Strange Battery Issue

Horsefly · June 2020

I just got back from a few days at our off-grid cabin, and in the last 36 hours of our visit I saw some things that confuse and worry me.

Background: I've got a 24V battery bank made up of a 2s2p array of 12V 155AH AGM batteries. It was installed in 2017 and we haven't noticed any degradation. I log the voltages of the bank every hour around the clock, and in the 3 years we've had them we have only gotten to 50% SoC one time.

One afternoon I was cleaning up some things, and had the PV array disconnected from the charge controller for some time. I then flipped the breaker to reconnect and forced the charge state to bulk. Once the bulk charge finished and transitioned to absorption (at 29.4V), I thought I could hear a faint hiss / squeal from one of the batteries. Since they are AGMs, any out-gassing is a pretty bad thing. I then measured the voltages on all 4 batteries. The suspect battery was at 16V, and the second battery in that string was only at 13.4V. Not good. The other string was more balanced, with one battery at about 14.6V and the other at 14.8V.

I got out my clamp meter and it appeared that the charging current was almost all going into the "good" string. This all seemed weird. I decided that I needed to swap two batteries between strings. I shut everything down, swapped the 14.8V battery with the 13.4V battery and fired things back up. By the time I got back to check on things the batteries were in float and all the voltages looked pretty balanced.

The next day (yesterday) I decided to do a test of the current under load. I shut off the PV input so that all I would see was current out of the batteries. I fired up our microwave, and measured the current at several points: Both lines to the inverter, and both lines between the strings at the ends and middles of the strings. The microwave caused a draw of about 43A continuous out of the battery bank. Here's the shocker, if I am to believe what I measured: Just over 40A came from one string of batteries, but something between 2A and 3A came from the string with the battery I am worried about. Whaaaa?!

To try and force a better balance between the strings, I inserted a temporary cross-strap between the mid-points. I asked about cross-strapping in an earlier thread, and it seems like there are mixed feelings. I had to leave before I could actually see any results. I plan to (probably) put in a more permanent cross-strap.

I plan to go back in a week or two and do some more testing. For one thing, I want to see what happens if I separate the two strings, having only one hooked up to the system for a day or so, then switching to the other string. If there is really a problem with one battery, I'm guessing I'd be able to see it quickly.

I respect folks here for their expertise. Are there other things I should be looking at?

BB. · June 2020

The string that supplied all of the current to the load--Probably the good batteries.

The ones that supplied little... You either have an open (or near open) cell in one (high voltage), or a shorted cell in the other (low voltage).

Ideally, if you can load test (or charge individually), one battery at a time--And see what happens. It sounds like the bad battery has an open cell, so it will not take charge, or supply load current very well.

-Bill

Horsefly · June 2020

Thanks @BB. That's my fear. If the suspect battery has gone bad, is there a high likelihood that the other battery in that string is bad?

It's about a 6-6.5 hour drive from my home to the cabin, which makes it hard for me to pop in and do some more tests. This time I had my wife and daughter who both appreciate the electricity that the solar provides them by really dislike when I shut it down. I'm going to try to go over solo soon after fathers day to do some tests. I was thinking of trying to test the entire system on each of the two individual strings. I would assume if I have a bad battery neither charging or discharging would go as expected / desired. I don't really have a way to test individual batteries, unless I buy something to do it.

The funny thing is, I can't see any difference in how the system is performing. Since I've never bothered to check to see how the current is going between the two strings, I am thinking that there is a chance that the one battery has never been working right. Is that possible?

Additional item that may be of interest: I was able to completely shut the system down mid-day and disconnect all the batteries for a couple of hours. All four batteries got to a very similar resting voltage. I would have liked to let them rest longer, but the natives were restless. Anyway, I'm not sure 2 hours is long enough to tell anything.

BB. · June 2020

It is hard to say... If the shorted cell battery cause the other in the string to "overcharge"--That is not good for the "not bad" battery in the string.

If you have an open cell, then the other battery in that string was not getting charged (or not charged much)--So it could have sulfated.

A shorted cell can also discharge the parallel connected string(s) too...

Given that the other string seems to be working well--Then my first guess is that the battery with the high voltage (during charging) sounds like the one with the failing open cell... But that is just a guess on my part.

