Battery Banks and Switches

[Tom.B]

I am planning an upgrade to our battery system (again). I am moving the switches and fuses to the main cabin so they are accessible in the cabin without requiring a trip into the engine room to operate. Currently, our house bank consists of three boxes with a total of eight batteries. One box with four and two with two each. They are a single bank that is hard-wired to a buss bar before sending to the breaker panel. They are not fused or switched at this point, but that is part of what I want to add. I understand the need for the fuses (protect wires... yea, I know I should have done this last time). I am wondering if there a need to switch off each box independently? Additionally, with a 275A switch (a BEP 701) be enough for the 4-battery box? I am thinking it is, but thought I might want to ask... just in case I am missing something. Thanks y'all.

 

[O C Diver]

IMO, if each battery box has an internal fuse, I don't think each box needs to be switched. My rational is that theoretically you could have all the batteries in one box with one set of cables. In my situation, my 900 AH bank is all in one box. If I had gone to 1,300 AH, it would be 2 boxes. With a bank that large, I would probably have had an A, B, and Both switch, but not at 900 AH.

Ted

 

[CharlieJ]

To be compliant, you do need a switch between the house bank collecting bus and the main panel circuit breaker.

For fusing, I would suggest Blue Sea Marine Rated Battery Fuses (MRBF) and their companion fuse holders. Don't cost much and provide superb protection.

Here is the rule: Every B+ conductor on a compliant boat, except for engine starting circuits, must have over-current protection.

 

[psneeld]

If referring to the battery post Blue Seas fuses...make sure your battery bank can't exceed the new ABYC interrupt (think it's that) rating.

 

[CharlieJ]

@psneeld
Good advice but the MRBF's have an Ampere Interrupt Capacity (AIC) of 10,000A @ 14VDC so he should be ok if his battery short circuit capacity is below that.

 

[Frosty]

I can't see what type of batteries you have so this may or may not apply. I have Lifeline AGM, and for my bank of three Group 31's, I had to go to a Class T fuse in order to meet AIC (Ampere Interrupt Capacity). The Blue Sea MRBF (which I like when I can use them) is 10,000 amps, but the short circuit amps of my 375ah bank is 12,771 amps (4,257 amps each). Class T AIC is around 20,000 amps.

(AIC is basically what the infrastructure of the fuse style can take without just failing on its own, regardless of fuse size.)

One thing I didn't consider at the time (I have since seen this on LiFePo banks) is fusing each battery individually. I'd have to think about it a bit more, but if that would work I could have used 3 MRBF, one on each battery in the bank.

Here is a chart for Lifeline AGM short circuit amps. They do say it can be +/- 35%. (It will come into focus if you click to enlarge.)

 

[CharlieJ]

@Frosty

One thing I didn't consider at the time (I have since seen this on LiFePo banks) is fusing each battery individually. I'd have to think about it a bit more, but if that would work I could have used 3 MRBF, one on each battery in the bank.

And that is exactly what the OP should do. The output of the combined capacity of the bank that is formed at the buss then has to be protected appropriately. A Class T fuse will do that.

This is all laid out in ABYC E-11.

 

[Frosty]

So you are saying that even if the batteries are fused individually, one still needs the same fuse for the group that you would have if you didn't fuse them individually? I guess I had better brush up (or maybe I never understood it correctly).

Not to hijack (I hope), but taking my bank as example.

Three AGM at 125ah and ~4,300 amps of short circuit possibility each, batts in parallel. I have a single, 225a Class T fuse on the single outgoing positive 2/0 cable, then a Blue Sea main battery switch, then on to bus bars. Wherever the cables drop to a smaller size from the positive bus bar, I re-OCP accordingly.

*************

But let's say I instead decided to fuse each battery individually. Let's say I put a 225a MRBF (or 250 if they don't come in 225) on each of the three battery positive terminals. Individually, MRBF has an ample AIC (10,000a to cover around 4,600a), so that's good.

But where I'm confused in thinking this through is the need you mentioned to still have the Class T fuse at the exit point for the bank of three, now-individually-fused batteries on the way to the switch. In my mind the individual fuses would supplant the single Class T. I must be missing something and it's annoying to not be able to think of it! Granted, I have only been pondering for fifteen minutes, and I would do more research and thinking if it were something I were implementing myself.

I'll probably be rolling my eyes at myself once it's pointed out to me, because it will be obvious.

 

[CharlieJ]

@Frosty
The ABYC Standards (and ISO Standards and the NEC for that matter) establish "rules". With these "rules" come exceptions with the most conservative approach being to build to the "rules" and only use the exceptions when absolutely needed. But your comments, so far, certainly imply that you understand this.

So you are saying that even if the batteries are fused individually, one still needs the same fuse for the group that you would have if you didn't fuse them individually? I guess I had better brush up (or maybe I never understood it correctly).

Short answer is "yes". Here is my logic.
The AIC rating of a fuse is based on:
1) The fusible element must melt as required by its designed time/current curve.
2) When the fusible element melts, the arc that is created must extinguish as it is a conductive plasma.
3) The fuse can not physically blow apart.

So for the AIC to really come into play, the protected circuit must experience a bolted fault. This is generally a very, very, very rare occurrence.

From a circuit design POV, for DC systems, voltage drop is generally the constraining variable that we have to design to. Conductor ampacity generally follows. In other words, if the voltage drop is accounted for the conductor ampacity will usually be sufficient.

