Series/parallel battery bank.

icarus

Working with a new client.  He has a 24 vdc battery bank, made up of 12 L16s Crown batteries.  I want to advise him on the best way to wire the series/parallel connections (including fusing the series batteries) He is eating batteries and we are working on trying to “clean up” his system and his operation so that the next set has a bit better performance.  

I am pretty familiar with 2 series wiring, but not quite clear on 3.   

Thanks in advance for advice.

Tony

(PS Pictures would be great!)

BB.

Hi Tony,

We have been pointing folks here:

http://www.smartgauge.co.uk/batt_con.html

By the way, have you figured out why he is going through batteries?

-Bill

icarus

Bill,

Nice to hear from you again!.

I just met the fellow today.  Interesting guy.  Way off grid for 30 years, now nearing 80, and he got in touch with me for some other energy issues (how not to have to cut so much wood!). I don’t yet know what the real issue is, but I think it is a number of issues.  The first is his cabling is not a crisp as I would like, and the batteries are very unbalanced (SG). I think the gist is also that he has his inverter to auto start, and the genny does (I think) a full charge regimen when ever it runs.  Because of his multiple fridge/freezers, the combined load often triggers a low voltage gen start command in the middle of the night.  The net result of which is very excess run times, AND the batteries are fully charged when the sun comes up, sort of defeating the whole PV concept.  

The really surprising thing is that while he has all kinds of ways to monitor his system (CC/Inverter/Trimetric etc) he has no idea of how much he routinely uses per day.  The system is pretty well set up...2800watts of PV, A pair of 60 amp controllers,(1 Outback classic, 1 Xantrex MPPT which seem to play all right together, I would prefer them to both be outback’s...)  running into 12 L16s, feeding a 3k Outback inverter. 

The other issue is he has a 15 KW inverter propane genny  which is greatly oversized.  By my calcs, a 6-7 KW would bre a much better fit. As it is, it is burning nearly 2 gph of propane.  (He can get bulk delivery so transport is not difficult.) 

I’m going to work with him for a while and see if we can define the loads, and try to teach him how to manage his loading a bit better.  

Thanks for the input,

Tony

icarus

Bill,

Following your link, I am still unclear about 3 parallel strings.  4 batteries in simple series, 3 strings of 4.  Trying to figure out the best cabling and take off points for this set up.I should have taken a picture of the cabling.  I try to get him to send me one.  

Tony

BB.

Remember each "rectangle" in the above link is a set of cells/batteries @ bus voltage (12/24/48 volts)... Each string of batteries should have the same total wire length (you can mix the location of long/short wire jumpers--But the total length should be the same for all parallel strings).

If you want to fuse each parallel string of batteries, I really like #3--Also, if you are using a DC Current Clamp Meter to monitor voltages, #3 is easiest to monitor (each lead to the +/- battery buses only carries that string's current--Easy to check balance). The X parallel connections from battery string to their respective bus connections should be the same (again, total length, a long+ and short- on one string matches a short+ and long- on its mate).

If you are going to fuse each string--The Blue Sea fuse blocks are neat (and small):

https://www.bluesea.com/products/category/16/72/Fuse_Blocks/Terminal_Fuse_Blocks (one and two fuses per "block")

As you know, good fuses/blocks and even DC breakers are not cheap... If you can swing breakers (space), they would be a nice on/off/debug switches per string.

A little sanity math... 12x 6 volt @ 400(?) AH batteries:

• 12 * 6 volt * 400 AH = 28,800 WH
• 4s * 3p = 24 volt bank @ 1,200 AH
• 1,200 AH * 0.10 rate of charge = 120 Amps (minimum for deep cycle 10% rate of charge)
• 1,200 AH * 0.20 rate of charge = 240 Amps  (max nominal for deep cycle 20% rate of charge)
• 15,000 Watt genset * 0.8 derate = 12,000 Watts (max suggested loading)
• 15,000 Watt genset * 0.5 derate = 7,500 Watts (minimum suggested "most efficient" fuel loading)
• 120 Amps * 29.5 volts charging * 1/0.8 genset derate * 1/0.8 charger eff+PF = 5,531 Watt/VA genset min rating
• 240 Amps * 29.5 volts charging * 1/0.8 genset derate * 1/0.8 charger eff+PF = 11,063 Watt/VA genset min rating
• 12 * 6 volt * 400 AH * 0.85 AC inverter eff * 1/2 days storage * 0.50 max planned discharge = 6,120 WH per cloudy day (or over night) battery loading (2 days storage and 50% max planned discharge)
• 1,200 AH * 29.5 volts charging * 1/0.77 panel+controller derating * 0.10 rate of charge = 4,597 Watt array @ 10% rate of charge
• 1,200 AH * 29.5 volts charging * 1/0.77 panel+controller derating * 0.15 rate of charge = 6,896 Watt array @ 15% rate of charge
• 2,800 Watt array * 0.77 controller+panel derate * 1/29.5 volts charging = 73 Amps
• 73 amps charging / 1,200 AH bank = 0.06 = 6% rate of charge

