Balancing aged AGM battery pack

[Nathan Barnard]

Looking for advise on matching and/or removing weaker batteries to maintain this battery pack.

I've been using it as is for the last two years. I has been performing well with daily discharges of 500-600ah @24v, it has stayed balanced up until recently. The lower side is dropping off much quicker than upper side. I regularly pull 250amps from it (for electric oven or clothes dryer) and charge around 100amps with solar.

Yesterday I charged both sides up to 14.2v and held them there overnight, then this morning I disassembled and tested each battery on a digital CCA load tester, marking down the cranking amps and internal resistance. I put it in a spreadsheet to try to figure out which batteries I should switch to balance average resistance and CA on pack (so they would discharge and recharge equally). I'm not sure if this is the best way to balance a pack for longevity, or even what a high resistance, low state of health battery does to a pack. Perhaps I should be eliminating those batteries altogether?

Specs:
26x Eaton Powerware 12v 390W VRLA AGM batteries, forming two strings in series of 13 batteries for 24v. (Appx 100ah 12v ea)

Ohms (Left 13 batteries)	CA	Ohms (Right 13 batteries)	CA
12.4​	300​	13.1​	280​
6.3​	585​	10.2​	365​
9.5​	395​	6.8​	550​
5.5​	675​	11.7​	315​
12.1​	305​	11.1​	335​
11.6​	320​	6.7​	560​
8.2​	455​	11.3​	330​
8.3​	450​	5.8​	645​
7​	530​	4.9​	750​
7.4​	505​	11​	340​
10.9​	340​	11.1​	335​
13.7​	270​	6.7​	555​
5.1​	735​	7.6​	490​
average Ohms 9.1	5865​	average Ohms 9.1	5850​
	TOTAL CA		TOTAL CA
	Lower 12v		Upper 12v

 

[sunshine_eggo]

As you have discovered, there's an inverse correlation between IR and CA. I assume you're reporting mΩ or mR values and not actual Ohms.

https://www.eaton.com/content/dam/eaton/products/backup-power-ups-surge-it-power-distribution/backup-power-ups/ups-batteries-resources/eaton-12v-390w-battery-brochure-BAT02FXA.pdf

Indicates 3mΩ internal resistance. Batteries pushing 2X or more this value may be ending their useful life.

I don't see the charge voltage? Is 14.2V appropriate?

I can't really tell from the picture. Is this a 13P2S bank? two groups of 13 batteries in parallel with the groups in series?

If so, you might shuffle them within each group moving the lowest IR to the highest IR positions.

In the left group, swap 12 with 13, 1 with 4, 2 with 5, etc

12​	13.7​				1​	13.1​
1​	12.4​				4​	11.7​
5​	12.1​				7​	11.3​
6​	11.6​				5​	11.1​
11​	10.9​				11​	11.1​
3​	9.5​				10​	11​
8​	8.3​				2​	10.2​
7​	8.2​				13​	7.6​	​
10​	7.4​				3​	6.8​
9​	7​				6​	6.7​
2​	6.3​				12​	6.7​
4​	5.5​				8​	5.8​
13​	5.1​				9​	4.9​

If those are aluminum bus bars, recommend you buff the crap out of them and apply a conductive corrosion inhibitor.

 

[Nathan Barnard]

Thanks for the reply. It must be mΩ! Yes, 2p13s. Yes, aluminium and covered with dielectric grease at the connection points. Finally i have never been able to find the recommended charging voltage other than float voltage.

Could you explain the benefit of sorting the batteries by resistance values? With the plates design they should all be equally discharged, or at least from outside in to center of string.

Finally i would like to understand the theory between removing older batteries vs leaving in back. I believe their resting voltages are similar and only that they discharge faster and are slower to recharge. (I used to recharge at 200-300 amps but now pack seems to only handle 100amps before hitting max voltage). Also if the weak batteries can't hold their voltage on discharge then by default it is the stronger batteries which are doing the heavy lifting. So it stands to reason removing old batteries vs leaving them in has similar effect of letting better batteries doing the work.

Maybe it is best to make sure both sides of battery have equal "good" batteries?


By the way, even just switching two batteries to make both sides have similar capacity and average resistance has been a big improvement overall. Now i can discharge pack further before it goes off ratio (voltage wise). I used to pull it down 600ah (of max 1300ah) but recently could only get it to 300ah. Swapping two batteries yielded 450ah at least last night.

Thanks for helping me understand how my batteries are working!

Nathan

 

[sunshine_eggo]

Thanks for the reply. It must be mΩ! Yes, 2p13s

13S would be a 156V battery.

You have a 13P2S, 24V battery bank.

Yes, aluminium and covered with dielectric grease at the connection points.

Dielectric grease is non-conductive. It should only be used to protect AROUND a connection and never part of it. There are many conductive corrosion prohibiting substances that can be placed between the contact faces.

I used this on my 28' of Aluminum bus bar:

https://www.amazon.com/NO-OX-ID-Special-Electrical-Contact-Grease/dp/B09239CNFN/ref=sr_1_3?keywords=NO-OX-ID+A-Special&qid=1678905644&sr=8-3

There are other, possibly better options out there.

Finally i have never been able to find the recommended charging voltage other than float voltage.

Are these intended for UPS operations rather than cyclic? It's pretty common to see only a float voltage in that application.

Could you explain the benefit of sorting the batteries by resistance values? With the plates design they should all be equally discharged, or at least from outside in to center of string.

The idea is that for whatever reason, the batteries are experiencing different conditions in different positions. Thus swapping the batteries to the different positions will place the healthier batteries in the higher stress locations. However, given your most recent reply, I don't recommend this as I didn't fully understand what you had done at the time.

Finally i would like to understand the theory between removing older batteries vs leaving in back. I believe their resting voltages are similar and only that they discharge faster and are slower to recharge. (I used to recharge at 200-300 amps but now pack seems to only handle 100amps before hitting max voltage). Also if the weak batteries can't hold their voltage on discharge then by default it is the stronger batteries which are doing the heavy lifting. So it stands to reason removing old batteries vs leaving them in has similar effect of letting better batteries doing the work.

As long as the older batteries will hold charge and not draw down the parallel batteries, they will add some benefit.

Maybe it is best to make sure both sides of battery have equal "good" batteries?

By the way, even just switching two batteries to make both sides have similar capacity and average resistance has been a big improvement overall. Now i can discharge pack further before it goes off ratio (voltage wise). I used to pull it down 600ah (of max 1300ah) but recently could only get it to 300ah. Swapping two batteries yielded 450ah at least last night.

Thanks for helping me understand how my batteries are working!

You have independently discovered that it is critical that each 12V portion of your 24V battery needs to be as identical as possible. It is also critical that when the 24V system is at peak voltage, the two 12V halves are at nearly identical voltages.

Lastly, I am completely unable to tell how your batteries are wired from the picture. Could you provide a top down sketch?

 

[Ryujin]

Dielectric grease is non-conductive. It should only be used to protect AROUND a connection and never part of it. There are many conductive corrosion prohibiting substances that can be placed between the contact faces.

Well now I feel like an idiot, especially after this triggered a little googling. Definitely didn't do my spark plugs right... But the car is running way better anyway.