For a real world design, your three Lifeline AGMs have a short circuit current of 4257A. The design should account for the voltage drop between the three batteries and the B+ and B- collecting bus bars.

An example design:
Battery whips. For this example the B+/B- circuit length from each battery to the bus bars is 6 feet. Most DC panelboards have a 100A or 50A main breaker. Assume 100A. So each battery will have to produce 100A/3 = 33 A for the DC panelboard. I would probably spec AWG 2/0 for the battery whips. Engine room ampacity is 280A and the 6 foot circuit to feed the bus bars would yield an IR drop of <<3%.

Panelboard Feed. AWG 1/0 will provide an IR drop of 3% to 30' of circuit length. AWG 2/0 will provide an IR drop of 3% to 44' of circuit length. For the example I will use 2/0 with an engine room ampacity of 281A.

Inverter/Charger. A 3kW inverter/charger is fed from the same collecting bus bars. A Victron Multi-plus calls for a 400A fuse and 100mm^2 conductors for up to 5m of circuit. 3000W/12.5VDC = 240A (assuming 100% efficiency). The example Multi is fed by 4/0 B+ and B- with an engine compartment ampacity of 378A.

Full Load. At full load the three batteries are going to have to provide 240A for the I/C and 100A for the DC panelboard for a total of 340A. Each battery must provide 340A/3 = 113A. The IR drop for the whips is still << 3%.

Whip OCPD. Finally, and I apologize for the length of this lecture, I would protect each battery B+ whip (250A MRBF to protect the AWG 2/0 whips (E/R ampacity = 280A) from the individual battery to the collecting buss.

Panelboard Feed OCPD. I would protect the DC panel board feed (AWG 2/0) with a 250A MRBF on the collecting bus. My rationale being that the total SSC of the three batteries, at 12.9VDC is 12.7kA. The AIC for the MRBF is 10kA at 14VDC. There is some headroom there as the nature of AIC is that it diminishes with rising voltage. If I was super conservative or designing for a USCG Sub-chapter T boat, I would spec a Class T fuse and fuseholder.

Inverter/charger feed. I would protect the B+ 4/0 feed (E/R ampacity 378A) to the I/C with a 400A Class T fuse and fuseholder. This last is permitted by the ABYC Standard as a Class T 378A fuse is not available. The Class T AIC is 20,000A @125VDC so there is a lot of headroom with this fuse.

The above design is conservative, meets the spirit and the letter of the "rules" and I can sleep well at night after installing such a system. The exceptions to the ABYC "rules" are attached. Clearly, Exception #2 could be used to eliminate the battery whip MRBFs. It is my opinion that the cost of the MRBFs is so low and the added protection is very worthwhile.

So the short answer is "no" if you want to use the exception that is clearly laid out in the "rules". I just choose not to.

 

[Frosty]

I'm going to have to read through all you wrote carefully. Some of it doesn't apply perhaps because I don't have an inverter or an inverter/charger. My system is likely much smaller and simpler than most here because I don't have a trawler (Covid interrupted my shopping).

A couple of things to clarify:

-- I'm not the type to try to "get away" with anything. I like ABYC as it gives me a written guide from which to design a good system. Not looking for exceptions (heck, I have my starting circuit OCP'ed).

-- I am not planning to fuse with the three individual MRBF's, because the Class T does cover my ~12,500 amps of short-circuit-ability. I was just thinking out loud about whether one could, in order to stay safely within AIC - since many here have much larger systems, maybe too many short circuit amps for a single Class T.

One thing I found disturbing when I was considering LiFePO for my house bank, is that the mfgrs I was looking at were not able to provide any short-circuit amp numbers. I know LFP is probably high, but how do you plan without numbers? I like that Lifeline publishes them.

OP: I'm sorry this ended up barging in on your thread; that was not my original intention.

 

[CharlieJ]

@Frosty #10

One thing I found disturbing when I was considering LiFePO for my house bank, is that the mfgrs I was looking at were not able to provide any short-circuit amp numbers. I know LFP is probably high, but how do you plan without numbers? I like that Lifeline publishes them.

All LFP battery manufacturers publish the internal resistance of their cells. It is a marketing point. Apply Ohm's Law and calculate short circuit current.

 

[Tom.B]

Okay thanks yall (for the record, I totally forgot I posted this and haven’t checked here in a few days... sorry)

Okay, so I will fuse each box of batteries and not use individual switches. There is a “master bank” switch. A 500A 1/2/off/both switch that routes the bank to the appropriate location, so I have that covered. Honestly, I assumed this was correct. I haven’t figured a reason to switch them separately. Besides, it will save me a bit of money anyway... and space is at a premium.

 

[Frosty]

@Frosty #10



All LFP battery manufacturers publish the internal resistance of their cells. It is a marketing point. Apply Ohm's Law and calculate short circuit current.

Thanks for that! It's interesting that it's so simple because when I contacted a well-known company's tech support, they were not able to provide a figure (which I wanted for AIC purposes).

Given that this season is kind of a bust (and who knows about the future ), I've decided to stick with Lifelines for now. Why buy expensive new batteries and let them sit. But I imagine I will reconsider LFP in the future. Being able to calculate the potential short-circuit current myself will be a plus, so I appreciate the info.