How to determine if Lifepo batteries are bad

Sonnyboy said:

My question was, assuming I do the very thing you are suggesting, what are the next steps in diagnosing the battery?

Click to expand...

I watched the video and that does look bad.

You need to measure your battery with a digital meter and see what the real battery voltage is. Maybe it’s the MT50 or it wiring. Maybe the batteries are actually fluctuating (unlikely I think).
What does the Epever display show when this happens?

But you need to narrow down the source of what this MT50 is showing.

Sonnyboy said:

Thanks and I appreciate the response. I understand everything you are suggesting. But it doesn't answer my question. My question was, assuming I do the very thing you are suggesting, what are the next steps in diagnosing the battery?

Additional info: I have three identical Weize 100ah batteries all claiming(and reinforced by a Will Prowse review) to have temp protection built in, in my model anyway. I just looked at the one installed in my van with a Renogy system(20a Rover scc). The battery was steady at 14.7v. I disconnected it and connected one of the batteries that I am trying to troubleshoot. It was reading 12.4-5 volts. Once I connected it, the scc immediately read 14.6-7 volts. I checked the battery that had originally been in the van, and disconnected it now read 12.5 volts. Does that make any sense?

When I removed the second battery from my shed and now only have one 100ah lifepO4 in there, it seems to charge to 14.1 volts but then bounces up and down between 12.5 and 13 and then 14 again. All with no load.

Click to expand...

I can't see that you've ever fully charged these batteries. If that's the case, then they are almost certainly grossly out of balance resulting in vastly diminished capacity. You may literally have both full and empty cells in the pack. If so, they may exhibit erratic behavior during charging and discharging or just sitting there for that matter.

In retrospect, because we're 3 pages in, and there hasn't been any movement on your part, if the batteries are under warranty, I recommend you contact them for warranty replacement. If they are truly as out of balance as I indicate above, it may takes weeks/months to get them sorted.

Until you have the means of reliably charging the battery to a known full state, or at least can characterize the behavior, you can't do any diagnosis. I know you want a black and white answer, but it's a process. The first step is a controlled charge. This first step may be all that's needed to diagnose, i.e., if the batteries refuse to take a charge under controlled conditions, they are bad. If they do take a charge, you need to test their capacity in some way. This requires a means of measuring their output. You don't have that either as far as I can tell. @Consumerbot3418 recommended a $21 tool that will allow you to monitor both charging and discharging and measure capacity.

Hopefully, you can be convinced to take action before we get to page 4.

MisterSandals said:

I watched the video and that does look bad.

You need to measure your battery with a digital meter and see what the real battery voltage is. Maybe it’s the MT50 or it wiring. Maybe the batteries are actually fluctuating (unlikely I think).
What does the Epever display show when this happens?

But you need to narrow down the source of what this MT50 is showing.

Click to expand...

Using a multimeter, while witnessing the battery icon on the MT50 showing full battery and 13.8v, the multimeter also read the same, within .05 volts. Every so often, the MT50 would report a spike in battery voltage to 15-16v and then it would abruptly drop to 12.5 and stay there for a while before returning to somewhere between 13 and 13.8 volts.

sunshine_eggo said:

I can't see that you've ever fully charged these batteries. If that's the case, then they are almost certainly grossly out of balance resulting in vastly diminished capacity. You may literally have both full and empty cells in the pack. If so, they may exhibit erratic behavior during charging and discharging or just sitting there for that matter.

In retrospect, because we're 3 pages in, and there hasn't been any movement on your part, if the batteries are under warranty, I recommend you contact them for warranty replacement. If they are truly as out of balance as I indicate above, it may takes weeks/months to get them sorted.

Until you have the means of reliably charging the battery to a known full state, or at least can characterize the behavior, you can't do any diagnosis. I know you want a black and white answer, but it's a process. The first step is a controlled charge. This first step may be all that's needed to diagnose, i.e., if the batteries refuse to take a charge under controlled conditions, they are bad. If they do take a charge, you need to test their capacity in some way. This requires a means of measuring their output. You don't have that either as far as I can tell. @Consumerbot3418 recommended a $21 tool that will allow you to monitor both charging and discharging and measure capacity.

