Adding batteries to an existing battery bank

[DeepBass9]

I have been running 8x150Ah Trojans for almost a year and would like to add another 4 or 8 batteries (48V configuration). Now the folklore is that you shouldn't mix batteries of different ages. What does that actually mean practically? If I added another 4 would the new batteries be damaged? Or the older ones? I'm not sure how this all works.

[Chris Hobson]

I am on holiday so should not be straining my brain at all . Batteries lose capacity and their internal resistance changes as they age. Let's ignore the changes in internal resistances for a moment and deal with just changes in capacity. The batteries will be discharged to the same voltage for each string. The newer batteries will be discharged more to be at the same voltage. When you charge the batteries the newer batteries are going to rise in voltage slower than the older batteries. Thus the newer batteries will end up under charged and the older batteries over charged. This is exasperated by the difference in internal resistance of the new versus older strings that we ignored in the beginning.

Sent from my SM-G925F using Tapatalk

[DeepBass9]

Thanks for that. What sort of time periods are we talking about though for 'old' and 'new'? Is less than 1 year still new?

Also how feasible would planned replacement of parts of the battery bank be? So I now have 8 batteries, I buy 4 more now. Next year I buy 4 more and sell the first 4. Assuming they are still functioning I should be able to get 25-50% of the original purchase price. So I will have a battery bank with batteries always between 0 and 2 years old.

I'm looking to try and avoid having to replace an entire battery bank at some (unknown and probably inconvenient) time in the future.

 

[Chris Hobson]

That depends on how hard you have worked them. If you connect a battery balancers to your strings, you should get you batteries to charge more evenly. There is some discussion on this forum in the review section. Some folk were unhappy about the parallel connections required for multiple banks. As long as those paralleled connections are fused I think it is OK. Read up and make an informed decision. With proper battery balancing I would risk mixing batteries of different ages so long as there is not a huge disparity. I think that batteries that are less than a year old and proper balancers and a BMV 702 to warn you of any imminent battery failure should work. I have nothing to base it on other than gut feeling and since I am on holiday my gut is feeling more than usual. You are doing something not recommended but I understand you problem. Have you thought of selling your battery bank in its entirety and replacing with the size you want.

Sent from my SM-G925F using Tapatalk

[DeepBass9]

Well, rarely less than 50% DOD, and I periodically equalise them though an overcharge using the generator.

It would be an expensive exercise to sell all, and then buy another set of batteries, what I have is adequate, but a little more would be nice?

Maybe another option is to have two battery banks charged from their own MPPT, and switch from one to the other? Maybe that is creating a problem that doesn't exist though

[Chris Hobson]

Having worked them hard I would not mix and match.

Sent from my SM-G925F using Tapatalk

[___]

Guys, read this paper: http://neuralfibre.com/paul/wp-content/uploads/2007/05/can-we-now-sin.pdf

Telecoms operators replace/add parallel strings all the time. Of course it will have an overall effect on the life of the other string, the important thing is to understand if and how this can be dealt with.

The long and the short of it is this: That when you discharge, dissimilar strings don't matter. Each string will contribute according to its own capacity (as long as they have the same chemistry). The trouble comes when you recharge. Now one cell fills up before the other, leaving you with two choices, either overcharge one, or undercharge the other. The obvious answer is to rather overcharge one string and avoid sulfation on the other. This means there will be a slight reduction in life on that string, but that might be acceptable.

You obviously cannot do this with sealed batteries. With flooded cells, a slight overcharge is often not too much or a problem.

As Chris said, balancing can help you get away with it too (that sorts out the distribution during charging).

[SilverNodashi]

Generally you don't want to mix-n-match sealed batteries but I know "traction cell" batteries can be mixed like this, since they can take this kind of punishment. At the end of the day you would then just replace a single 2V cell, or perhaps a 6V battery instead of a 12V battery so the costs would be lower in such a case. 

Sealed lead acid batteries on the other hand don't like this very much. As explained above, the two sets won't charge the same and one set will charge a bit harder than the other - which isn't ideal. 

[Guest]

If Trojan batts are more than 6 months older to new ones, I was told not to mix and match, for battery supplier.

[DeepBass9]

So it doesn't make much sense then to have a huge battery bank, because if one battery fails, you have to replace the whole bank, rather have a small bank and work it harder and then replace sooner. Seems like a lose-lose proposition either way.

[Guest]

Ditto DeepBass9, unless you replace the damaged battery/ies knowing full well the new one/s will be brought down to the older ones levels, and the sums still make sense.

Or you sell the older working batteries to a new person starting to learn, getting a whole new bank.

[Guest]

The less batteries you have, the better it is in the end. Try and stay above 50% DOD and, you need to check the water levels and gravity per cell as frequent as you can to find the problem before it is too late.

That way you can have a set of batteries lasting you 10 years plus, if you bought the right model.

[Guest]

So if a balancer makes sure all the cells are the same level why can you not mix old and new as the balancer will keep all on the same level. 

[Guest]

You can as long as you are aware that the older batteries will 'bring down' the new battery/ies to their current operating levels, so the new battery/ies will have the same life expectancy as the older bank.

It depends a LOT also on how hard the older bank worked.

[Mark]

I agree with jdp!  I have installed the HO02 balancer (discussed here) and have found they have made a huge difference.

I have 2 banks of 4x 100ah (Excis sealed, maintenance free el cheapo's ;)) - SO (ready for some discussion) I have also linked the batteries ACROSS the banks.

They are linked anyway (at the middle of each bank) because I measure the midpoint voltage variation (using the BMV702).

