Why Has My 6-Month-Old "Once Perfect" Solar System Failed?

rickst29

Solar Addict
You should review this setting. The charger puts current into the battery until the boost charge voltage 14.6 is reached. It then holds the voltage at this level for the boost duration. That is holding the perhaps fully charged battery at 14.6 volts for 3 hours. This seems a poor method of charging a lithium battery where a fractional C charging rate is used.
Victron for their lithium batteries recommend 14.2 charge voltage, no or a maximum 30 min absorption time, (the boost duration in Epever speak), and a float of 13.5 volts.

Mike
I set 14.6V because it also works slightly faster for charging from a gas-guzzling Generator. In my experience with charging my own LFP battery bank, the batteries need about 40 minutes of additional charging, after 14.6V charging voltage is reached, before current drops to roughly zero amps. My MPPT charger can only do about 30A, while my test charger does a bit more than 60A - so I conclude that it needs about 80 minutes from the MPPT. I can certainly drop from BattleBorn's 180-minute value to perhaps 80-90 minutes maximum (and still finish the job). But charging at only .17 C maximum current, I need more than 30 minutes to reach 100% SOC.

I wonder about Victron's advice. Is 14.2 Volts really adequate, when most BMS boards are willing to individual cell charging to reach 3.65 Volts? It seems to me that if one or more "happy" cells wants to absorb up to nearly 14.65, a single "recalcitrant" cell might be left significnatly lower than that. (The average of all cells would be 14.55, but some could be higher, while another is low). If BMS was reprogrammed to terminate cell balancing at 3.55 per cell, they would all get pulled up to that (by a 14.2 Volt total). But most BMS defaults to 3.65 on each cell.

Thanks Mike, you brought up great points and my "Boost cycle time" does appear to be too long (by at least a factor of 2). I'm not sure, however: Does it stay in boost mode for the whole time, even if "Float Voltage" has been exceeded?
 
Last edited:

mikefitz

Solar Addict
Most BMS start to balance at 3.4v per cell, 13.6 volts on a 12v battery. If your cells are already top balanced do you need to run up to 14.6 volts every charge cycle and activate the balance for long periods? If the cells are reasonable quality they should stay in balance, so why stress unduly each cycle.

My personal view is that the trend to push each cell to 3.65 each time you charge will reduce the service life of the battery.

I can understand why you want to reduce generator run time, perhaps accepting charging to 98% capacity rather than 100% would be a solution.

Mike
 

rickst29

Solar Addict
Most BMS start to balance at 3.4v per cell, 13.6 volts on a 12v battery. If your cells are already top balanced do you need to run up to 14.6 volts every charge cycle and activate the balance for long periods? If the cells are reasonable quality they should stay in balance, so why stress unduly each cycle.

My personal view is that the trend to push each cell to 3.65 each time you charge will reduce the service life of the battery.

I can understand why you want to reduce generator run time, perhaps accepting charging to 98% capacity rather than 100% would be a solution.

Mike
Thanks again, I LOVE this answer. I will drop my charging SCC charging Voltage, maybe to around 14.4V at the same time I shorten the 'Boost Duration'. (It's easy to change things on the "MT50" display/control unit I have connected into my Epever BN3215).
 

jesfl

Solar Enthusiast
Thanks.

So, there's no "typical" configuration for a 150-volt LiFePo4 battery? (I'm clear about the "standard" 100 Ah battery from numerous videos and articles.)

And, the only way to know what is inside my 150 Ah batteries is to cut one open? (Since I am able to get no information from anyone selling the batteries, and who knows who the Chinese manufacturer might be?)

I'm just not ready to start tearing apart batteries, yet. Truly last resort. But, probably edging closer.

Again, one more beginner question. Does every battery in a system have to be the same exact brand, size (Ah), etc.? I get they all would have to be 12-volt batteries. I have not yet searched for that on the Forum, but I will.

Again and again, many thanks you's to all.

Jim S.
 

jesfl

Solar Enthusiast
P.S. Forgot to mention. When I tried the three different series of values for setting the Tracer for lithium, I also tried 30, 90 and 180 as the Boost duration. Batteries still did not charge and differently . . . i.e., essentially none.

Jim S.
 

GXMnow

Photon Sorcerer
I went back to the original seller posting for the battery and looked at the specs. The dimensions are shown as 12.2 inches high. That is just a bit longer than 4 x 65 mm. This could be a stack of 18650 cells. Then it is 7.7 x 9.4 inches. If 18 mm diameter cells were packed against each other, that would hold 10 x 13 cells, but let's just call it 10x10 with the BMS stuck on the side. That would be 100 cells in parallel. They would only need to be 1.5 AH each and that is easy for LiFePO4 cells. They may be less fatter cells, who knows, but it is large enough to hold that capacity with standard small cells. If they used prismatic or pouch cells, it could be a bit smaller for that capacity.