-Bill

mike95490 · June 2020

I would not use the cross strap and would plan to replace a battery in the low current string. ( the one venting )
Since the batteries are vintage 2017, they are approaching expected cyclic usage end of life and you should prepare to replace all of them.
AGM batteries cost more, and generally have shorter lifetimes, than flooded.

Horsefly · June 2020

Thanks @mike95490. I'm going back up there before the end of the month. I intend to see what happens if I run the entire system on one string (the "good" one) and then on the other string (the "bad" one). Mostly to confirm what I suspect going on. Then I'll have some decisions to make.

I knew these were cheap batteries when we started. I kinda hoped they would last a bit longer than this, but in my worry world I knew I was rolling the dice. I don't know if this was caused by something we did or bad manufacturing. Based only on logging the voltages, the batteries have not been abused. As I mentioned earlier, it seems like the battery voltage (the only thing I have for an SoC) goes up and down about the same every day just like it did when the batteries were new. It seems like if we suddenly lost half of the capacity, we would notice. So it may be that one battery has been bad from the start. I just wish I checked the current between strings under load back when we started.

If I'm going to replace them, I have to make a choice. I think we would prefer LiFePo4, but it would take some extra engineering. Besides a BMS, I think I'd need to come up with some method to heat the bank before attempting to charge it when we have sub-freezing temps. At ~9,000 ft elevation in Colorado, freezing temps when the charge controller first fires up in the morning are common for the ~6 months that no one is there.

If I can't go LiFePo4, I'll be sticking with AGMs but will go with a single string of 4x 6V batteries from a better company (maybe LifeLine or Full River?).

If I go AGM, I'd need something between 300Ah and 350Ah (at 24V). If I go LiFePo4 I obviously don't need the 50% SoC cushion, but I think I'd still want to shoot for between 180Ah and 200Ah.

At this point I haven't shopped at all since we first installed the system. I'm sure I'll get sticker shock once I do shop.

Horsefly · June 2020

Just an update for anyone who is interested.

For review: My battery bank is a 2x2 array of 4 12V 155Ah batteries. That is, 2 parallel strings of two batteries in series. The two strings are connected on the positive and negative ends with a short (13") 4/0 cable, and longer 4/0 cables to the E-panel for the inverter and charge controller. Earlier this month I was using clamp meter to check the currents while running a microwave oven. With the microwave running the draw was between 40A and 44A according to the system control panel, and I could confirm that with the clamp meter on either the positive lead or negative lead to the inverter. However, when looking at the leads within and between the strings, almost all the current was coming from one of the two strings.

This past weekend I was able to go back up and (without wife and daughter) was able to shut things down and do some experiments to see how each batter was. I ran the system with only one string of two batteries, no PV input, and tried different combinations to see how each battery was performing.

I did discover that one battery seems to be failing. The first time I ran the microwave on a string which contained the bad battery the voltage collapsed only a few seconds into the run, and the voltage on that one battery was only 10V as the inverter shut down for LBCO.

During all the other combinations I did find that the other three batteries seem to be fine. I could run the the microwave on high (44A) for four minutes with no problem using any combination of the other three batteries.

I have a smart charger (Noco Genius10) that has a "repair" function. I will admit that I'm not sure what this function does, although it is supposed to help reduce sulfation and stratification. I tried it on the failed / failing battery. It ran for about 6 hours, and it seemed like it was pulsing a relatively high voltage (maybe 16V) for some amount of time, then settled back down to something like 14.7V. After some time, it would go back to pulsing a higher voltage. After 6 hours the charger declared victory and went into standby mode.

After that first repair run, I tried the microwave experiment again with the failed battery, and this time the inverter did not shut down until about 90 seconds, as opposed to less than 10 seconds or so the first time. Based on the improvement, I did another repair cycle on the battery before I left to head back to Denver. I didn't have time to test it again.

So my questions:

Does this repair thing work?
Is it at all possible that I can do repeated repair runs and get this battery back to a usable state?

I was at the cabin Friday, Saturday, and Sunday night. All three nights I ran the system with only one string of two batteries (different combinations of the three good batteries), so only half of my designed Ah capacity. Although it is hard to say for sure, it seemed to me like the performance of the batteries was the same as when I had all four batteries in the bank: I was able to run the well pump and microwave after most of the sunlight was gone, and the next morning the bank was still registering between 25.2V and 25.5V. So, I still think this battery that I now know to be bad may have been bad for quite a while.

