# Combining strings of different batteries.

Options
Solar Expert Posts: 410 ✭✭
I'm working on a project with three large VRLA battery banks. Absolyte GP. They have been in service for 4 years and all seem to be doing pretty well. (one slightly week cell will be replaced soon).

Two banks are rated at 2088ah ( C/8 ) the other is 1800ah ( C/8 )

They were originally connected to seperate SMA Sunny Island systems but because of inefficiencies in generator runtime due to uneven load sharing, the system may need to be reconfigured into a single battery bank. The additional capacity is really tempting and may actually be necessary but we really need to do the homework to justify the expense of replacing any of these.

I've got my own reservations about doing this but I would appreciate your thoughts. I'll be checking with the mfgrs (not today) and will let you know what they say.

Thanks,

-Alex Aragon

• Solar Expert Posts: 410 ✭✭
Options
Re: Combining strings of different batteries.

Yes, I know, it is an ugly proposition.

My fear is that the smaller bank will mess with the voltage readings of the Sunny Islands and lead to a chronic deficit charge of the larger batteries. Ken fro SMA Sunny Island tec support agrees. His comments were that the small battery stack would "burn out" while the larger ones would be chronically under charged.

My problem is that I just don't have enough letters after my name for corporate big-wigs to take me seriously.:roll:

Is there any way to quantify the issue here?

-Alex
Options
Re: Combining strings of different batteries.

There is a good engineering rule of thumb... Basically anything different by a fact or 2 or less is roughly the same. Anything that differs by 10x or more, the smaller thing/effect can be ignored.

So, if your battery banks differ in AH capacity by 2x or less, you probably can combine and "bite the bullet".

If they are more than 2x different, then it starts becoming too small to matter (i.e., would you go to all the work to combine banks if you the smaller bank is only 33% or less of the total bank capacity--Remembering that lead acid battery cells are only within +/- 10% or so of each other capacity anyway--i.e., a 20% spread between low and high capacity cell in the same battery--worst case).

Batteries, if they are the "same" model/technology/brand/electrolyte fill, they probably can be paralleled together OK. Basically, the chemical reactions set the "voltage" of the battery (charging/discharging/resting), and the current is going to be a combination of the battery chemistry (voltage--hopefully matched) and AH capacity (resistance and chemical reactivity of the area of the plates--smaller battery will output/absorb less current based on its size)...

All that sounds good, but then we start with paralleling things that "are different"--And many times, those differences amplify/diverge. What happens if the larger bank gets "hotter" (less surface area per unit volume, so it might not cool the same). That means that string may absorb more current (hot batteries, lower charging voltage). Which reduces available current to the smaller batteries, which under charges them, etc...

But--who knows, what if the smaller bank runs hotter (less thermal mass). And they started more sulfated (lower operation voltages)... They get more current and get even hotter--and age faster (another engineering rule of thumb, things that are 10oC hotter, age 2x faster).

Now you are chasing your tail, trying to balance loads/currents/temperatures... Trying to catch weak cells (open/shorted) that may take down more batteries or even the whole bank, and in there every few months replacing "one more battery"... And what do you replace it with, the smaller battery in the current string--because you cannot replace it with the larger battery (miss-matched series AH capacity amoung cells--big no-no)... So you end up replacing the entire small string at some point anyway to match the rest of the bank?

Anyway--Lots of hand waving. Perhaps you can use this as a test site (lots of cool test equipment, trips to middle of no-where for camping/hunting/fishing) and document different solutions and what happens 2-5 years down the road (is any management looking 2-5 years down the road right now????).

In the end, my 2 cents... I would not combine strings where the batteries had a AH capacity greater than 2x different from each other (i.e., 300 AH vs 600 AH)--It is not worth the extra capacity... And I would have "a plan" on where to go next (replace up to 10% of a string before replacing the whole string--And would that string be same capacity or equal to the larger bank capacity... And what if the larger bank needs replacing--would be be downsized or kept the same? etc....).

-Bill
Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
• Solar Expert Posts: 10,300 ✭✭✭✭
Options
Re: Combining strings of different batteries.

uneven charging from a single generator to 3 separate systems would make me think 3 separate gennys or at least charging separately with one genny, but you state a limited charge time. i don't know if there would be inherent problems in combining the battery banks, but i suppose you could try it to see what happens and that's easier said than done.

as far as letters after my name go i was thinking i don't think peon or us citizen works well either. btw, the title moderator doesn't either excepting for naws itself and most of the membership here.
• Solar Expert Posts: 410 ✭✭
Options
Re: Combining strings of different batteries.
niel wrote: »
as far as letters after my name go i was thinking i don't think peon or us citizen works well either. btw, the title moderator doesn't either excepting for naws itself and most of the membership here.

I'd buy you a cup of coffee.
• Solar Expert Posts: 410 ✭✭
Options
Re: Combining strings of different batteries.

Thanks Bill, for your "lettered" response.

Definately some food for thought. The idea of +/- 10% for new batteries is pretty scarry. Have you ever heard of mfgrs making "matched sets" to increase performance. It seems that even a matched set of "lows" would be better than a mixed set of ≥ 90% and ≤ 110%.

I was actually musing the idea of feeding the smaller bank with longer cables. More resistance on charging and discharging could theoretically help to even things up. It would be a great case study. Unfortunately this site is snowed in more than 6 months / year, so it really needs to be right or it causes lots of headaches.

-Alex
Options
Re: Combining strings of different batteries.

"They" do matched sets for Lithium/NiMH/etc. cells for R/C (radio controlled) cars and planes, etc... (either done by the manufaturer, or with battery chargers that can do measured discharging/charging of cells by the end user).

This allows them to drain a pack dead without worry that a cell will "reverse" bias/charge and be killed (vs the cost/weight of adding monitoring to every cell).

Given that lead acid batteries need cycling from new to obtain their maximum capacity (down 25 to 30% until there are a few dozen cycles?), you will have fairly large spread until the batteries are "broken in".

In some ways, this is actually a bigger problem with 48 volt banks... You lose 1 cell and the bank is now 44 volts--Still within standard operating voltage range (2 volt failure out of "48 volt" was "hard to see").

With 12 volt bank, you lose a 2 volt cell, and you are at 10 volts (2 volts out of "12" was much more obvious)--You know you have major problems.

When I was working on 48 volt projects years ago, I looked for per cell/per battery voltage monitoring wondering how Telco's were monitoring their cells for best life/performance.

In general, it turned out they did not really care. Either the cells sat at "float" for pretty much their entire life, or were drained well beyond their standard "low state of charge" (which may have been below 75% SOC--very shallow cycling for many Telcom/float service batteries) such as when a generator ran out of fuel or failed... They would then expect the battery to run until failure or power was restored--And at that point the bank would be replaced anyway.

-Bill
Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset