AIC and house bank fuse danger

[NoRain]

So far this is looking like the best solution for me without reconstructing my setup.

Sorry a bit late to respond, but I didn’t see it mentioned since (maybe I missed something?). Also not sure my Reply function is working correctly here, but this is in response to your Post #18:

Where did these images come from? 800A fuse is not acceptable to protect 4/0 wiring.

 

[Mac2]

Sorry a bit late to respond, but I didn’t see it mentioned since (maybe I missed something?). Also not sure my Reply function is working correctly here, but this is in response to your Post #18:

Where did these images come from? 800A fuse is not acceptable to protect 4/0 wiring.

Chat gbt

 

[Mac2]

Sorry a bit late to respond, but I didn’t see it mentioned since (maybe I missed something?). Also not sure my Reply function is working correctly here, but this is in response to your Post #18:

Where did these images come from? 800A fuse is not acceptable to protect 4/0 wiring.

Fuse is within 12 inches of house bank.

 

[Frosty]

The "thing" that it's protecting against are fuses (or breakers) that fail to open in a bolted short. That's the cause of the problem. If they fail to open then a metric s-ton of energy will dump into the short, be it a parallel battery with an internal short, a wrench dropped across two terminals, or any other fault. That's the consequence and it's likely very bad.

And yet I never saw this recommendation before LFP batteries. And if I had dropped a wrench across my big AGM bank I'm sure it would not have been pretty. They each had like 4,500 amps of short circuit current (so I had a class T even then, but just one for the bank, which also protected the downstream wire (so it had the AIC plus amp rating for the wire size).

@Mac2 I edited this just as you were "liking" it, so things may not match now. At first I read @twistedtree as explaining why we have "normal" downstream fuses, when "the thing" from my thread was specifically why do we now fuse individual batteries IN ADDITION to the bank as we always have.

 

[Mac2]

Thoughts? I know this doesn’t cover the individual battery “thing”.

 

[BDofMSP]

The approach on the left will subject the two batteries to different loads and discharge rates, and each will have to be individually charged. Note sure if that's your intent.

In the approach on the right, there is a segment as marked up that is subject to 900A and is likely overloaded at that rate.

View attachment 172130

Regarding the model on the left, yes I get that there are (probably unacceptable) limitations. But I wanted to make sure I really understood what was being said. Regarding the right, I corrected myself that the 4/0 cable wouldn't be there (although I haven't looked up the capacity of the connector bars that the Pro Installer system uses). But the 650 amp buss bar is undersized. And it's the biggest one they make in that series I believe. So now we're back to some other "roll your own" solution I guess.

 

[BDofMSP]

Regarding the model on the left, yes I get that there are (probably unacceptable) limitations. But I wanted to make sure I really understood what was being said. Regarding the right, I corrected myself that the 4/0 cable wouldn't be there (although I haven't looked up the capacity of the connector bars that the Pro Installer system uses). But the 650 amp buss bar is undersized. And it's the biggest one they make in that series I believe. So now we're back to some other "roll your own" solution I guess.

The connector bars are 650 amp capacity, but can be doubled for higher loads. The bus bar is the limitation. What are others using to combine their banks? I see Blue Sea makes a 1000 amp power bar but not sure how well it would align with the BEP components.

 

[NoRain]

Sorry a bit late to respond, but I didn’t see it mentioned since (maybe I missed something?). Also not sure my Reply function is working correctly here, but this is in response to your Post #18:

Where did these images come from? 800A fuse is not acceptable to protect 4/0 wiring.

Chat gbt

Yeah, that’s what I was afraid of. It is wrong, 800A fuse is not acceptable to protect 4/0 wiring.

Fuse is within 12 inches of house bank.

That’s not relevant, technically. And BTW to the letter of the standard it should be within 7”. I know that’s an almost impossible task in many cases, that’s why there is an exception possible to extend that distance if the wire is “enclosed”. Like with split loom. Just be aware that enclosing de-rates the capacity of the wire, so you’ll need to take that into account.

So much for tech specs - in a practical case I wouldn’t hesitate to do a 12” run if it is clear from obstructions ie no friction points, and/or is strapped tightly so it cannot move and chafe. Personal opinion only, and I will live with the “risk”. However a surveyor if/when you go to sell or re-insure may not accept that and quote the ABYC spec at you. Just as long as you are aware.

