacity left.

Since the SOC gauge is really there for convenience and (the rest of the) non-technical users, it would be nice if there were an implementation that would adapt based on the load, a defined Ri, and perhaps a defined hysteresis interval.

And as long as we're dreaming... it would also be nice if the inverter would allow its LV cutouts and warnings to shift based upon load, too. Unless I want to set a low cutout that would damage the cells under low loads, I end up with one that is breached too easily at 10 or 15% SOC under very high loads. And of course a bunch of spurious LV warnings and emails before that.

[200MPH]

Nebster - correct about wanting/needing a low voltage alarm/cutoff that varies with load. Would be easy to achieve with a simple logic lookup table. Wonder if any existing battery monitors incorporate such functionality? If not then I guess you could make your own with a little effort.

For instance, my low voltage alarm briefly sounds when starting the diesel engines. Yet the Lead starter battery is 100% OK. It's all a compromise I guess! Ultimately I'm happy knowing what's going on with the yacht systems and can deal with alarm systems that are not 'inteligent' or 'learning'. Not sure the other partners in the yacht are, so I've set the sytems to be conservative for that reason.

[OceanPlanet]

Originally Posted by 200MPH

So, I did a full discharge test on the new battery pack last night so I would have a baseline test I can consistently repeat every year to see how much the battery is degrading.

Battery: 300Ah Winston 4 cells to make 12.8V.

Charged to 13.8V setting on Victron charger, using dedicated voltage sensing function it has. Bulk charge 90A, then absorbtion for 1 hr (it can't terminate on trailing amps), let battery settle for a few minutes while I fooled around getting the wind controller set correctly, then start a constant discharge overnight. Load was a radiant heater run through the Victron AC inverter. 465W or thereabouts (35A at 13.2 V). No other loads during discharge test.
Victron low voltage inverter cut set to 12.0V. Capacity during discharge monitored using the yacht BEP battery monitor, Pukerts set to 1.03 as an estimate.
Battery voltage logged from the pack terminals (pack level, not individual cells) using a DataTaker DT85, logged every 15 seconds. Before you ask, it is a professional quality precision logger, that is way, way overkill for this application. So I'd trust the results of it pretty much over every other reading. We calibrate our fleet of these instruments regularly against national standards.

The short version of the results are:

Volts (mV) SoC % Capacity (Ah)
13473 100 300
13197 90 270
13159 80 240
13078 70 210
13029 60 180
13000 50 150
12958 40 120
12874 30 90
12774 20 60
12640 10 30
12131 0 0

Interestingly the Victron cut the AC inverter a bit above 12V (12.131V) within a minute of the BEP SoC reading reaching 0 Ah. Coincidence of course, couldn't have planned it that way.

I don't know how to post a picture of the graph otherwise I would.

The above values are very, very similar to those published by MaineSail from a test done last year on a 400Ah Winston battery discharged at 30A, although that test went down to 11.6V and hence showed more than 100% nominal capacity.

What does all this prove? Not much, but a good baseline for me and just thought I'd share it. It also verifies the 12.8V low voltage buzzer alarm I've setup so that everyone who uses the yacht gets a notification at around 80% discharged. The actual DC and AC voltage cut levels are set lower than this of course, but I figure we don't need to discharge it to less than 20% SoC in normal use.

Excellent test, well done!

[john61ct]

Balmar asking for VR ideas

http://www.cruisersforum.com/forums/...d.php?t=194917

[Recy]

I have a 360 Ah 12v LFP bank and a Victron Multi Plus 12/1600/70 charger.
When charging there is only max 47A even if I start charge below 12V.
I had expected the charge to be close to 70A under these conditions.
What can be reason?

[DotDun]

Originally Posted by Recy

I have a 360 Ah 12v LFP bank and a Victron Multi Plus 12/1600/70 charger.

When charging there is only max 47A even if I start charge below 12V.

I had expected the charge to be close to 70A under these conditions.

What can be reason?

House is using 23A ??

[Recy]

No there is no load on the batteries they are only being charged.

[JayH]

Originally Posted by Recy

I have a 360 Ah 12v LFP bank and a Victron Multi Plus 12/1600/70 charger.
When charging there is only max 47A even if I start charge below 12V.
I had expected the charge to be close to 70A under these conditions.
What can be reason?

The charge current of the MultiPlus is limited to 75% of the max by default. You may need a computer and their USB adapter to change it. I don't see it mentioned in the 12/1600/70 manual as configurable by DIP switches.

https://www.victronenergy.com/upload...L-FR-DE-ES.pdf

[Recy]

I have the usb adapter so I was able to change it to 70.
Do you know the reason for setting it to 75% by default?

[JayH]

Originally Posted by Recy

I have the usb adapter so I was able to change it to 70.
Do you know the reason for setting it to 75% by default?

The manual for the 24/3000/70 says this:

5.5.1.3 Battery charge current limitation (default setting 75%) For maximum battery life, a charge current of 10% to 20% of the capacity in Ah should be applied. Example: optimal charge current of a 24V/500Ah battery bank: 50A to 100A. The temperature sensor supplied automatically adjusts the charge voltage to the battery temperature. If faster charging – and a subsequent higher current – is required: - The temperature sensor supplied should always be fitted, since fast charging can lead to a considerable temperature rise of the battery bank. The charge voltage will be adapted to the higher temperature (i.e. lowered) by means of the temperature sensor. - The bulk charge time will sometimes be so short that a fixed absorption time would be more satisfactory (‘fixed’ absorption time, see ds5, step 2).

[john61ct]

Originally Posted by Recy

I have a 360 Ah 12v LFP bank and a Victron Multi Plus 12/1600/70 charger.
When charging there is only max 47A even if I start charge below 12V.
I had expected the charge to be close to 70A under these conditions.
What can be reason?

What are you using to measure amps, and at what point(s) are you measuring?
Have you checked on the Victron's charge profile programming? Any derating options set there?

Link to your cells?

Is there a BMS or other devices between the bare cells and the charger outputs?

Link to those too please.

Have you checked for resistance in the wiring, make sure properly crimped terminations, no loose connections etc ?

What do you mean by "start charge below 12V"? Is that the resting voltage of the cells when isolated in a depleted state?

Have you ever allowed the cells to go dead flat?

[evm1024]

Measure the voltage under charge at the battery terminals and then again at the output of the charger.

I am assuming that the chargers voltage sense is internal. Thus, if you have too small of gauge wire from the charger to the bank the voltage drop across the charge cables will not give the charger proper feedback.

I use 2/0 charge cables to my LiFePO4 bank and even then I see this happening with my Xantrex inverter/charger.

[Recy]

The problem is solved.
Victronic sets the charge level by default to 75%.
I have now reset it to 70A so now there is full charge.
Thanks for all your advices.