Guessing at the sizing of the system... Close to a 20% rate of charge, he could justify that size genset--But getting ~200 Amps charging to the bank is a pretty big charger... And at ~1,200 AH--That bank would be a good candidate for a 48 volt system instead... And looking at doing a single string of ~600 AH batteries (or cells)--Size/weight of cells/battery/battery bank if issue of moving monoblock battery(ies) at the place. And a 2x larger array would not hurt...

Of course, a 48 volt battery bank means new AC inverter (or inverter-charger), new AC charger, rewiring/possibly new panel(s) for solar array @ higher voltage...

And depending on where the system is installed/amount of sun per day (and trees/mountain shade, especially in winter)?
http://www.solarelectricityhandbook.com/solar-irradiance.html

Bellingham
Average Solar Insolation figures

Measured in kWh/m2/day onto a solar panel set at a 41° angle from vertical:
(For best year-round performance)

Jan	Feb	Mar	Apr	May	Jun
1.65	2.79	3.53	4.36	4.76	4.93
Jul	Aug	Sep	Oct	Nov	Dec
5.44	5.42	4.80	3.10	1.88	1.45

If you aim at 25% discharge per day (2 days, 50% max)--This works well for lead acid batteries (especially during the shorter days around winter--Hours of "on charge" during the day). Guessing at 6,120 WH per day (25% of 400 AH @ 24 volt bank):

• 2,800 Watt array * 2.79 hours of sun (Feb break even) * 0.52 off grid AC system eff = 4,062 WH per day

Obviously, not enough for winter--And still ~2,000 WH per day "short" of break even for February (based on 25% discharge).

To help with longer "hours on charge"--A tracking array or a virtual tracking array (1/2 panels facing SE, other 1/2 facing SW) will give you better current at beginning and ends of day.

Then there is the age factor--If not an issue now, could be an issue down the road (5+ years)... A single string of batteries, fewer cells to water and connections to maintain. A 600 AH @ 2 volt cell is about as heavy as a typical 6 volt @ 200 AH "golf cart" battery--Easy to move--Less cost to replace a single "failed" cell.

Going with a new inverter-charger--Looking at one that has full AGS control (quiet times), and should be able to better estimate state of charge and generator run times (such as start at ~50% and shut down at 80% state of charge).

Any chance of bringing power to the property (i.e., cost of system upgrade/maintenance vs cost of running power lines)? I also suggest that putting $$$ into the solar system vs $$$ into power to property (I am not sure that $$$ in solar+battery+genset gives much return if/when the property is ever sold--Utility power to property probably adds near 1:1 to land value--pure guess on my part--check with local real estate agent).

Anyway--Just to back of the envelope calculations... What do you think Tony?

-Bill

icarus

I think I found it...

I thought I posted it last night...T


  

BB.

Yep--That drawing is correct and probably the simplest to implement and maintain.

-Bill

icarus

Thanks Bill,

Looking at his system, he has ~ 1200 AH @ 24 vdc, or ~ ~28kwh of capacity.  

That battery bank would like to see ~120-150 amps of charge current.  

His Outback inverter/charger only puts out a maximum of ~82 amps into 24 vdc.  

HIs PV on a “perfect day” is only going to put out ~90 amps...and even that will be rare.

I just got an update from the client.  It turns out he has 4800 watts of PV, not 2800 as noted.  Leaving a charge current potential of ~150 amps...better!

Bottom line.  He doesn’t have enough charge capacity for the size of the battery ban (coupled with a few other problems!). 

He might very well be better off down sizing the battery bank to 8 L16s which would now leave ~800AH of capacity, ~18kwh of capacity, 

That battery would like ~80-100 amps of charge current, within the range of what he could do.

