LifePo4 Voltage at .5C - .6C Discharge

[otrfun]

Our two GC2's (with 100% charge) drop below 12v while under a 100a - 125a load. When they reach 50% (at 100a - 125a), voltage drops into the low 11's while under-load.

What kind of under-load voltage can I expect with two 100ah LifePo4's (paralleled) with a 100a - 125a load? At 100% charge? At 10-20% charge?

[BFL13]

Ha! He said knowing the SOC tells you how much "power" is available. I think he meant to say, "energy".

I don't see why it is so hard to use the BB table just because it is different from a FLA's. An AGM's table is different too, and so is an SiO2's. I thought the issue is the flat mid range of the LFP table.

[jaycocreek]

Will Prowse says you can tell the state of charge for a Lifepo4 battery,or get a good idea without a load..He is talking about the Battleborn..

Will Prowse on SOC lifepo4

[3_tons]

Not totally sure how my stand alone meter arrives at SOC, though it would seem that with LFP coulomb counting would play an important role...

3 tons

[FWC]

If otr is considering LiFeBlue, their app provides SOC. I think most BMS have the ability to calculate SOC as they are measuring the current anyway so it is only a few lines of code to calculate SOC. It is just a matter of a having a way to display it. With bluetooth, that is easy.

[BFL13]

Trillium has a sort of SOC gauge. (Rustofer said he has 110AH LFPs, so maybe these?) I don't know if it uses voltage and if it does, whether it can tell it is a loaded voltage vs resting voltage.

Using the Trillium's gauge, and wanting to know if it safe to start a 10 minute or so 100a load (on one batt) you would not run it if the yellow blinking lamp is on at 35% SOC and you wanted to be above about 32%.

With two of those batts with 220AH, now your 100 amp load is much less of a C rate, so you could pick a lower zone, but that 35 -20 is the same lamp. Solid red indicator to start off would be for the brave or desperate I guess. 🙂

IMO, I would rather have an AH counter to get a better idea of SOC.

http://www.wegosolar.com/products.php?product=Trillium-Battery-%252d-Trojan-Intelligent-Lithium-12.8...

[FWC]

Resting or open circuit voltage provides a very reasonable estimate of SOC for lead-acid batteries, as the SOC vs Voltage curve is quite steep and fairly linear. Most every lead-acid battery manufacturer provides this data. This is not the case with lithium chemistries, and BB is the only company that I have seen provide a similar table, probably because folks coming from lead acid want to see this (even if it is not useful). For LiFePO4 the change in voltage with SOC is so small that it is often eclipsed by changes due to temperature.

I have been using LiFePO4 batteries in a series of campers since 2010 or so, and various lithium (and lead-acid) chemistries in various instrumentation professionally.

[pianotuna]

FWC wrote:
I am not sure what your obsession with this is, but as we have tried to explain before, you cannot use voltage to determine the SOC for a lithium battery. Anyone who with experience with lithium batteries knows this.

The only successful device that used voltage to determine state of charge for lead acid batteries (smartgauge) doesn't work for LI chemistries. It probably never will, even though it checked voltage 60 times per second. So other than that "one" device, voltage is a pretty poor way to "guess" at state of charge for most battery chemistries.

All the other meters 'drift' over time because capacity is a "moveable feast" being affected by temperature, load and age.

What is your experience with Li?

[BFL13]

Whatever the V/SOC on the almost flat curve, the OP wants to know how low he can go and still run the 100a load before voltage crashes.

We seem to all agree he wants the 100a run to end at about 15% SOC, so when can he start? 100a for 10 minutes is 16.7AH so 15AH +16.7AH is still 31.7AH = 32% SOC.

Somehow or other you have to be able to tell your SOC so you can tell if you are below 32% in this example. If voltage doesn't tell you, then AH counting is needed.

FWC is saying BB's voltage/SOC table is bogus. It looks pretty flat along the mid-range. Whatever.

Note that the inverter alarm and shut-down uses voltage. You need to have some idea what the voltage is per SOC so you can figure what the inverter will see, while you are using SOC.

[FWC]

I am not sure what your obsession with this is, but as we have tried to explain before, you cannot use voltage to determine the SOC for a lithium battery. Anyone who with experience with lithium batteries knows this.

Look at the discharge curve posted earlier in this thread - as soon as you use an LiFePO4 cell it will drop to ~3.3V/cell (or ~13.2V for 4 cell pack) at around 98% SOC and will stay there until ~10% SOC.

At a 1C discharge rate each cell will be about 3.25V, or 13V for the pack, so the Lifeblue BMS/wiring is dropping about 0.3V at 100A, or 3mOhm. At 15% SOC and 1C the cell voltage is ~3.1V, so the pack would be 12.4 - 0.3V for wiring = 12.1V. Right on with OTRs estimate.

