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

[deckofficer]

BTW, good advice on your copy & paste article. The #1 killer of all batteries is leaving them in a deep discharge for any length of time.

[Lagoon4us]

Below is the return email Pavel sent me, interesting to note the initial charging voltage he recommends, it contradicts most on this thread, these guys are quite qualified to make recommendations what thoughts? Why the difference?? Cheers


Hello Frank,


thank you for email.

1. BMS123, cheaper variant to RT-BMS including all functions and is even better for car/boat/anything that moves projects. You would need control unit for the pack, 4 cell units and I would recommend current sensor for more detailed monitoring also. You can check all here : EV-Power | BMS123 System

2. Ideal relay might be EV-Power | DC Power Contactor 100A, Coil 12V or EV-Power | DC Power Contactor 200A, Coil 12V

3. Before first use, it is important to charge the cells to 3,95V. Optionally you could use single cell charger EV-Power | Charger 3.6V/6A for LiFePO4 cells (1 cell) or any other charger with manual control to prevent the damage. No further processing is required, just keept he BMS connected for the rest of use and it should be fine.

4. balance at low voltage is not required, the cells are rather cut from the load once any of them reach their minimum to prevent damage

MPPT regulators are perfect for LiFePO4 cells.

For much much more information about the BMS and actual connections schemes, check our blog : lithium & solar power LiFePO4, Search results for: BMS123

Let me know if you have any further questions or if you decide which parts to order.


Best regards

Pavel Radomersky
Export department

i4wifi a.s.
Prumyslova 11/1472
102 00 Praha 10
Czech Republic

[deckofficer]

An interesting piece from Capt Mike motor sailing the Hudson River going to NYC. In the beginning before he had any wind or current assist, it looks like he was able to do 3.2 kt on about 720 watts from the battery. I couldn't help but calculate his range if he bought those 700 a-hr cells that Balqon still has on their clearance sale. At 3.2 kts with zero currents or wind assist, Capt Mike could motor for 160 nm on (16) 700 a-hr cells in series for his 48 volt propulsion motor.
THE BIANKA LOG BLOG: ELECTRO SAILING AROUND NEW YORK: PART ONE

[deckofficer]

Frank,

On his point #3 he says 3.95 volts for that first charge but then says his 3.6 volt 6 amp charger will work fine too. As to why I picked 3.7 volts per cell is because that would be the cell's voltage if all are in perfect balance, when I use my 14.8 volt 50 amp quick charger. My pack #1 which I deemed fine on arrival, I will now perform the same bottom to top balance that got pack #2 with its two errant cells so nicely balanced. To speed things up, as per your email, I'll charge pack #1 back up at 3.95 volts as recommended by your shared email. Thanks.

[T1 Terry]

So, no confirmation on that claim the manufacturer demands a BMS unit be used with their cells. Reseller will add this requirement because they also sell BMS units. It's not "would you like fries with that" it's "you must have fries with that"

I think it's time the whole BMS thing was explained again, it's a term trotted out quite often and then linked to something someone is selling.
We all use a BMS system, no matter what sort of batteries we use. BMS is just a Battery Management System, as soon as a voltage regulator is attached to an alternator we have fitted a battery management system, we plan to limit the max volts the battery will charge to, a form of management.
The only difference with lithium ferrous batteries we can now monitor individual cell voltages, we couldn't really do that before, this gives us a much better feel for how the battery is travelling, is it full, is it over charged, is it empty, we know this at a cell level if we monitor them. The only way you know there is a problem when only the terminal voltage is monitored is when the voltage suddenly drops, one cell is flat.
If you don't want to monitor each cells that's fine, just don't expected the very long cycle life, if you don't charge at an excessively high voltage and limit the low voltage you stand a good chance of never pushing a cell outside it's comfort zone. 13.8v tops and 11.8v bottom and you will be fairly safe that an individual cell hasn't wandered outside it's comfort zone, you just don't know that for sure.... If the battery only last 2 or 3 yrs and you are happy to replace it then I won't bother with monitoring cell voltages, but for the small cost of a Junsi cell logger I can't see the logic in that line of thinking.