Hopefully, you can be convinced to take action before we get to page 4.

Click to expand...

Consumerbot3418 said:

I had a whole reply typed out, but @sunshine_eggo beat me to the punch--buy or borrow a bench power supply, set it to 13.8 - 14.0V, and see if you can't get those batteries topped off and balanced. This is the first step you have to take if you want to test the batteries' capacity.

You might also consider picking up a shunt that can count coulombs (something like this), and discharging the batteries one at a time, so you can compare their output.

Click to expand...

Thanks Consumerbot3418, I will pick up a shunt.

After reflecting a bit, maybe it’s best to follow @sunshine_eggo ’s advice and claim warranty, if eligible. They’d probably ask you to follow some steps to confirm the battery is actually faulty, so that way, you could kill two birds with one stone.

A shunt is a handy way to see power in use, and guesstimate the SoC. But you could also just discharge the batteries with a steady load, and time it. Watts times hours is Watt-Hours. So a 55w automobile headlight should take about 24 hours to deplete a 100ah lifepo4 battery.

Sonnyboy said:

Every so often, the MT50 would report a spike in battery voltage to 15-16v and then it would abruptly drop to 12.5

Click to expand...

I dunno, sounds a little like this

Sonnyboy said:

Using a multimeter, while witnessing the battery icon on the MT50 showing full battery and 13.8v, the multimeter also read the same, within .05 volts. Every so often, the MT50 would report a spike in battery voltage to 15-16v and then it would abruptly drop to 12.5 and stay there for a while before returning to somewhere between 13 and 13.8 volts.

Click to expand...

MT50 can't show full battery. It's % charge meter is essentially worthless. It's purely voltage based and can't be trusted unless it's measure a lead-acid battery that has sat for 24 hours with no charge or discharge.

Surge is almost certainly caused by BMS over-voltage protection as evidenced by the spike as the MPPT can't adjust fast enough to prevent it.

The low voltage reading is very concerning as over-voltage protection normally still allows the battery to discharge while preventing charge. Low voltage at the battery implies low-voltage discharge protection as that will drop the battery voltage.

sunshine_eggo said:

At this point, you have what appears to be poorly performing batteries, and it's been established that you don't have suitable equipment for charging. The answer to your subject line is, *YOU* can't. *YOU* lack any diagnostic capabilities whatsoever.

If you don't want to buy a Victron, go cheaper and get the 30V/10A adjustable power supply that Will recommends on his page. You will become master of all voltages and currents between 0-30V and 0-10A. This will enable you to set exact voltage and current limits and observe actual voltage and current real time. This can double as a "dumb" battery charger as well. You just have to set the constant voltage level.

If you want to get the same function and include 3 phase charging, the Victrons are great. You can set most to operate in a constant voltage mode achieving similar functionality as the power supply, but only for voltages more appropriate for 12V. With BT, you can monitor voltage and current on your own.

These things may need to be held at elevated voltage for a week or more if balance is the issue.

Click to expand...

I’ve been searching for guidelines on bench charging the lifep04 with one of the 30v/10a power supplies. Does anyone have advice there?

Sonnyboy said:

I’ve been searching for guidelines on bench charging the lifep04 with one of the 30v/10a power supplies. Does anyone have advice there?

Click to expand...

13.8V for 24 hours monitoring current. If it drops abruptly to 0A, it's protecting cells. If it's a smooth taper to 0, it's likely good to that point.

ALWAYS set your voltage when the leads are disconnected.

when used in conjunction with the battery monitor mentioned, you will see the actual # of Ah that goes into the battery.

Quickest way to test the quality of a battery is to do a load test. The load needs to be high enough and last long enough for battery to reach equilibrium at the discharge current. LFP should be charged between 40% and 80% state of charge and allowed to reach rested stable equilibrium open circuit voltage. If dealing with a self-contained 12v battery with four series cells you also need to ensure they a fairly well balanced. Test should be done at 20 to 30 degrees C. Voltage slump increases at cold temperatures.

