Opinions sought - Are my batteries nearing end-of-life

[off-grid-geeks]

Thanks in advance for anyone who weighs in with their opinion...

Based on the following, do you think my batteries are nearing end-of-life?

Battery bank consists of 2 parallel strings of 4 six volt GC2 FLA 220AH energizer batteries (making a 24 volt 440AH bank), 5 years old, all the same age. Running a desulfator which I realize may be snake-oil.

A couple of weeks ago, in the middle of the night, my inverter started alerting a low voltage alarm when the fridge would kick on. This is on days which was sunny and a full manufacturer's absorb was reached and float was attained, WBjr showed 100% SOC. Just prior to the alarm, my nominal middle of the night load was drawing 2.7 amps from the bank. When the fridge kicked in, the load jumped to 9.7 amps (7 amp increase).

I know that voltage is not a great way to measure SOC, but since this issue started, the two readings are way off. For example, last night, when the alarm went off, the WBjr showed 85% SOC, and the battery bank showed 22.6V. According to my rough calculations, 85% SOC should be 25.89 volts. The battery voltage of 22.6 volts should be down to 30% SOC.

All appearances are that the battery bank voltage is dropping faster than the WBjr says it should be, based on amps out starting at 100% SOC. And the bank voltage is dropping faster than it did over the past months/years even though the loads have remained the same. There are no new loads on the system, and the WBjr is showing the same consistent loads over the past months/years.

Prior to a couple of weeks ago, we NEVER had low voltage alarms at 85% SOC.

[ClassicCrazy]

Take the voltages of each battery separately and see what you get. Could be that one or more cell in one battery is bad.
Since you have two strings if just some batteries have failed at least you may be able to use the best batteries in one string to run your system to get you by .

There is some more info that would be good to know.
How deep did these batteries usually get cycled down ?
Have the batteries always been temperature compensated ?
Have the batteries been equalized regularly ?
And how much distilled  water have you usually needed to add to keep them topped off ?

Problem with having two parallel strings of batteries is that it can be tough for them to be balanced in charging and discharging and lead to not as long of life span.

So depending on how hey have been used , what the setpoints were,  how deep discharged, and maintenance  yeah in five years they may be all done doing their job. My bet is just some of the cells have gone bad - so let us know what the voltages are of each battery. And of course since they are flooded batteries you can charge them up and then get the specific gravity of each cell to find out which ones might be sulfated up and gone bad.

Larry

[Westbranch]

If you do not have a clamp Amp meter , I would look at the individual battery voltages while or just after the alarm goes off. You might be able to pinpoint the weak one.
Have you removed, cleaned all cables and torqued all connections from Batt. to inverter?
Check the voltage at the inverter and compare to that obtained at the Batt buss bars.

Have you checked the SG of all cells? Any loafer cells?

[Ron Swanson]

Seat of the pants says that is a very small battery bank for somebody with an electric fridge and we typically see only 4 years out of any brand of L16 so my money is that they are almost definitely "plumb wore out." 

Especially when you consider that they have probably been cycled deeper more repeatedly than a larger bank would have, and the smaller the battery the more "consumer oriented" the quality seems to get if you know what I mean.

[krementz]

It sounds like you have been regularly getting low voltage alarms, just not with such a high SOC reading.

If you are  getting alarms (regardless of calculated SOC readings), then you have probably got your SOC below 20% often, and you have worn out (abused) your batteries.

You probably need a larger PV array and new batteries.  $orry.

[dbcollen]

we typically see only 4 years out of any brand of L16. 

I have consistently gotten between 5 and 9 years out of L-16s in my customers systems, 7 yrs being typical. I have used Rolls-Surrette, Deka, Interstate and Centennial, all with similar lifetimes. If you only get 4 yrs, then you are not sizing properly or charge parameters need tweeking.

[Ron Swanson]

I have consistently gotten between 5 and 9 years out of L-16s in my customers systems, 7 yrs being typical. I have used Rolls-Surrette, Deka, Interstate and Centennial, all with similar lifetimes. If you only get 4 yrs, then you are not sizing properly or charge parameters need tweeking.[/quote]

None of those brands are common around here for whatever reason.  It always amazes me how regional off-grid stuff is.  Here it is US Battery or Trojan mostly.