As far as reverse charging a cell in a 4 cell in series battery, with a cut off point at 11.8v it's not possible, even 10v it would not be possible to reverse charge a cell, but you could pull one to near zero volts and that does damage them, especially if they held at that voltage for an extended time.

Would I ever go back to lead acid? As far as throw away priced batteries for starting functions, lead acid serves a purpose, for house batteries, lead acid was never fit for purpose, we just learnt how to operate within the limitations and accepted they would fail long before they should. I have a collection of still functioning lead acid batteries building, I don't ever want to be a servant to a battery again and have to live within strict limitations so I would never switch back to lead acid.

T1 Terry

[T1 Terry]

Originally Posted by deckofficer

Frank,

On his point #3 he says 3.95 volts for that first charge but then says his 3.6 volt 6 amp charger will work fine too. As to why I picked 3.7 volts per cell is because that would be the cell's voltage if all are in perfect balance, when I use my 14.8 volt 50 amp quick charger. My pack #1 which I deemed fine on arrival, I will now perform the same bottom to top balance that got pack #2 with its two errant cells so nicely balanced. To speed things up, as per your email, I'll charge pack #1 back up at 3.95 volts as recommended by your shared email. Thanks.

If you plan to do the 3.95v conditioning charge, don't walk away from the charger without turning it off once the voltage goes over the 3.7v mark, from 3.8v onwards the graph line will be near vertical. Don't do this if you want to use the cells as propulsion or traction batteries, the resulting coating of the cathode? (anode) will limit max ion flow resulting in voltage sag at higher amp draw, it isn't a problem with house batteries though, we are never likely to expect more than 0.5C and a bit of voltage sag can be handled by an inverter anyway, that's only thing that would pull those sort of high amps in house battery situation.
Don't do this condition charge more than the first time, limit the max cell voltage to 3.6v if you want a long cycle life, continually recoating will reduce the amount of free lithium ions available in the electrolyte. The conditioning charge does virtually stop the build up of dendrite growths though greatly extending service life.

T1 Terry

[goboatingnow]

Originally Posted by Lagoon4us

BMS only there as 'protection of last resort' below is what EV-Power state:-

Some SBM models monitor the temperature of the high-current component to detect the high current over-discharge. However the SBM boards do not have any temperature sensors to monitor the temperature of the battery pack.
Warning: Please note that that the lithium cells need to be charged regularly if protective electronics like SBM are connected to the cells. The lithium cells need be charged to full before letting them stay unused for longer periods of time.

I'd question this last line of advice. Li technology is actually best charged to about 80 % for long shelf life and some manufacturers recommend 40-60% SOC for long shelf life. While Li ferrous don't suffer the life degradation that other Li types do they still degrade at full charge. That's why float charging them is a bad idea.

Dave

[deckofficer]

Originally Posted by T1 Terry

If you plan to do the 3.95v conditioning charge, don't walk away from the charger without turning it off once the voltage goes over the 3.7v mark, from 3.8v onwards the graph line will be near vertical. Don't do this if you want to use the cells as propulsion or traction batteries, the resulting coating of the cathode? (anode) will limit max ion flow resulting in voltage sag at higher amp draw, it isn't a problem with house batteries though, we are never likely to expect more than 0.5C and a bit of voltage sag can be handled by an inverter anyway, that's only thing that would pull those sort of high amps in house battery situation.
Don't do this condition charge more than the first time, limit the max cell voltage to 3.6v if you want a long cycle life, continually recoating will reduce the amount of free lithium ions available in the electrolyte. The conditioning charge does virtually stop the build up of dendrite growths though greatly extending service life.

T1 Terry

As always, glad your here to help out as you were the real early adopter. The service these cells will see is the very light load of a kayak propulsion motor, so around 375 watts, which is nothing. Since the cells were 2 1/2 years old when I got them, then stopping dendrite growth for extended service life would be a goal.

Thanks again Terry.