What we (he) don’t know is what is daily consumption and max current draw is.  Until we know that on some reliable basis, we are shooting in the dark.  I have encouraged him I to get an Emporia metering system so that he can monitor all the relevant circuits (mostly fridges and freezers)  so we actually know where to begin.  I did suggest to him that upgrading to 48 vdc would be a good (albeit) expensive option.

My WAG on his consumption is that the bulk of the draw is the two freezers and fridge.  These are not modern, inverter compressor units, so I’m guessing the large freezers draw ~500 watts with a duty cycle of ~40%, or  ~ 4.5 KWH/day

The fridge may draw ~300 watts with a 50% duty cycle or 3.5 KWH

So his two freezers and fridge consume ~12.5 kwh/day maybe.  Simply changing out the fridge would save ~3kwh/day, changing out the freezers might save another 3 kwh each. Changing all three would save ~9 kwh/day...a not insignificant number given the problems with this system and it location.  


Gonna have to gather, and then analyze some data,  thanks for the input,

Tony

BB.

If he has 10+ year old fridge/freezers--They can really suck the energy. Modern ones are closer to 1 to 1.5 kWH per day. I personally like frost free--No having to defrost once or twice a year (or more often)--That is really nice.

The old EnergyStar website is a quick way to look up what is "possible" for low energy appliances:

https://www.energystar.gov/

If in Canada--Do you/they have similar?

Have fun,

-Bill

Dave Angelini

Tony,  really to do this right, is a string of 2V cells for a 24v bank. Or, go to 48v.  As you know,  the same problem will just keep happening, no matter how you wire strings. Just a poor way to do this long term but if funds are tight, it can help, for a bit.

Bill, they have the Canadian version of energy star on many US  refrigerators also. I find their number much more accurate. The US data, I think, is based on a warmer summer room temp. I have seen over 100 kwh differences and the Canadian are closer to what I see with my clients measurements. I also do not think that there is a significant savings in linear compressors as far as normal sized family refrigerators in the 20+ cubic foot NA units. Like you said 1 to 1.5 KWH is pretty typical no matter of the compressor design. You just have to pick a good one and measure the power.

examples of a unit I have been using for clients for 10 years now. The Canadian number has always been accurate.

https://images.homedepot-static.com/catalog/pdfImages/68/687a4db7-4c9b-4fc3-a238-dd8550c3aef3.pdf

https://www.homedepot.com/p/Whirlpool-22-cu-ft-Bottom-Freezer-Refrigerator-in-White-with-SPILL-GUARD-Glass-Shelves-WRB322DMBW/205065355#product-overview

icarus

Thanks for the input Dave,

One of the flies in ointment so to speak is I don’t have too much pull here.  The client (I’m doing it probono) has had a relationship with an off grid guy for years who seems to MOSTLY know his stuff.  The lack of fuses is a bit disconcerting. The client doesn’t want to mess up his relationship with the solar guy, which I can understand. This solar guy is going to build him a lithium battery bank based on some used batteries, the origins of which I know nothing about. 

Bottom line, if he can get his fridge/freeze use in order with modern ones, (and he is willing to give up one chest freezer) the consumption should drop enough to make some sense.  I still need to get some real data however.

Tony

Dave Angelini

Sounds like my recent client who had back surgery and now his wife is lifting batteries while I watch her on a camera. Same thing though, someone else's work and parallel strings. One string had a SG cell reading 0.00. The genset has been running every morning for 2 months and she found me through the Outback Forum.

 Her hubby is on Vicodin and morphine and really can't say much that I understand. No Probono for me ever, unless they come do summer work for me first. Below is a picture of another one a installer friend sent me of a power system he turned down.

Not shown and really close is a genset and 400 gallons of Diesel. He claimed to have got his permit by parking a tractor in front of the door. Notice the inverters are on! Just needs TLC right?

icarus

Wow...

I don’t do this for a living and while my knowledge is pretty good, I don’t like to change for advice. That holds true for much of what I do.  If I were 20 years younger, I could make a great retirement living “fixing things” for people, simply because few people are willing to fix stuff.  I still do a fair bit of it, but only if the project interests me.

T

Dave Angelini

Nice to be fully retired ! Not much longer for me and the 350+ offgrid who do lean on me are great folks. It has been a pleasure to know them and except for the last C19 year, not much stress living offgrid. I do not charge for general advice either. My favorite work is with Engineering in Burnaby, up your way. Really talented group of folks.