Mostly they are all customer supplied/installed.  We just get called to clean up the mess when it all goes sideways. 

We don't find any reason to sell L16s.  Industrial/fork lift type batteries are the same cost.  And they seem a whole lot more robust.  And they can be gotten in much larger capacities while still a single string.

Also, in the past, they seemed to go longer.  But the 4 years is what we see here and now.  Just that there does not seem to be any reason, to us, to use L16s of any type.

[off-grid-geeks]

Thanks everyone for all the input.
I'll start by answering the questions that were posed.

Oh, I wanted to make one correction... My wife pointed out that these batteries have been running for 6 years, not 5.

1) Batteries never taken below 80% SOC, typical start of day for the bank is 85%

2) The electric fridge is a repurposed high efficiency small chest freezer, and has never taxed the PV system, the 440AH bank has served it well untill this past week.

3) These low voltage alarms are new to this situation over the past week. We DO NOT regularly get LV alarms, never had one until this past week.

4) All cables and connections are clean and tight.

5) Yes, the bank is temperature compensated, always has been.

6) Battery voltages match within 0.1 volts at the battery, buss bar, and inverter.

7) Last equalized 2 weeks ago.

Last watering of the 8 batteries (24 cells) took a tad over 1 gallon of distilled water. (hmm, this forum software turned my point '8' into an emoticon with sunglasses!)

Now for my findings based on your suggestions.

1) I disconnected the strings from the system and each other, and checked the voltage of each battery after the low voltage alarm went off around 4AM today. String 1 each of the 4 batteries was at 5.9 volts. String 2 battery 1 and 2 and 4 were each at 6.2 volts. Battery 3 though was only at 5.2 volts. Oh oh.

2) Checked the SG of each cell while still in the alarm stage, did not recharge first (that will be another test soon). And on battery 3 of string 2, the one at 5.2 volts, cell 1 and 2 were in the same range as the cells on all the other batteries, but cell 3 was at zero, off the bottom of the scale on my hydrometer.

So for the time being I have left string 2 disconnected until I can replace all the batteries. This is one of those times when the prohibition of mixing old and new batteries hurts. One dead cell requires the complete replacement of all the batteries in the bank.

That should conclude this thread, again thanks for all the input! It sure is appreciated!

-Alan

[Westbranch]

Yup the SG' TELLS THE REST OF THE STORY'.  that weak cell is pulling your whole system down gradually by causing an overcharge to be applied to the whole bank.....

While you have one string you might want to monitor the V and SGs and see if they respond to a more even charge, in comparison to the 2 string results.

[mike90045]

Although you seem to have located the problem, another thing could have been a dieing compressor in the fridge, running too often

[off-grid-geeks]

Eight new batteries installed, almost identical specifications from the old ones, still Sam's Club GC2, but they are now Duracell (Deka) and no longer Energizer (Johnson Control). [430AH @ 24v]

Prior to the problem that started all this, our system would typically start the day at 85% SOC. This morning the day started at 94%. I'm pleased. No issues of any kind with the fridge system.

[dbcollen]

6 years service from a set of GC2s is impressive, the longest running set I have seen was just under 5 years.

[ClassicCrazy]

Eight new batteries installed, almost identical specifications from the old ones, still Sam's Club GC2, but they are now Duracell (Deka) and no longer Energizer (Johnson Control). [430AH @ 24v]

Prior to the problem that started all this, our system would typically start the day at 85% SOC. This morning the day started at 94%. I'm pleased. No issues of any kind with the fridge system.

I got my first set of golf cart type batteries as a Sams Club - they used to sell them for a lot less than you could get them for anywhere else. And they performed well for me.  East Penn Deka make good batteries too.
Here are their charging specs - solar are the same batteries with different label
http://www.eastpennmanufacturing.com/wp-content/uploads/Renewable-Energy-Charging-Parameters-1913.pdf

Larry