Batteries have begun to fail

[Wolvenar]

Well the last few days I noticed my reserve capacity in my batteries seemed to be very much reduced..
So today I got out the hydrometer, took a bank at a time offline, let them stabilize and started testing.

End result is I had four pretty much dead batteries, one in each bank.
One battery was totally shot, each of the three cells seem shorted. 2 of them had a dead cell each, with the remainder having two bad cells.
 
So after a number of tests per battery I rearranged them all to best match one another for each of the 3 banks of four I have left. Hope they hold out well enough for me for a while yet, until I can get some more.

Out of the batteries I seem of have one bank that is (best estimates) at least 98% of rated capacity
And the other two are about 95%

But having the one battery in each series bank having problems, none worked right.

[Wolvenar]

For my future reference * or any other it may be useful for.


Found here
http://www.solarseller.com/battery_state_of_charge_charts.htm

Does this seem correct to everyone ( before anyone that does not know any better like me relies on it)

[Wolvenar]

% Charged	Specific Gravity
100%	1.255  1.275
75%	1.215  1.235
50%	1.180  1.200
25%	1.155 - 1.165
0%	1.110-1.130

Comparison of Battery State-of-Chart to Open Circuit Voltage for 12-Volt Batteries
% Charge	FLA	Gel	AGM
100	12.70-12.60	12.95-12.85	12.90-12.80
75	12.40	12.65	12.60
50	12.20	12.35	12.30
25	12.00	12.00	12.00
0	11.80	11.80	11.80
Divide values in by 2 for 6v, mulitply by 2 for 24v, 4 for 48v

[Volvo farmer]

What is the configuration of your battery bank?

I am in the same boat with my seven-year-old L16s  I lost a cell, found a replacement used battery for it, and immediately lost another cell in the other string.  Have been running one string with a dead cell for several weeks now in parallel with the string with no bad cells.  So far, the lights have not gone off and the inverter hasn't disconnected on LVD. Been using a LOT of water though.

It's kinda like running down the road on seven cylinders.  It's not very pretty or efficient, but at least you're not sitting motionless on the side of the road. 

I have almost saved up enough money for a set of HuP Solar ones. I think I can limp through summer like this but no way am I going to get through winter with this battery bank.

[ghurd]

For my future reference * or any other it may be useful for.

<chart>

Does this seem correct to everyone ( before anyone that does not know any better like me relies on it)

No.
11.9V is usually considered "dead", like the text you posted 1 below it.
12.4V is usually considered 50%, which the text says, but that is a lot different than what the chart says.
I don't think I like that chart.
G-

[Isaiah]

If you can maintain your required bank voltage remove the bad battery and work on it by its self.
 Sometimes a transplant dosent adapt well with the others in the bank.
Isaiah

[ChrisOlson]

11.9V is usually considered "dead", like the text you posted 1 below it.
12.4V is usually considered 50%, which the text says, but that is a lot different than what the chart says.
I don't think I like that chart.

I disagree.  NOBODY that lives off-grid can ever rest their battery bank long enough to get an accurate open cell reading to determine SOC.  As far as I'm concerned Rolls-Surrette is the only manufacturer that gives specific instructions on how to determine 50% SOC based on voltage with the bank under load, and how to use that information to properly set up your inverter/charger to recover the bank when it reaches true 50% SOC.
http://www.rollsbattery.com/content/motive-power-bulletins?q=node/51&phpMyAdmin=0610e516bf803196b5feee0b1ad65c08&phpMyAdmin=3jSJ-jdC5E7b53DHgV8TGvpSCF6

It's not unusual to see our Surrette bank (twenty four T12-250's) at 23.5 volts under load, and the SOC is still well above 50%.  On consecutive days when we've had poor power production I've seen our bank below 24.0 volts for two days, and it hangs right there between 23.5-24.0 volts delivering ~1.5 kW of power, average, 24 hours a day.  Our master inverter brings the generator online if the voltage drops below 23.2 volts for more than two hours, which is 11.6 volts for a 12 volt battery at its rated 100 hour discharge rate (3 amps for our batteries @ 850 watt load on the inverters), and that's right at 50% SOC.
--
Chris

[Wolvenar]

I was able to take all but one bank off for about six hours, then test each cell with a hydrometer.

By taking the initial voltage of each battery, drawing each down with a load for 10 min noting the voltage under load, and retesting the voltage then taking another test after sitting for 10 min unloaded. At the same time I also took new cell hydrometer tests.
I then sort of graded each and matched them as best as I could per series of four.
I traded in the best set I could for the live bank, and repeated.

Not sure if this is the best way to do this but it seems to have worked well as can be hoped.
I still don't have good numbers to give but it still seems that the banks are less than they should be.
Or maybe this is just my way to convince myself I need more/new batteries?

[madlabs]

I agree with Chris. I see the same sort of voltages, scaled down to my 12V system. The three ways I keep track of my SOC are watt hours out, S.G., and how long it takes for the batteries take to reach 14.4V when charging and the current starts dropping. The last is the one I use the most on a day to day basis.

