You are correct but to suggest that we can connect two batteries that are at considerably different potential is misleading, wet cell batteries have low internal resistance which means if we parallel two that are at different states of charge and therefore terminal voltage, massive amps will flow albeit for a brief period of time.
Allan, yes that was a bad thing to say now that I look at it. You, of course, are correct. Amps will flow from the fully charged one to the one that's at a lower state of charge and they will equal out. I guess that was my point - you can do it and charge or discharge them and the two will equal out with no adverse effects on the batteries (as long as they are both healthy).
Truthfully, there is no way to configure a bank where in the event of a failure, there will be no adverse effect on the other batteries. It seems all these arrangements have been discussed and documented by someone else in the past and this arrangement is nothing new
This is correct, Watt. I think the key is early detection of that failure so it can be dealt with before it causes failure of other batteries in the bank. I made a video of what it does when I hooked a little 2 amp 12V battery charger to one string in the bank and posted that video in an earlier post. Within seconds it caused enough imbalance to see it on the meters.
My theory here is that if the amp-hours of both strings are identical, there should never be an imbalance in voltage. If I lose even one cell in one battery at some point, that changes the amp-hour capacity of one string. And it will show up as imbalance on the meters during load (lower voltage on the string with the failed cell) or charging (higher voltage on the string with the failed cell).
I read your post earlier and I tried something else this afternoon. We went walleye fishing last night (didn't catch any fish) and the batteries in the boat were down on charge so they read 12.4 volts at-rest. Those batteries are 95 ah each - Group 27 marine deep cycle batteries. I got an idea from your post - 95 plus 95 is 190 ah. The batteries in my bank are 200 ah. This is all "nameplate" stuff because one T12 weighs double what those two Group 27's weigh combined. But that's what it says on the tags.
Our bank was floating most of the day and I float those T12's at 28.0 volts (temperature compensated was 27.7 volts today). I disconnected one battery in Bank 2 and replaced that battery with the partially discharged Group 27's by hooking them into the bank with jumper cables to see what happens. The voltage on Bank 2 instantly dropped to 13.7 when I hooked them up. The voltage on Bank 1 went to 14.0.
Now, with these two boat batteries we have 10 less amp hours on Bank 2 than we have on Bank 1. I left them hooked up all afternoon (for about 3 hours now), and they're still hooked on to Bank 2 as I write this. The balance meters evened out within a couple hours, but now after about three hours the voltage on Bank 2 is showing higher than on Bank 1.
I am completely satisfied that if I lose even one cell in one battery I will be able to detect it with those meters and this bank wiring configuration. The reason is because the bank always runs at a certain system voltage - let's say it's 26.0 volts for easy example. The balance meters should show 13.0 volts each. But if you have even a slight imbalance in one string from a failed cell it will cause either a drop in voltage on that string (under load) or a rise in voltage (under charging). At the same time it causes a corresponding rise or fall in voltage from the 13.0 on the other string, which makes it very easy to detect on the analog meters.
In the above example, if the system voltage is 26.0 and Bank 1 is at 12.95 because of a bad cell, Bank 2 is at 13.05 - a difference of .1 volts which is easily visible on those meters.
I've done enough experimenting now to know it works and will detect even a slight imbalance. My batteries are due for service again in June under my previous regimen. I'm going to leave it hooked up this way and skip that service interval. Servicing involves starting the generator and switching all our loads to genset, shutting down the inverters, solar and wind. Then remove each battery from the bank and record at-rest voltage, load test it and record loaded voltage, then leave it set and record at-rest voltage after it recovers from the load test. It's a long drawn out affair that takes 4-5 hours, and I hate doing it because my back is about shot after handling those freaking heavy batteries.
I'm going to leave it this way until the September service interval and just check water once a month. I'll service them as normal in September by load testing, etc.. If they all show identical like they did for the last service interval I will then consider this wiring method to be "proven" and it works.
--
Chris