I've recommended to my siblings / partners that we should be fine on just two batteries for the rest of this year, and will have the winter and early spring to decide about replacement batteries.

Sorry for the treatise, but I look forward to hearing what people think.

mike95490 · July 2020

The "repair" seemed to add a few seconds before the shutdown, I've no idea if it is a Equalize repair or a Desulfate repair. It could be simply one dud cell in the battery and this could be a time to try a single battery replacement and see if it works OK, or if the rest of the batteries are going out next week.

mcgivor · July 2020

Being AGM batteries it's not possible to determine which cell is affected if that is the case, applying a higher repair voltage will force the ballance of cells to overcharge, causing venting before the low cell has time to be corrected. With FLA monoblock batteries it is possible to perform a recovery, which I have done successfully using the water treatment process, however it was extremely time consuming and unfortunately can't be done AGM.

Perhaps running the single string is the best option whilst considering a replacement, rationale being the ballance of cells in the bad string may have been stressed by over voltage if in fact there is a bad cell. Such problems are common with LA in general, especially in the beginning when knowledge is limited, resulting life expectancy being less than expectations, it is unfortunately the nature of the game.

My LA bank is now on borrowed time, I too had to recently remove one string from a parrallel arrangement to extend the lifespan of the good one, four years isn't bad considering the higher temperatures here, most automotive batteries last three years, give or take.

BB. · July 2020

And leaving a "crippled" battery/cell in an operational string can cause other problems too... Depending on the failure, a shorted cell can discharge the rest of the good battery strings. And, again, depending on the fault, it is possible to overheat and even cause a fire/hydrogen explosion in the bad cell/battery when cycling.

Better to just remove the bad battery/string from service. You can keep any of the "good batteries" left in the string charged and ready to replace the next failing battery in the "good string". For AGM, they can go 6 months between recharging (if battery is good), and cold ambient temperatures slow down self discharge even more--Helping to extend the life of the batteries in storage. Just toss the stored batteries on a 12 volt charger 24 hours every ~6 months.

-Bill

Dave Angelini · July 2020

You have got to know when to hold them, and when to fold them. Kenny Rogers

Horsefly · July 2020

So my brother took his family up to the cabin yesterday morning. He had his two portable refrigerators / coolers plugged in all night, and four adults doing whatever they needed, all operating on just the two good batteries that are still connected to the system. He got up this morning and reported that the bank was still at 25.7V. I think that is more evidence that the bad battery was bad at least for the past year, maybe two.

Yes, new batteries for next season. Seems like we can limp along the rest of this year with just half of the designed energy storage.

Dave Angelini · July 2020

Hard not to do well in summer. Good Luck!

softdown · July 2020

If it was me I would replace the bad battery. I've replaced five bad batteries in the past five years. It isn't like your world will collapse if it fails to deliver another 2-3 years from the rest. I don't think these batteries are heavily worked and they are likely cool on average at that altitude.

That's just me of course.

I think if we understood battery manufacturing better we would understand how easy it is for one cell to go kaput.

Oddly I had to replace as many fairly young batteries as old batteries. Pretty sure that line of batteries suffered a collapse in quality control around 2013. Sam's Club wound up ditching them as a result. Also - they were never good solar batteries. Long story that.

Marc Kurth · July 2020

softdown said:

If it was me I would replace the bad battery. I've replaced five bad batteries in the past five years. It isn't like your world will collapse if it fails to deliver another 2-3 years from the rest. I don't think these batteries are heavily worked and they are likely cool on average at that altitude.
That's just me of course.
I think if we understood battery manufacturing better we would understand how easy it is for one cell to go kaput.
Oddly I had to replace as many fairly young batteries as old batteries. Pretty sure that line of batteries suffered a collapse in quality control around 2013. Sam's Club wound up ditching them as a result. Also - they were never good solar batteries. Long story that.

As a distributor or AGM batteries, I may have a different perspective than other people. I sell between 2,500 and 3,000 AGM batteries per year and "bad batteries" are exceedingly rare in my world. Battery abuse/neglect runs rampant due to a lack of knowledge. I would estimate that I see around one actual factory defect per year.

Much has to do with the quality of the product line. At the risk of being repetitive: All AGMs are not the same.