The picture above still cracks me up. I like the alligator clips and high voltage sign. There is not a fuse or breaker within 20 miles of that,  just like yours. Happy Hour soon !  Cheers! 

BB.

I am a big fan of placing Li Ion batteries outside the living space (separate outbuilding, etc.)... If there is a fire (whether solar related, or something completely different), the product of Lithium Battery combustion can include can include Hydrofluoric Acid--A very nasty chemical that generally cannot be "cleaned" from surfaces (pull the contaminated materials and take to dump).

Dave and Tony--I hope you have the next generation of solar installers/technicians properly trained and ready....

I have a friend that is a journeyman machinist (probably the last class to go through the full training program in California ~4 decades ago). There appears to be virtually no talent pool to bring the knowledge forward after he and his friends retire (or most have already retired by now). He tried to restart the program through the local community college he got his training--His old instructor said that the math/geometry/understanding of measurements skills of any new students were so low as to make the program infeasible. 

-Bill

icarus

Dave,  not fully retired, but not fully working either.  I take on some challenges when I want to and that keep the world away.  I guess I should say, I charge clients selectively buy usually in the broader case of a larger scope project.  I do a bit of energy audit/analysis mostly, and the solar comes in there.  

As for the future of the trades.  I think that is a trend that we see all over. The trades were “looked down upon” by many for too long, and too many people were college tracked who might well not be.  What has happened is a dearth of qualified people to do routine work that needs doing.  Western countries have a pretty difficult relationship with immigrants and immigration.  In most places immigrant classes have filled the void of skilled trades people, from carpenters to mechanics to electricians.  Without getting too political, the work ethic (and ability) of many puts many native born to shame.  When I started in the trades many decades ago there were some minorities, but few immigrants.  Now the trades are nearly exclusively immigrants of first generation.  The geo politics is pretty broad and complicated and I don’t want to open a can of worms here.

as for the next generation of what I do...I’ve seen a fair number of pretty clever, younger guys working in the solar biz, mostly grid tie.   Grid tie being fairly predictable.  As for what Dave and I do, it is a very specialized trade, that in my case boils down to simple (yet complex) problem solving.  What I bring to the table is a diversity, depth and length of experience to draw on, that encompasses many “trades”. I also avail myself of a few special experts to bounce ideas off of.  (Even including some here I might add).  My real specialty is solving problems in remote locations where you can’t just “call the plumber”.  You need to be the plumber, but also need to make the fittings to build the pump to run the water to use the plumbing so to speak, much of it with little access to outside material.  These days outside expertise is just a click away however.  

Anyway, enough rambling.  My current intellectual hill to climb is figuring out how to use lithium batteries (which I know next to nothing about) and how to use them in cold environments.  

Thanks for all the help, I’ll keep this thread going when I get some data from the owner.  

Tony

Dave Angelini

Hey Bill / Tony

Along this same thread of strings of batteries, one of the ways I have fixed a multi string system is to wire a Marine battery switch to allow easy ways to EQ one bank that may be of different age or one that has problem cells.

This Marine switch becomes a problem in 48V batteries that EQ at 65 VDC. In these cases I use 2 or more of the breaker/disco in the pix below. None of the Marine I can find are designed much above 48vdc max.

I would like to know if you guys have ever seen a switch only that could be used to lower the cost. There is already a 250A breaker in the power distribution panel. All I need is a switch that can run 65vdc and last at typical 150adc currents. This is to help folks get the last life out of a bank. Kind of what Tony is asking.

706jim

Dave Angelini said:

Sounds like my recent client who had back surgery and now his wife is lifting batteries while I watch her on a camera. Same thing though, someone else's work and parallel strings. One string had a SG cell reading 0.00. The genset has been running every morning for 2 months and she found me through the Outback Forum.

 Her hubby is on Vicodin and morphine and really can't say much that I understand. No Probono for me ever, unless they come do summer work for me first. Below is a picture of another one a installer friend sent me of a power system he turned down.

Not shown and really close is a genset and 400 gallons of Diesel. He claimed to have got his permit by parking a tractor in front of the door. Notice the inverters are on! Just needs TLC right?