[BFL13]

otrfun wrote:
Appreciate everybody's inputs. Always learning something new.

Called LifeBlue (lifebluebattery.com) on Mon and inquired about the voltage drop with their 100ah LifePo4 battery while under heavy load. Surprisingly, he offered to test one and get back with me. Received an email yesterday with the results. With a fully charged battery, battery voltage was 12.8v at 80a, and 12.7v at 100a. Pretty impressive. Based on the graph FWC posted, I'd guess-estimate battery voltage while would drop to approx. 12.0v at 15% SOC with a 100a load.

Using BB's v/SOC table, 14% SOC is 12.5v, and 9% is 12.0 resting. But you want the loaded voltages while the inverter is running that 100a load to figure this out.

Using BB's full at 13.6v (assumes your batt uses the same--been told here the LFPs all do ?) then initial drop was 13.6-12.7 = 0.9v

You have fat wires to inverter, so call total drop 1v for this exercise. that drop happens whatever the SOC is when you start the run.

Next thing is the drop while running due to the voltage going down as AH are removed from the batt. We have the BB v/SOC table for that too.

Say you run the 100a load for 10 minutes. that is 16.7AH down from whatever SOC you started at.

Staying above the knee so the voltage does not crash from say that 14% at 12.5v (loaded will be 11.5v then, so above the 11v inverter alarm), and 14% is 14AH remaining of your 100, then your 10 minute run has to start at 14 + 16.7 = 30.7 AH ( 13.0v resting and 12v loaded)

If you start the 100a 10 minute run below 30% SOC, your voltage will crash before the run is done. And don't let the furnace come on during the run!
--------------

Meanwhile, what do we know about the internal R of that battery now that affects that initial voltage drop? R = V/I

Note: somebody mentioned that the R of an LFP included the wiring etc with the BMS, so it could be higher than you would expect.

We have 0.9v from 100a so R = 9 mOhm a 100AH AGM is rated at 5 mOhms so 9 seems too high.

We have 0.8v for 80 amps so R = 10 mOms. So try it using the diff between them where 0.1v is 20 amps so R = 5 mOhms.

That is better, same as an AGM, so that means the starting voltage of 13.6 was too high for 100%.

If R is 5 mOhms and I is 100a, then V is 0.5v so 12.7 + 0.5 = 13.2v

That 13.2 looks awfully familiar! Meanwhile BB says 13.2v resting is 70% SOC.

Beats me! You LFP guys can figure it out if you need to 🙂

[otrfun]

Appreciate everybody's inputs. Always learning something new.

Called LifeBlue (lifebluebattery.com) on Mon and inquired about the voltage drop with their 100ah LifePo4 battery while under heavy load. Surprisingly, he offered to test one and get back with me. Received an email yesterday with the results. With a fully charged battery, battery voltage was 12.8v at 80a, and 12.7v at 100a. Pretty impressive. Based on the graph FWC posted, I'd guess-estimate battery voltage while would drop to approx. 12.0v at 15% SOC with a 100a load.

[3_tons]

My Keurig draws about 124a from a single 200a/h LFP, but this last for less than 2 min - repeated cups are no problemo. With inverters, proper battery cable sizing is paramount to reduce voltage drop...I went a bit overkill with 0004 aught, 3’ length...

3 tons

[Rustofer]

I have 2 110 Ah lithiums, currently at half-charge. With minimal load (propane fridge and a couple USB chargers running), the batteries are at 12.9 volts. My inverter and household toaster draws about 75 amps, pulling the batteries down to 12.2 v. I switched the fridge to 12 v as an experiment, so now drawing 95 amps, and got 11.4 v. In the past, using a Keurig, I’ve seen 11.8 v on the batteries. These days I’m making coffee on the propane stove.

[pianotuna]

otrfun,

Using a 12 volt form factor rather than twin 6 volt batteries improves greatly on voltage drop under load, simply because there are twice as many cells.

Some LI battery management systems limit discharge rate to 1 C.

There are a host of other hoops to jump through with LiFePo4.

That said, the LI chemistries behave much like the old Nicads--where they supported the voltage well and then dropped off a cliff.

My next battery bank will be SiO2 chemistry because they meet my needs, do not sulphate and there are many fewer hoops to jump through.

Any battery type can be designed to fit the needs of a particular end user.

Please wire your new bank in a balanced manner.

[3_tons]

This is just one reason why I finally chucked my Vector charger (labeled B&D) into the garbage can - it would always dump a code to find a reason not to work regardless of battery type - BD service was a waste of time...For some strange reason, I found myself feeling much better after tossing it in the garbage... No more B&D for me!!