Max parallel banks of AGM?

[entaran]

Hi Guys, new here.

I have a not-trivial off grid setup consisting of 110 250 watt panels and two banks of 600ah 2v AGM's in 48v configuration. (total usable 600ah @ 50% dod).

I also have a non-trivial load on my inverters sometimes which is causing the banks to discharge at well over 100 amps each in some cases. This is obviously not optimal.

Is there a working limit on the number of parallel banks of 48v 2v cells I can setup? Like 4 or 6 or similar? I've read that charge controllers start getting odd towards the higher end but I have not been able to find anything to verify. I would like to increase the number of parallel banks to reduce the high demand load from the banks when my industrial machinery is running in my workshop. My property is fully off grid. Grid connection isn't feasible (WELL OVER $150k to connect, and 3phase not available).

Important Bits:
2 banks of 24 2v 600ah cells in parallel (48v).
2x Victron 250/100 Charge Controllers each with approximately 60ish panels connected in 6sxp (peak voltage about 242v OC)
2x Victron Phoenix 5000va 48v Inverters running split phase.
1x 15kw 48v three phase inverter (custom made by manufacturer) to run heavy equipment (This can pull 300 amps alone @ 48v)
110 250w panels in 6s18p configuration. I think the last two are actually 5s2p because you need 114 panels from memory to make it all 6s.
Yes, I run my system hard. Currently my batteries almost never discharge below 80% and I like it like that.

Thanks in advance

 

[snoobler]

Welcome to the forum.

Pulling 300A from a 600A bank is a 0.5C discharge rate. For AGM, you're not going to get anywhere near your rated capacity. Since you don't discharge below 80%, you must not use 300A for very long.

Wiring Unlimited indicates 3-4 strings. I've seen similar recommendations from Rolls and Trojan. It also demonstrates proper wiring to optimize balance between the strings.

Don't know what you mean about charge controllers "getting odd" towards the higher end. They don't care. They just provide voltage and current until the battery says "enough."

Normally, I'd say you have WAY too much solar for those batteries, but since your charge controllers are maxed at 5800W, and they're limited to 100A each, your 200A charge current isn't to bad for 1200Ah of batteries.

In other words, you have 110*250 = 27500W of solar on charge controllers that can only use 11600W.

242Voc on a Victron 250/100 is dangerously close to the limit. If your temps drop below 18°C, you're going to hit the limit and risk damage to the MPPT.

They also have an input limit of 70A on the PV side. You're not supposed to violate this on a Victron.

If you have all of these positioned at different orientations to get more even power throughout the day, that's fine. If it's one big array on each MPPT, then I'd be taking a hard look at the numbers.

Lastly... if by "last two are actually 5S2P" you mean that you have 5S2P in parallel with 6SXP, you are sabotaging the entire array. Each parallel string needs to be the same Vmp/Voc. You will likely get more power from the array by ditching the extra two panels. If you've factored orientation into this, then it may be a non issue.

 

[Topper]

I ran 12- 24v strings 170ahr each for a number of years. Each string had a failsafe fuse and 100A breaker. These were 12 volt telecom batteries. All was decent enough for a while, but 12v packs did wander high/low outside of what i considered to be my safe limits. To resolve the issue, installed a number of QNBBM-12 active balancers, which worked flawlessly for me. To save some cost, i grouped 4 strings per balancer pair and ran a 250mA fused center tap per bank. I'd guess this got me another 3 or 4 years out of them on average.

 

[Topper]

http://www.cpspower.biz/6-shelf-battery-rack.html

 

[Luthj]

Typically 4 parallel strings is the max. Though for low C applications you can sometimes get away with more.

Another option is to build a parallel then series battery. This requires putting cells in parallel with each other first. I would not recommend doing this with anything other than new cells which are identical.

How long of duration is this peak load for? Short periods typically aren't a problem as long as you are within the manufacturers published limits.