Well, I've seen wiring and I've seen wiring but this "system" has to take the cake!

icarus

Dave,

A lot of the Marine battery A/B switches i see are rated for 48VDC systems, up to 250 amps.  Wondering if the 65vdc eq volatage would really be an issue.  In my (admittedly limited) experience, I would be more concerned about switch (and passing) high(er) currents than higher voltage.  The 48VDC limit may simply be a liability thing.  

https://www.amazon.com/Perko-8501DP-Marine-Battery-Selector/dp/B00144B6AE/ref=sr_1_9?dchild=1&gclid=EAIaIQobChMI8463_J_C7wIVEwutBh3GyQzmEAAYASAAEgIJvPD_BwE&hvadid=153739150036&hvdev=c&hvlocphy=9033357&hvnetw=g&hvqmt=e&hvrand=5856006323953663767&hvtargid=kwd-370617341&hydadcr=10060_9464266&keywords=marine+battery+switch&qid=1616359403&sr=8-9


You kinda hardly coulda run a 48 volt system and not exceed 48 vdc.  

T

BB.

Decades ago I asked our UL inspector about voltage ratings... specifically about -48 VDC TELECOM systems.

-48 volt battery that could go over 60 VDC  during charging. Given that UL has access to the data used to List components (the data is not open to public).

He mumbled, and approved  48 VDC components for use on our system rated for 72 VDC max.

 

There are safety factors... Such as 1,800 VAC might tests for 600 VAC rated insulation and 120/240 VAC nominal hipot testing for all of our AC appliances.

For the marine switches,  I am more concerned about switching under load.

Sustained arcs are easy with simple moving contacts (marine switch) vs a snap action contacts used in breakers and disconnects.

Xantrex had a simple twist switch for DC disconnects in their GT inverters. Later units had very heavy duty/physically large snap action switches in their newer systems.

I would suggest that the marine battery switches moved when no/low current flow.

Even then, not sure all can carry rated current. In times past a fair number of complaints about overheating or burned contacts.

Bill

Dave Angelini

I have used marine A or B or both rated at 300A continuous, but they do not last at 65vdc even when switching under small load. There still is a quite a surge when you do this at 55v BTW! Try it in a dark room!  This is for EQ at 65VDC max for Surrette and Trojan. Certainly not UL listed or CSA for use at an insured offgrid home !

I will search for Cantrell, thanks Bill.  I really need something less expensive than the midnite, but that will last like the midnite with it's rating. Might just have to tell them to use that or similar from outback or Schneider.

If anyone see's something like this without the circuit breaker cost, I would appreciate a shout!

BB.

Cantrell was an auto incorrect on my phone.

That was Xantrex/Schneider up grade to the GT 3000 family a decade or so ago. Appeared to be an engineered solution for that product. Not an off the self upgrade (as I recall).

Bill

Dave Angelini

Yes no joy on cantrell.  Oh well, maybe Midnite could build a version of the marine switch that was not made of plastic, had the ratings to last more than a summer and not melt. The ones I have built in a metal box are too much work for me and too expensive for them. 

In most of these stringed battery systems I get called when it hits the fan as above. Money is an issue and in these times I want to do more than say no.  I will send an email to Bob at midnite later today. My installer buds are just doing the double breaker in my picture above.

Did you see the energy star link I posted about the Whirlpool in this thread Bill ?

icarus

Maybe a stupid question, but why would one (absent an emergency shut down) need to switch things under high current loads?  Is it no possible to reduce the loading (either charge current or load current) to the point of small current to protect the switch gear?    The load current should be pretty easy, by simply reducing demand.  The charge current might be more problematic depending on how a PV array is wired however.

Tony

BB.

Bluesea (and others) have been having a decade or more of failures (just breaking in 1/2, etc.)... 2010 and 2017 dates seem to come up a lot... In 2017, Bluesea released a much more robust construction--Physically stronger, but still no "snap action" on the switching itself:

https://www.tidewaterowners.com/warning-blue-seas-m-series-switches-are-failing-t2255.html

The UL testing sounds pretty good:

https://www.bluesea.com/support/reference/119/UL_Marine_Listed_Battery_Switch_Testing

I know that Canada had the numbers on the tags... Just was not sure if there was a website for looking up brands/models. Your Whirlpool tag *US and Canada numbers):

https://images.homedepot-static.com/catalog/pdfImages/68/687a4db7-4c9b-4fc3-a238-dd8550c3aef3.pdf

I agree that the US numbers tend to be on the high side--In my moderate climate, I usually use less energy than the US tag has.

