Renewable Energy Questions/Discussion > Automation, Controls, Inverters, MPPT, etc
Battery Bank Configuration
ChrisOlson:
I service my batteries (24 Rolls-Surrette T12-250's) every three months. The service procedure involves fully charging the bank as the first step. Then, with the generator running and my RE system shut down, I flip my big Ronk 200 amp manual transfer switch. The transfer switch disconnects gen power to the inverters while simultaneously disconnecting the inverter output to the loads, and switches all the loads over to gen power. It totally removes the inverters and battery bank from the loads so the inverters go idle.
I takes about 4 hours to service the bank. Those batteries are heavy and hard to handle, but I pull them off the bank in pairs to check them out. I stick a 2" wide piece of painter's tape to the top of the battery and record the at-rest voltage of it on the tape with a marker. I check the electrolyte level in the cells, and check the SG of each cell with a hydrometer and record it on the tape. I then apply a 50 amp load to the battery with my Sun AVR for one minute and record the loaded voltage that it shows after one minute on the tape. If the battery checks out it goes back online.
When I serviced my bank in September 2011 I had the batteries hooked up this way:
What I found was batteries all over the place on SOC. Some of the batteries varied from others by up to .28 volts at-rest and the ones with low at-rest voltage fell flat on their face during the load test.
After servicing all the batteries I sat on my upside down five gallon pail staring at the bank while drinking a beer and decided this is not good. I got 12 series connections in the bank and 24 cables going from the bank to the bus. No two batteries are ever identical. But the challenge is, when you connect them series, just how do you determine which ones are matched so they stay even during charging and discharge?
So I changed all the bank wiring around to this:
I eliminated 12 cables from the bank to the bus. I eliminated 6 series connections in the bank. I figured, what the heck - with pairs in parallel like that, and connecting two pairs in series, maybe it will stay balanced better. It's worth a try.
I got a little behind on my bank servicing this last time with the holidays and stuff. I just got it done yesterday, for the first time since I had changed the wiring around. It worked! The furthest off any two batteries were in the entire bank was .04 volts this time. They all tested 1.270-1.275 SG, and all load tested so close to identical that it's not even worth mentioning or recording the difference between any two.
--
Chris
WooferHound:
Are you sure about that second battery diagram ?
It looks like some of the batteries are shorted out . . .
Watt:
--- Quote from: WooferHound on February 02, 2012, 07:47:58 pm ---Are you sure about that second battery diagram ?
It looks like some of the batteries are shorted out . . .
--- End quote ---
They are not shorted, they are in parallel series. Equal to
l-- + 24v buss bars - ----- i
l l
l l
l --- + 12v group - - l l
l l l
l l l
l l l
l- +12v group - - l -- l l
l l
l l
l - l -- + 12 v group - --i l
l l l
l l l
l l l
l l l
l -- + 12 v group - -- l --l
that. IF it worked.
ChrisOlson:
--- Quote from: WooferHound on February 02, 2012, 07:47:58 pm ---Are you sure about that second battery diagram ?
It looks like some of the batteries are shorted out . . .
--- End quote ---
Yep. It works. You take two batteries and make one big one with parallel connections. Then hook the parallel groups in series.
My theory on why it works better is because batteries in parallel tend to support one another even if they're a little bit different. Series connections between batteries are always bad because if, for instance, the charging voltage is 30 volts and one battery is at 14 volts, the other will be at 16. The one that's "lazy" gets deficit charged and the other one gets the snot boiled out of it.
By grouping the batteries in pairs with the parallel connections it must've eliminated the problems with slight mismatch between batteries that tends to show up more with series connections.
I thought about doing groups of three in parallel. But then I would only have 8 cables from the bank to the bus instead of 12 cables. It's 9 feet from the bank to the bus and my inverters draw 720 amps @ 24 volt nominal at full load. With 12 cables I can get by with 1/0 from the bank to the bus, as each cable only carries 120 amps. If I used a triple parallel configuration, then each cable would have to carry 180 amps and I'd have to use 2/0, and I'd still get more voltage drop to the bus than I do with the 1/0 and 12 cables.
I got 4/0 cables from the bus to each inverter, but those are only 6 feet long.
--
Chris
Volvo farmer:
--- Quote from: ChrisOlson ---
My theory on why it works better is because batteries in parallel tend to support one another even if they're a little bit different. Series connections between batteries are always bad because if, for instance, the charging voltage is 30 volts and one battery is at 14 volts, the other will be at 16. The one that's "lazy" gets deficit charged and the other one gets the snot boiled out of it.
--- End quote ---
I have the feeling we have been over this before, but everything I have ever read by solar installers and people who have been working with batteries for over 20 years, recommends the opposite. I have heard that it is parallel strings tend to stray in their SOG and that this tendency is reduced if one has a single string of big old honkin 2V cells.
Isn't a 12V battery just a series string of six 2V cells? Why are 6 2V 100ah batteries in series bad but a single 12V 10 ah battery in parallel with nine others good? Do all ten of those parallel strings really support each other?
Navigation
[0] Message Index
[#] Next page
Go to full version