The obvious first choice is to see if the seller will repair, exchange, or refund on them. They list a 30 day full refund, and a 1 year warranty. How long have you had them? But I am guessing the seller won't warranty the batteries. If that is the case, you don't have anything to lose. I would choose the one that measures no voltage, seems totally dead etc. Then carefully slit and peal off the outer black shrink wrap to see what all is inside. If you can peal down until you can see the buss bars, you should be able to measure if the cells have voltage on them. If it is just a BMS failure, you should have close to fully charged cells. Each cell group should measure about 3.4 volts, with the 4 in series, making about 13.6 before it goes to the BMS. If the voltage is low, it may be the BMS is drawing power. Hopefully all of the cell groups are still above 2.5 volts, 10 volts for the whole pack. As long as you have that, you can wire in a new BMS and be back up and running pretty easy. If you are getting no voltage on the buss bars, then it may have cell level fusing. This is very common and large parallel groups of small cells. They intentionally use a very short length of a thin wire or nickel strip to connect each cell to the buss bar. With 100 in cells in parallel, each one will never see more than 2 amps, so a thin wire will take normal conditions, but if something goes bad, they can pop pretty easy and protect the wiring from melting. My large battery bank fully shut down on me just because of a single bad connection at one of the balance wires. The problem could be that simple. But since all 3 of yours seem to have dies together, my guess is the charge controller raised the voltage too high and the BMS shut it down. With no load on the cells, it can take a long time for them to blead down enough for it to reset and turn back on.
 

jesfl

Solar Enthusiast
Wow! That is a tremendous help! You have truly gone above-and-beyond. THANK YOU ! ! !

(1) Because I've never seen any variety of lithium cells, I don't really understand your description of the types and sizes. But, I probably don't need to know that right now. I'll do a little research for different types of lithium cells later.

One of the new eBay sellers of the BtrPower battery responded to my inquiry a few minutes ago. I've been badgering the sellers asking questions, mostly no response. I asked about the cell construction of the battery. The response was: "The battery is composed of 4S150Ah lithium iron phosphate. The voltage of each string of chips is 3.2V-3.65V." That doesn't help, of course. I try again for a more thorough explanation.

A.so, I would not be surprised that these "facts" are incorrect for the 150Ah version of the BtrPower batteries, but sorta' standard response for all of the eBike batteries they sell? Who knows? I will ask again and try to pin down the gent. At least he responded. That's a start.

(2) The eBay seller from whom I bought is no longer to be found. I had two email addresses and have been sending messages to him for nearly two months. Zero response. He is history. eBay itself has been no help. Two of the new sellers on eBay haven't been very helpful either. It turns out the seller who responded this morning has eBay listings under 3 different seller names, probably at different price points, or something. Even if the one-year warranty was valid, none of the new/current sellers will honor it. I've asked in many different ways.

So no, no 1-year warranty. Bought in April 2020, delivered mid-May 2020, first put to use in late August 2020 when I got all the parts and finished the first install of my RV solar system. Many, many modifications/corrections have followed.

(3) You make it seem easy to open up one of the batteries, and it may be. So, I'm edging closer to doing that with one battery as you and others have suggested.

Opening up a battery is not overwhelming to me. But, I started poking around right after my first post on March 19 to find BMS sources and to watch videos on building lithium batteries. The specs of the BMS are where the tech-speak lost me. Will Prowse's video, of course, makes it look easy. It ain't ever that way for a beginner, I attest. Then I started reading about "(top?) balancing the cells" and I totally drowned in the tech-speak. When you have absolutely no basis of knowledge, every tech term requires an explanation, sometimes two or more explanations for layman clarity.

That latter point is why I sincerely appreciate your layman-like approach in trying to explain my mess to me, and to explain the options in a way I can understand (mostly).

(4) I've decided to move from my current boondocking spot to an RV park, probably Friday. I'll hook up to 30 amp shore power. That way I can use the Progressive Dynamics lithium charger I installed in the RV's WFCO power panel to charge 2 of the batteries 24/7 for a few days. It will be a good test. And, I'll have power for living, instead of, as now, no power but minimum LED lights after dark when sun-power goes away.

I may even begin to "skin" the weakest battery before that. FYI, the batteries are built with a heavy-duty rubber "shrink wrap" on the outside. There looks to be a fairly thick paper/cardboard material under that, and then on the sides, at least, there is some sort of 3" by 3" geodesic-like case/support grid (boxes with X inside them). I'm assuming that is plastic. At any rate, it doesn't seem that difficult to open up the one battery.

(5) Lastly, you wrote: "With no load on the cells, it can take a long time for them to bleed down enough for it to reset and turn back on."

Does that always occur with lithium-BMS-managed batteries? The "reset and turn back on" part? If the answer is yes, that begs a different approach as a test, at least to me? You have aroused my curiosity, again.

Is there is any way to have one battery isolated (the weakest of the 3) and connect it to a small "load" of some sort to speed that "bleed down enough to reset and turn back on" process you described? For instance, I have an old 700-watt inverter (from my sailing days). It's not pure sine wave, and I don't know if that matters. But, it works. So, could I just connect that small inverter directly to one battery and plug-in some low-load something to deplete the battery? If not the little inverter, is there something else safer I could connect for a slow discharge? Would I need inline fuses for safety? If so, do you have a suggestion for fuse amperage?