I agree that linear compressors do not seem to save power on US models... Perhaps the "other countries" models are also smaller refrigerators which seems to use less than the full size US fridges. I guess surge current is probably less--But have not seen/measured a "real unit" to see what (for example) the defrost heaters draw (around 500-600 Watts draw for heaters that I have seen).

The "upgraded" Xantrex GT 3kw inverter snap action switch covered both DC panel and AC power--And was an impressive sight when compared with the original "twist to turn off" switch on the bottom on the older units. Xantrex/Schneider had to add a bottom section the unit to even fit the new and improved switch gear.

From this post: https://forum.solar-electric.com/discussion/comment/393116#Comment_393116

Notice the large handle to take the increased torque required.

-Bill

icarus

Meanwhile...back at the ranch..Anybody else see anything wrong with this wiring scheme?  The RED in the upper right corner is the positive lead to the inverter/controller/charger.  The long diagonal RED is the negative.  MY hunch is this might be a large part of the short lived battery issue.  

So much to do...so little time.

T

Dave Angelini

It hurts my head to see this kind of thing. Corroded terminals. Start fresh and rewire. At least in the system with the diesel soot everywhere, and the tractor blocking the door you knew that bad things were happening 

As for your question on "stupid". A system needs to be designed and implemented in a way that meets code. The meeting of code generally means that anyone should be able to disconnect a battery and the disconnect needs to be listed. Knowing about reducing the loads first can't be guaranteed. Using a plastic battery switch for Marine is just not going to fly, very long!  It is a good idea but not really a safe thing for me in state with more lawyers than firefighters.

Will look at the links tonight! Thanks Bill !

BB.

Wonder if the batteries were topped up/too full at times... During EQ/heavy charging, heating (expanding electrolyte) and gassing (displacing electrolyte) was causing batteries to spill electrolyte and corrode the terminals.

Some cables are longer than others--Could be imbalanced charging to a degree... But don't see any "red flags" (other than corrosion/possible contamination of electrolyte while servicing while popping caps).

If you (Tony) wire per your post #6, you can replace fuses with breakers (rather than trying a marine A/B/All/Off switch).

You have a wide variety of 2 volt battery solutions for a single string (NAWS, our host's site):

https://www.solar-electric.com/residential/batteries-battery-storage/deep-cycle-batteries.html?nav_battery_voltage=387

Would AGM be a good solution here (little/no maintenance, should be good for 5-7 years of normal service)?))? FLA can last longer (equivalent good quality batteries--AGM catalysts do wear out during charging/gassing/low voltage EQ events).

Of course, until you nail down electrical loads (and hopefully some serious efficiency upgrades)--Sizing is just a guess at this point.

How much maintenance will the owner perform (or outsourced to local guy)?

Does he have any tools on site (Hydrometer, AC+DC Current Clamp DMM, knowledge how to use, log book, etc.)?

-Bill

icarus

I hear you about the switch...

I’m going to have to convince the owner to make some changes.  (May not be an easy task) 

Actually, pretty easy to rewire.  Clean up all the terminals, add a negative and positive bus bar/terminal, make some home run cables, colour the properly, and add three BlueSea series fuses.  (Begs the next question, with ~150amps of potential charge current, do I fuse each series to say 150 amps, or each series to say 50 amps?)

T

BB.

My first suggestion... If 3 strings, drop 1 string (2 strings operational) @ maximum charging and/or inverter draw current * assume 65%/35% worst case sharing * 1.25 minimum NEC derate. Say 150 amps charging (optimistic) for the 1,200 AH battery bank.

The below is my though processes--Ignore if does not make sense/help you...

• 150 Amps / 2 strings (assume one string not working) = 75 Amps if sharing
• 150 Amps * 0.65 unbalance sharing = 97.5 amps worst case string
• 97.5 amps worst case * 1.25 NEC derating = 122 Amps derated branch wiring/breaker rating (120 amps+)
• Typical max supported surge current for FLA around 2.5 hours derate:
• 400 AH battery string / 2.5 hour surge = 160 Amps per string... So max "useful" current around 160 Amps surge
• 3,000 Watt inverter * 1/0.85 inverter eff * 1/21.0 volts LVD = 168 Amp max draw for inverter
• Typical FLA max continuous draw (5 hours worst case support):
• 400 AH / 5 hour discharge = 80 Amps per battery string "practical draw"...
• 168 amp max draw * 0.65 sharing (2 strings unbalanced) = 109 amps
• 80 amps max practical draw * 1.25 NEC derate = 100 Amps per 400 AH battery string (long term, not surge)

Sorry for all of the above math--Just looking for "worst case" things happening and the 1.25 NEC derate for breakers (breakers will not trip at 80% rated load, will trip at 100% rated load--minutes/or hours)... So, I would use 100-120 Amps per string as a "comfortable" not causing any false trips/etc. down the road unless something really goes wrong.