My suggestion:

1. Mix the new and old batteries placing the new batteries across the strings (and not all the same new string).

2. Install a battery balancer (eg HA02) or multiple HA01's (like Chris has done).

3. Periodically (every 6 months) rearrange the batteries in each string (as per TTT or Plonksters suggestion - can't remember) so that batteries get an equal chance to the in positions 1, 2, 3, 4 (4x Xah, banks)

Thoughts anyone...

 

PS:  Here is a question for debate (given the balancer above)...

Can you add a 24V MPPT charge controller to the center of the 48V bank to top up the bank (low amp 5-8A) in the late afternoon (another small PV String facing the setting sun)?  AND let the balancer do its stuff... batteries 2,3 will then get some extra charge.

[Guest]

Why? It does not make sense? If the balancer keeps all the cells at the same level then they are working the same amout and when the old bank fails you just replace those cells and the cycle continues. That is what we do with our big ups banks at work. Why is it deffrent now? Everybody says that but can not explain why. It seems everybody says that as that is the thing to say. 

[Guest]

JDP, suggestion, why don't you ask a few battery supplier technical departments, because just maybe you have opened a can of worms by adding battery balancers on the bank, which is not something I have heard anywhere else bar here?

Mark, reshuffling the batts I got from a local Trojan battery supplier. Trojan themselves says it is not necessary but it will not hurt if you want to do the effort.

I would not add a 24v controller to the middle of a bank. The main controller will get readings that makes it assume the charge cycle is at X, instead of Y.

[Mark]

Thanks TTT... good enough reason for me... although there are many instances where there are banks with multiple MPPT controllers.  ie An Infini and an Axpert both with their own PV strings...

[Guest]

Yes, you can add multiple controllers to the same bank IF they are connected to separate arrays, provided they are the same brand and model, for different makes / models may have have slightly different charge algorithms.

AND the matching controllers need to be set on the exact same settings, then they will not argue on who is in bulk charging mode and who is in float. I tried it. 

In my case the one controller ruled, as I used 2 different makes. The other one was 'off' as it saw what was going on as the batts being fully charged, even thought they where not.

Interestingly Morningstar was the first to inform all their customers they can do so if the T&C's are met as above, without having to interconnect the controllers like some other brands require.

[___]

1 hour ago, Mark said:

Can you add a 24V MPPT charge controller to the center of the 48V bank to top up the bank (low amp 5-8A) in the late afternoon (another small PV String facing the setting sun)?  AND let the balancer do its stuff... batteries 2,3 will then get some extra charge.

Balancers are limited in how much charge they can divert, and they are not particularly efficient either. The above does not sound like a good idea :-)

42 minutes ago, The Terrible Triplett said:

AND the matching controllers need to be set on the exact same settings, then they will not argue on who is in bulk charging mode and who is in float. I tried it. 

Argue? Well I think I get what you are saying, but more likely, as the bank starts reaching any of the critical levels, you will find that some chargers will back off sooner than others. So I could not for the life of me figure out why this might be bad... and then suddenly I could. Illustration. You have two chargers. One charger decides the batteries are full and goes to float (27.6V). The other remains in bulk (28.8V). Then a cloud moves in front of the sun, and now we're not making enough power to lift the bank to 28.8V. Lets say we still manage to lift it to 28V. So now one charger sits there doing nothing (its in float, which is lower than the present battery voltage) and the other isn't capable of getting up to absorption voltage.

If they were in agreement, the full capacity could be used for absorption.

Overall though, I'm not sure if it's really "bad" for the batteries. One of your chargers is going to "win" the argument, and as long as all of them are adjusted more or less correctly, it shouldn't matter too much which one wins. On a good day, all of them will eventually go to float. On a bad day, all of them will remain in bulk/absorb mode. It's those intermediary days that might pose a problem... how big a problem really?

[Mark]

I wonder if you set the (smaller) MPPT to just float the bank... and the main controller does both.  That way they won't "argue" past the float stage.

[Guest]

Although I agree with your logic, also thinking that the real effect may be negligible, IF they are close, but we have to assume that there are more like me out there that will go and do it, not fully understanding the way it works.

I may miss a very salient point that makes it problematic. So I presume the manufacturers know to the T what their products can do, how they wrote the software, to know of a potential problem that may be caused over a period of time on the set of batteries being charged.

In my particular case the 200w panels was made redundant, added no value.  = wasted power.

So I would prefer to follow manufacturers guidelines than reasoning it out for myself i.e. using non-identical controllers charging the same bank. 

[viper_za]

4 minutes ago, Mark said:

I wonder if you set the (smaller) MPPT to just float the bank... and the main controller does both.  That way they won't "argue" past the float stage.

I'm thinking the same here, does not even need to be MPPT for this one.

Only thing worrying me a little here is how would it take begin configured to float and seeing the voltages are way above it when the other charger is charging the bank

[___]

My next plan, once there is a bit of spare money again, is to add some west-facing panels. Separate controller. That should avoid any confusion. In the morning, the main array is boss and wins all arguments. In the afternoon, that one doesn't make much power, but the West-facing one does, so it now calls the shots. Doesn't matter if it switches to float during the day... in fact that is what I would expect.

[Guest]

8 minutes ago, Mark said:

I wonder if you set the (smaller) MPPT to just float the bank... and the main controller does both.  That way they won't "argue" past the float stage.

Took 2 controllers, separate panels, settings identical between the small and larger controller.

One day I accidentally saw the smaller controller showing zero watts coming in most of the time. Wasted that 200w all this time.