If there is a possibility that a low-low state would turn on the BMS, I'm up for trying that. My reservation is whether that will do any permanent damage to the cells if they go way down to a near-zero state? My understanding is that lithium can be discharged to zero, but I don't know if that is true?

Obviously, I'm still holding out a little hope for a small miracle before I start tearing things apart.

And, once again, a huge THANK YOU for the helpful response and layman-level explanation.

Jim S.
 

HRTKD

Boondocker
The time to "bleed down" a battery from full 100% charge to a resting voltage can vary. I would expect it to take no more than a few hours, call it eight hours. Probably less.

Top balancing is not possible until the cells have been removed from the battery. The idea is that they need to be reconfigured from the 4s arrangement to a 4p arrangement. Once you have them 4p, a bench top power supply can be used to charge the 4p pack to 3.65v.

You can certainly connect one or more of the batteries to a load and see what happens. Check the battery voltage before, during and after. If the battery voltage is less than 12.6v I wouldn't push the battery much lower. I base the 12.5v on the voltage chart found here:

 

jesfl

Solar Enthusiast
Again and again, thank you. You are immensely helpful. I admire and respect your thoughtfulness.

I will try to "drain" the weakest battery which yesterday tested 13.29 volts via my multimeter disconnected/isolated from everything.

However, the curious conundrum I face is that I have no idea what the "real voltage" is in each of the batteries?

That's because the same battery when reconnected to the system by itself shows 10.5 volts on the EPEVER Tracer controller remote screen.

No one here on the Forum has yet been able to explain it.

If the actual battery voltage is really 10.5 volts, it is well below your suggested bottom number already and, essentially dead.

When all three batteries are in parallel and connected to the system, the Tracer SCC remote meter reads 10.5 volts, too. That's even though I know each battery when disconnected and tested individually with my multimeter read 13.38V, 13.32V and 13.29V.

My point is that given the way the batteries acted when all three are connected to the system (after I charged them each for 4 days with the 120V 10A charger that came with them), I am betting on the 10.5 volts as reality. And, the BMS is shut down allowing no charge.

That's because the three together will run nothing in the system except a few LED lights on the 12-volt side. Trying to just turn on the inverter (no load other than it's own internal start-up and standby need) causes an alarm and immediate shutdown.

Looking at the LiFePO4 voltage charts in the thread you sent, I see that the voltage in the batteries, when tested with my multimeter, means somewhere in the 70% to 90% full range. Therefore, a 13.29V charge is plenty of power to run the inverter with no load except its own very low stand-by voltage requirement. But, no dice. So it must be a completely dead battery (10.5 volts charge)?

Again, no one has been able to tell me why?

P.S. I used 2 different multimeters. I am confident I am not somehow screwing up the simple test for DC voltage.

Jim S.
 

HRTKD

Boondocker
If your battery is at 13.29 volts, then it should be in good, but not great, shape. Maybe it's allowing a voltage reading, but simply not allowing any - or not much - current. But to have three batteries fail at the same time? That's too much coincidence.

Once you get the battery taken apart, there are quite a few things to try. For example, bypass the BMS and connect the battery straight to a load , like the inverter, that it currently can't handle. Or, bypass the BMS and connect a charger to the battery. Note that neither of these tests would be performed for very long. Just enough to verify that the BMS is or is not the problem. Before all that, you'll want to check the voltage of each cell. If one is way different then you've found your problem. "Way different" would be a voltage difference of more than one or two volts.
 

jesfl

Solar Enthusiast
Good evening,

I have just spent a few more hours reading posts here.

I think I have learned that every battery in a 12-volt bank of LiFePO4 batteries does not have to be the exact same battery or even the exact same Ah size?

Beginner here. I didn't know that. Is that true?

If yes, I then have a possible solution for my 3 "dead-and-won't charge" 150 Ah batteries (450 Ah total, once upon a time . . . ).

I could:

(1) Buy one new 100 Ah LiFePO4 battery and pedal along slowly for a month or two with it alone. It would be better than trying to get by with the single SLA I am now using.

I have also considered two other options for temporary use: (a) a couple more SLA batteries, $160 more wasted: OR a couple of 6-volt golf-cart batteries, $400? wasted. Either is just throwing more money away, to me.

(2) With enough battery to barely get by while boondocking, I could concentrate on cutting apart one of the old batteries and trying to rebuild the cells into a new battery with a new (better) BMS. Assuming the cells are OK, of course.

(3) If I can successfully rebuild one battery from the dregs of one of my won't charge units, then over a few months I can probably make a total of four 100 Ah batteries from the corpses of the 450 Ah of cells I have?

With decent BMS units installed, that would give me a total of 500 Ah of battery and put me in better shape for total available Ah than I ever planned or dreamed.

Does this make sense?

Will it actually work?

If yes, it surely provides a little light at the end of the dead battery tunnel for me.

Thanks for any and all thoughts and suggestions.

Jim S.
 
Top