Horsefly · July 2020

I'm pretty sure that it was @Marc Kurth who recommended 3 years ago a voltage logger to keep an eye on batteries. Based on that, I bought one for our system when we first installed it: The Lascar EL-USB-3 Voltage Data Logger. (https://www.amazon.com/gp/product/B003DS0BS2). I've logged the voltage of the battery bank every hour now for three years. I think I've even posted some of the graphs here. I was especially worried about what would happen when I wasn't there to watch it. I suspected that the hardest period would be during hunting season, when there is less sun and 5-6 people (most of whom don't have a clue about solar or energy conservation). What I've seen is that there was one time - during hunting season - that the voltage dropped below 24V to 23.7V. So I think the DoD should have been OK.

On the charging side: Even though the Schneider MPPT60-150 defaults to a 2%C absorption end amps and no way to change it, I programmed the Ah rating to be half of what it was, so the end amps would in fact be 1%C. My system is over-panelled, so the only time the batteries didn't get back to float was on the days that snow covered the panels. That has been a total of 4 days over 3 winters.

Anyway, I'm not sure what I / we could have done to wreck this battery. We've managed both the discharge and the charging pretty conservatively. I know @Marc Kurth isn't specifically accusing me with the abuse/neglect comment, and I don't think I'm being sensitive. I just can't figure out what I could have done to prevent this, or what anyone did to cause it.

Dave Angelini · July 2020

I think Marc said it was a 2nd tier battery and I agree. For offgrid I only use the best and never any strings of batteries.

In my mind for offgrid, AGM's are not easy batteries to own offgrid. You did well Steve on what you have. I would have specced flooded because they are just easier to tweek with SG. Flooded batteries have been successful in cold places at altitude before there were AGM's.

Happy 4th ! Not sure President Washington would be very happy about 2020.

Horsefly · July 2020

Thanks @Dave Angelini, I don't disagree with most of what you said. Flooded would have never worked for us. There are several people / families who are up there way more than me, and I would be the only one to tend to the maintenance of the batteries. In non-pandemic years, the wife and I travel elsewhere in the world too much to get up there more than 2-3 times per year. Also, my brother-in-law and sister have a 42 ft sailboat with solar on board, and so his head was already tied to AGM over flooded all the way.

If I can come up with some ways to do thermal regulation, I'm leaning towards a LiFePO4 solution. Time to put on my engineering hat again!

Dave Angelini · July 2020

Well I would just say that before there were AGM's people could make that work (with engineering) so let's just leave it there. Happy 4th!

Horsefly · July 2020

Thanks @softdown, but I don't want to insert a completely different battery into this mess. We'll be fine on the single string for the rest of this year, and by next May will should have saved up enough pennies to buy a new set of whatever we decide on.

It occurs to me I should probably get some recycle money for these batteries when I take them out, but I have no idea how / where to do that. I guess I have multiple types of research to do now.

mike95490 · July 2020

Hold the batteries till you get the replacements, and use them for the core charge.

Check the price difference for the $5 payment you get for recycling a battery for the $11 core charge if you don't bring the dud in when you buy the new ones.

mcgivor · July 2020

The battery dealer here in Thailand offered $15 for a 12V 130 Ah flooded battery, the recycling depot paid $37 by weight.

softdown · July 2020

mcgivor said:

The battery dealer here in Thailand offered $15 for a 12V 130 Ah flooded battery, the recycling depot paid $37 by weight.

Sounds like lead is paying wel lright now where you are. Really well.

Marc Kurth · July 2020

We are down to $0.33/lb. so our 150 AH is worth $32.00 scrap lead price here in my little corner of Texas.

Dave Angelini · July 2020

Have you ever used one for a boat anchor? You guessed right, I have on a windy night in Costa Rica. Be well Marc!

softdown · July 2020

.33 sounds pretty good. I was quoted .22 last time I called.

Marc Kurth · July 2020

Dave Angelini said:

Have you ever used one for a boat anchor? You guessed right, I have on a windy night in Costa Rica. Be well Marc!

Bluewater guys are by nature resourceful and they do what they gotta do!

Thanks, Dave! We are alive and well thus far. We had a great time this last weekend - I'll post over in the Weeds section.

Hoping all is well with you and yours.

Dave Angelini · July 2020

Hunky Dory here. Loved the F35 pix on the lake.

Brother works on the Navy and Marine variants of F35 in hook-ups with the Navy drone. Should have put this in the right thread....

Strange Battery Issue

Comments

Categories