[Lagoon4us]

Originally Posted by goboatingnow

I'd question this last line of advice. Li technology is actually best charged to about 80 % for long shelf life and some manufacturers recommend 40-60% SOC for long shelf life. While Li ferrous don't suffer the life degradation that other Li types do they still degrade at full charge. That's why float charging them is a bad idea.

Dave

Makes sense. Cheers

[T1 Terry]

They appear to ship the cells at around 50% SOC and they are still at close to that yrs later. The self discharge chart on the Winston site shows a 10% discharge over the first 60 days, the next 10% over 100 days, the next 10% over over 160 days, it still appears to be around 65% at the end of 360 days. I'm guessing charging or discharging to around the 60% mark would allow a bit of settling without them self discharging much further than that, but there doesn't seem to be much data about regarding long term storage, just the findings of those who have bought cells that have been on the shelf for a long time.
I'm guessing Balqon are clearing out all the LFP cells with the old Thundersky logo on them, 3 yrs plus is a long time to have stock not bringing in any returns.
Pity the freight out of the US is so expensive, those 1,000Ah cells look so inviting.

T1 Terry

[diugo]

More theory for ya, my 25-month-old cells were at 20% SoC when I received them. At 3% discharge per month, that suggests they were shipped at 0.20 * 1.03^25 = 42% SoC.

[NahanniV]

Made a wiring harness for individual cell monitoring today.

Disconnected the paralleled cells and reconfigured them in series.
The individual cell voltages immediately drifted apart by about 10mV like they were before I did the parallel balance charge. They had charged in parallel for days (up to 3.75V) and then I left them paralleled for another day.

Next I tried them with a load. I was aiming for 20A with a heater element, but it was a bit smokey, so I shut it down quickly (see first graph). Second attempt was with an 8A load.

Looks like one cell has at least twice the internal resistance of the others. How bad is that ?

My test rig is not quite up to unattended operation, so I did not start a load test, maybe tomorrow.

[T1 Terry]

Originally Posted by diugo

More theory for ya, my 25-month-old cells were at 20% SoC when I received them. At 3% discharge per month, that suggests they were shipped at 0.20 * 1.03^25 = 42% SoC.

What brand and how did you determine they were at 20%SOC? I would be worried if I received a cell below 3.2v still fresh in the box.

T1 Terry

[deckofficer]

The weight of those 1000 a-hr cells are higher than proportionally expected.

1000 a-hr cell 92 lbs.
700 a-hr cell 47 lbs.

I'm guessing a much thicker case?

As much as I preach for single series strings, if I was to do hybrid diesel-electric propulsion on a cat, I would go with (2) 48 volt banks of the 700 a-hr in parallel than a single bank of the 1000 a-hr.

[T1 Terry]

Originally Posted by NahanniV

Made a wiring harness for individual cell monitoring today.

Disconnected the paralleled cells and reconfigured them in series.
The individual cell voltages immediately drifted apart by about 10mV like they were before I did the parallel balance charge. They had charged in parallel for days (up to 3.75V) and then I left them paralleled for another day.

Next I tried them with a load. I was aiming for 20A with a heater element, but it was a bit smokey, so I shut it down quickly (see first graph). Second attempt was with an 8A load.

Looks like one cell has at least twice the internal resistance of the others. How bad is that ?

My test rig is not quite up to untended operation, so I did not start a load test, maybe tomorrow.

If you can repeat the first test but with the scale closed up it will tell you a lot more. you need to be able to see down to 3v and up to 3.45v, that is the cells operating range, any voltage higher than that is just surface voltage and will burn off very fast, the turn internal or external resistance won't show up till the cell voltage drops below 3.35v, this is the load voltage of a fully charged cell, 0.5v either side of 3.4v shows the point of effect of load or charging on a fully charged cell. If you find a cell reading lower than the others, check the link plate temps around that cell, you may have a high resistance joint giving you the odd readings, also double check the cell logger readins at the cell itself with a multimeter, spent hrs trying to find a fault in the batteries that was actually a crap wire join once, I always double check now.

T1 Terry