Jonathan

[Volvo farmer]

One the one hand, I agree with Chris, that being off grid makes it hard to rest a battery long enough to measure it, but on the other hand, when you have TWELVE parallel strings of 12V batteries in a 24V system, it would seem easier to rest a string or three than for folks running a single string of 12 ginormous 2V cells.

[rossw]

One the one hand, I agree with Chris, that being off grid makes it hard to rest a battery long enough to measure it, but on the other hand, when you have TWELVE parallel strings of 12V batteries in a 24V system, it would seem easier to rest a string or three than for folks running a single string of 12 ginormous 2V cells.

I have 3 strings of 2V/500AH AGM in a 48V nominal configuration.
I can take a bank offline for work - but my greatest concern with 8-12 hours offline is that it's very hard to find a time where you could put the isolated bank BACK online and have it close to the same SOC as the rest of the bank.

Take it offline at 6am when I get up - the bank is about as discharged as it gets, then put it back on at 6pm when it's fully charged and floating...

Or take it off at 6pm when fully charged and put it back on at 6am when the rest of the bank is discharged.

Midday/midnight, isn't ideal, but probably getting "closer" to  being around the same SoC both ways.

[Wolvenar]

My setup is not live in the way people fully offgrid would be.
The only reason I could not take the entire thing down is this website still runs on it :-)

I have 4 banks of 4 (6 volt) golf car style batteries that are about 9 years old now

As far as matching charge states, all banks were full charge by 9am, the draw down only took the batteries about 10% with exception of the dead batteries. 
SOC diff was not to bad.. no huge amperage spikes across the banks when reconnected.

[ChrisOlson]

I can take a bank offline for work - but my greatest concern with 8-12 hours offline is that it's very hard to find a time where you could put the isolated bank BACK online and have it close to the same SOC as the rest of the bank.

Same here.  I have my batteries wired like this:
[ Guests cannot view attachments ]

I have six banks of four batteries like this, with each bank having a positive and negative feed to the bus.  Theoretically I could take any bank offline.  Or even take four banks offline and leave one bank on each inverter while I do whatever with the four offline banks of batteries.  But they get too far out of balance way too fast when you have 1-1.5 kW load on your bank 24 hours a day.

When we first got the new bank I tried them in series pairs and they got out of whack within one week due to slight differences between batteries.  I asked the dealer where I got them what to do and he told me to wire them up this way.  It took a lot of cables to wire my bank up this way, but it keeps all 24 of my batteries PERFECTLY balanced.  See here for the last time I serviced them:
http://www.anotherpower.com/board/index.php/topic,443.0.html

It might be just me, but I seem to read WAY too many sob stories about finding dead cells in series strings, and batteries in series strings that have gone tits up before their time.  My whole bank of 24 batteries only has 6 series connections in it, and I firmly believe in keep the number of series connections between different batteries to the bare minimum.
--
Chris

[Wolvenar]

My 9 year old batteries are all identical, and were all bought within a couple weeks of each other. There are a series of 4 (6 volt) to make the 24 volt of my system, then each of these banks are parallel via wire to a terminal with like length wires. These then connected to a the main "bus" wires that go to my disconnects/fuses.
 There are 16 batteries total batteries so nothing about it should be abnormal or *possible* to create the correct voltage.

@ Chris;
 How could this be significantly different than what I see (partly) represented in your image?

[ChrisOlson]

You got three series connections in each string, plus four strings.  So you got 12 series connections - double the amount of series connections I got.

I could reduce my series connections to three.  But my inverters need 720 amps at full surge load.  So I'd have to use a lot bigger wires from the bank to the bus if I cut my series connections to three.

At full rated load, my inverters pull 360 amps.  Each sub-bank of four batteries has to deliver 60 amps to the bus, which requires 30 amps from each battery.  Each one of my batteries, from 100% SOC, can deliver the 30 amps for 2 hours and still be at 60% SOC.  For us this is important because we have electric clothes dryer and range (240 volt).  If the bank couldn't deliver the required kW to the inverters like when my wife puts a load of clothes in the dryer, the generator would start on peak load management all the time.  We have plenty of battery capacity to run it, but it takes the six parallel feeds to the bus to keep the voltage up at the bus with the inverters at full load.

So in a nutshell, our setup is way different than yours because I have designed ours to run 8 kW loads with no incoming power.

Edit:
Wolv, I should clarify that our batteries are 12 volt, not 6 volt.  Let's say you had 24 batteries.  So you'd have six series strings.  At 360 amp load each string would have to deliver 60 amps just like mine do.  Except with the six volts and three series connections to make 24 volts, each of your batteries has to deliver 60 amps.  With the 12 volt batteries each one has to deliver 30 amps.

You would need 400 ah 6 volt batteries to have the same kW capacity as our 200 ah 12 volt batteries.
--
Chris