 

[time2roll]

I agree with 4 strings max if just connected to a bus. More than 4 and you should fuse each bank separately between the battery and the bus.

That is a lot of AGM. Price some DIY LFP and you might be able to expand your capacity for less money.

 

[Topper]

Personally, i am a strong believer in fuses and breakers on EVERY parallel string between the battery and buss bar.
My mindset:
I'd much rather see a casade failure of system protections (fuses, breakers) then the battery unit in part or as a whole. We've all seen or heard of a single bad cell taking out a parallel set or an entire bank due to cascade failure without proper protections in place. My largest fear is always the worst case scenario of a thermal runaway and fire.. I may be overly cautious, but that's just me. In my humble opinion, fuses, breakers and a properly designed system are the cheapest insurance on your investment, your home and your family.

 

[entaran]

Hi guys,
Thanks for your insightful replies.
I have adjusted the solar array to be 6sxp only. I have removed the extraneous panels (simply by disconnecting them for now).

I have individually fused each bank of batteries prior to hitting the bus. They are then again fused after the shunt and before the inverters/charge controllers with larger fuses.

Peak loads are usually only for a few seconds as large machines spin up. Like my Milling machine or Lathe. When they are cutting under load they can use high loads for 20-30 seconds at a time but it will revert down fairly quickly as the cuts don't take overly long. Generally *worst case* the inverters are providing around 4-6kw of power for long periods when washer/dryer/dishwasher/etc may be running. This would result in around 80-120 amps of draw for a couple hours, but this to my understanding should be split across the two banks for 40-60 amps per bank?

The batteries are already wired correctly as per "Wiring Unlimited", as I use 2v cells that have the positive and negative connection at each end and everything in between is in series. The wiring between each bank is exactly (to the mm) the same length on both banks to their respective fuses and then to the busbar. It is also the same size (75mm2) with the same size lugs hydraulically crimped to the cable.

My 250/100 charge controllers are running ~48 panels each in 6s8p. These panels put out ~232v and 8.36amps OC which results in a max load of ~232v and ~64 amps to the CC. This is at approx 5 degrees C when I measured it which is way below 18. There could be some creep below zero but it's never sunny and below zero here, I live in Australia. The excess panels have been disconnected because apparently I was overloading the CC without being aware (didn't know about the 70 amp limit until I went and re-read the manual on it, I missed it apparently which could've been expensive but I got lucky).
However, of those 6s8p arrays, 24 panels face east and 24 face north on CC #1, and 24 face west and 24 face north on CC #2. I am at 38 degrees south latitude so they are tilted 38 degrees to the north for me and the east/west panels are tilted at around 45 degrees to catch early morning/late afternoon. I have so many panels I'm not overly stressed about actual perfect orientation. I mostly went with a huge number of panels because I get a lot of heavy cloud for days on end so generate at like 10-15% of actual capacity. (And my house is off-grid 100%, no grid available).

 

[Brett V]

I built this as a UPS for basics in the house using "new" 180ah telco batteries in a 2s4p 24 volt arrangement using a balanced wiring harness. By new, I mean the batteries were never placed in service but sat in a warehouse for about 12 months without a refresh charge when I acquired them. Not great but they were free. It mostly sits and I float at 27.2 with temperature compensation. It has a small <1 amp load for the standby on the inverter and other parasitic loads such as the low voltage disconnect and desulfator. Things were pretty stable for about 6 years before I started having trouble with balance and I bought four active balancers which helped a lot. I believe if I'd had the balancers from the beginning, I'd probably not be looking at replacing them now. One string is toast and I've since split them up in to three separate strings going through individual 100 amp breakers to the bus bar. I can only get about 3 1/2 hours at a 45 amp load before reaching the 50% DOD point. It was a fun learning experience and I have very little invested in it. I'd never invest a bunch of cash in AGM batteries from what I've learned. Flooded 6 volt cells would probably be a better choice for my application but I don't want the hassle of maintenance that goes with them. Lfepo4 will be replacing them.