For LFP battery equilibrium will take 60-180 seconds. Wait 5 mins to get rested open circuit, no load voltage. Load current should be 0.2-0.4 C(A) and measure voltage, directly on battery terminals, after 3 minutes with load current. Subtract 3 min loaded voltage from original open circuit voltage. The greater the voltage slump the worse the battery condition. Matched cells will have the similar voltage slump for given load current.

A LFP battery will increase in terminal voltage slump under given load current 3x to 5x over useful lifetime. The amount of voltage slump as cell ages correlates to amount of aging capacity degradation of cell. Cell voltage slump at higher load current may make cell unusable as it gets older.

Do not do load test from a fully charged battery since it may have a surface charge that falsely increases unloaded open circuit voltage.

sunshine_eggo said:

13.8V for 24 hours monitoring current. If it drops abruptly to 0A, it's protecting cells. If it's a smooth taper to 0, it's likely good to that point.

ALWAYS set your voltage when the leads are disconnected.

when used in conjunction with the battery monitor mentioned, you will see the actual # of Ah that goes into the battery.

Click to expand...

I have a DC meter and bench power supply arriving today. In the diagram for the meter, for the purpose of fully charging these batteries, do I ignore “load”? Is the bench ps hooked up directly to the battery? Or should is my power supply the “external power supply” represented in the diagram? Thanks

'Load' is where your charger connects. The L and N inputs to the monitor display are to power that unit where the battery under test is lower than 8 volts. Leave these wires disconnected.
Regard the left hand terminal stud of the shunt ( as shown in the picture) as the 'new' negative in the system, all loads or chargers connect to the battery positive and this 'new' negative.

You measurements of 14.6/7 volts at the battery connected to the charger falling to 12.5 when removed, suggests the BMS was in protection mode, perhaps due to the high voltage set in the charger.

With the new power supply and instrumentation hopefully the problems with the batteries can be resolved.

Mike

Oh gawd, that drawing is ALL wrong.
I'm just waking up, but ill sketch something for you. The battery *is* the load.

Edit: not really wrong, I was still asleep. See my next post.

So the drawing isn't totally wrong, I was mostly looking at it before I was fully awake. That drawing is from the perspective of a meter on an installed system, not being used with a powersupply for charging. So it looked wrong for what you are doing. I redrew it, almost exactly the same, but maybe easier to "see."

Important note. DO NOT trust the voltage reading from the battery meter for charging purposes. It might be as much as 200mv off, which makes a difference. To measure BATTERY voltage, use a DMM directly on the battery. To measure charger voltage, use a DMM directly on the power supply. These readings will sometimes not be the same. And never adjust the power supply when it is connected. Disconnect, adjust, then reconnect.

Thank you. Totally easy to understand. Is there a reason why I need this during “topping off”? Is it in case the power supply is not accurate?

Neither the battery meter or the meter on the power supply will be accurate. Both good enough for many things, but not "top balancing" which is what you are trying to do.

Update: first battery charged to 13.8 in just over 24 hrs. Second has been charging but stuck here(picture) or near here for most of the day. Will leave it overnight and see tomorrow. This was the battery that I had mistakenly connected pos to neg when hooking up two in parallel. Hope I didn’t fry it. The DMM reads 13.6

Sonnyboy said:

Update: first battery charged to 13.8 in just over 24 hrs. Second has been charging but stuck here(picture) or near here for most of the day. Will leave it overnight and see tomorrow. This was the battery that I had mistakenly connected pos to neg when hooking up two in parallel. Hope I didn’t fry it. The DMM reads 13.6

Click to expand...

Is the BMS in place?

sunshine_eggo said:

Is the BMS in place?

Click to expand...

No but I can put it on there. From what while we said, I thought using the DMM would be more accurate. I’ll hook it up.