When "looking for problems" with paralleled cable/connector runs... One would thing that "higher resistance" run (longer wire, dirty/loose electrical connection, less than perfect connector) would be the "problem" string...

From my experience with high end test equipment that used "parallel +/-" DC power runs between boards--It was the "low resistance" path that "failed first" (typically burned connector pins/sockets). I could see the wiring "unzip" down the 5+ shared paths (one failure, things still worked--Getting out to 3 or 4, too much voltage drop affected operations--just an example).

When you look at one of the power equations-- Power = Current^2 * Resistance -- As an example, a "low resistance path" carrying 2x more current is 2^2=4x more heating (cable/connections)--And it was those low resistance paths that failed first.

When you get to AGM and Li Ion batteries--Their own internal resistance is quite low--And what "steers the current" is the cable and connection resistance (typically)... And why cable lengths and solid connector/bolt up connections are critically. And confirm with a DC Current Clamp meter under heavy load/charging, that the strings are reasonably sharing current (what is reasonable--Your guess is as good as mine--But one string carrying 2x more than "shared current" would be a concern for me (look for bad connections, bad wire/crimps, etc. on the "low current" cables).

Using the conservative NEC wire ampacity chart:
https://lugsdirect.com/WireCurrentAmpacitiesNEC-Table-301-16.htm

You could justify 4 AWG (THHN @ 85 Amps) to 2 AWG pretty nicely for 3x battery parallel interconnects (if you go with a single string, then simply 1.25x NEC derating for wiring and battery breaker/fuse.... Then there is voltage drop... For a 12/24/48 volt system, I would suggest the max drop be 0.5/1.0/2.0 volt drop respectively for loads...

And for charging, 0.05-0.10/0.1-0.2/0.2-0.4 for 12/24/48 volt wiring drop max (for accurate charging voltage measurements back to the charge controllers

If 150 amp charging * 1.25 NEC derate = 188 Amps

From NEC table, at least 2/0 Wiring (THHN @ 195 Amps)

Say 24 volts @ 150 amps charging @ 0.2 volt max drop @ 5 foot cable (one way run for this drop calculator):
https://www.calculator.net/voltage-drop-calculator.html?material=copper&wiresize=0.2557&voltage=24&phase=ac&noofconductor=1&distance=5&distanceunit=feet&amperes=150&x=0&y=0

Result

Voltage drop: 0.12
Voltage drop percentage: 0.49%
Voltage at the end: 23.88

That 0.12 volt drop for charging current back to charge controller (at least for the battery to battery bus cable) is just about right... Need to calculate the total drop from battery to bus common through battery fuse to charge controller fuse/breaker to charge controller is max 0.2 volts suggested drop...

Anyway... That how it looks from a 1,000 miles away. Pretty conservative numbers--And if the Charging current is <150 Amps, or goes to 48 volts or single string of 1,200 AH (or whatever) 2 volt batteries--That will all change.

Engineers like "conservative" (hopefully reduces the chances of something going wrong down the road. Marketing and Customers don't want to pay "too much money"... Balancing act (i.e., cost of fuses and holders + switches if needed vs cost of a breaker that does both fusing and switching functions).

-Bill "way too much information" B.

Dave Angelini

The DC switch in your GT300 looks alot like the Schneider Square D I use for the 600v mppt. Too bad it can't handle 250 adc

The Whirlpool has in my reference been sold as Maytag and Amana and ?. It is a lady pleaser offgrid and when I tell them they can use it, I get the are you married question.      Rock solid performance for 10+ years.

Tony,  you know this but make the interconnects the same length on your list. I only use 2/0 or up. It makes changes (go to 48v)  down the line much easier. Also by doing this I end up with a large supply of cables for repairs/installs later. Same with fuse /breakers, 175a or 250a.