 

[drewread]

Chat gbt

I've seen this a number of times lately.. Is is not ChatGPT?

 

[NoRain]

Quotes got screwed up….that was Mac2’s response.

 

[ksanders]

For a neater implementation of the multiplicity of class T fuses that E-11 seems to favor, take a look at the BEP Pro Installer offerings. They have thought this through reasonably well, with all components dimensioned in modular blocks so they pack tightly, and premade bus bars to hook it all together.

Thanks!!!

The BEP Pro Installer is PERFECT!

This is something folks should really be looking at.

 

[Mac2]

Is it realistic to have a fuse to protect for AIC and a second fuse to protect the wire to the inverter from the house bank (leave the thing that could happen at the battery csbles out of this equation). My inverter will put out 300 amps max. Im trying to tie in a 1,200 amp house bank to my 4 kw inverter.

 

[NoRain]

Your terminology is a bit confusing, but if I understand the question: 1) you cannot leave the wires out of the equation, they are what determines the *size* of the fuse. 2) So once you’ve determined the size of the fuse, you select the Classification of the fuse….Class T in this discussion. The class of the fuse is determined by the AIC rating that is determined by your particular installation.

You are 24V right? So a 4KW inverter can pull 167A max (more for a brief period as per its surge ratings). You have run 4/0 wire so you’re good for say 350A continuous (several variable here, but let’s just say). So you use a 400A fuse. Because you’re connected to a large lithium bank you will need that fuse to have a large AIC rating. Many posters above have tried to explain this and the options available. But again, say your particular config requires an AIC rating for your fuse to be 50,000A.

Therefore you would need an 400A, Class T fuse with an AIC rating of 50kA.

 

[Nick F]

Is it realistic to have a fuse to protect for AIC and a second fuse to protect the wire to the inverter from the house bank (leave the thing that could happen at the battery csbles out of this equation). My inverter will put out 300 amps max. Im trying to tie in a 1,200 amp house bank to my 4 kw inverter.

I was wondering the same thing!

Put a 400A fuse with a low AIC rating (say 3000A) to protect the wire and, separately, put a (say) 3000A fuse in series which has a (say) 50kA AIC capacity. If the fault current is <3000A the first fuse protects and, if the fault is beyond the AIC of the first fuse (i.e. >3000A), the second fuse will protect. This seems like an elegant way to cover the protection. Note that the second fuse could be in the common supply line to many loads.

Comments?

P.S. I now see that this is shown in Post #65.

 

[NoRain]

Yeah, why? Why 2 fuses?

 

[DDW]

I think the presumption is that with a 40KA short circuit fault, the 400A fuse arcs and welds itself closed, then the 3000A fuse blows. Much easier to find 3000A fuses in a high AIC rating.

My presumption is that I am much more likely to be struck by lightening (with the commensurate 500,000A surge), than to have this fault occur. There is quite a long discussion about this over on the DIY solar forum, with very little evidence that it is a problem, or more of a problem than in the lead acid world where we didn't do much hand wringing about it.

 

[Nick F]

Yeah, why? Why 2 fuses?

Valid question!

One less expensive fuse (or breaker) to protect the wire for the individual load and the second (expensive high AIC type - one for the whole boat) to protect everything fed by the main battery.

 

[NoRain]

Sorry, but no. Whatever is upstream from that 3000A fuse needs to be able to take 3000A of current. Like the wire connected to it.

The diagram in post #65 is misleading. Note it shows bus bars on either side of the Class T. So in that diagram alone those had better be 1000A rated bus bars. Secondly, how are the batteries connected to the bus bar? With wire, that’s rated at 1000A? Or is the bus bar somehow bolted right to the battery terminals. And what size bolts are connecting it; M6, M8? Don’t know off the top of my head and don’t have time to look it up, but I’d bet they are not rated for 1000A. So there’s that.

And this discussion is using 3000A as its number? I’ve seen 1000A busbar (Blue Sea makes them) but I’ve not heard of a 3000A one in the marine world. Maybe/probably in the industrial world, but that’s a whole ‘nother thing.

I’m experiencing a nagging feeling that I’m missing some fundamental element of this discussion, because what I’m saying is not novel or radical. It’s pretty well established science and practice.

 

[Marco Flamingo]

I think the presumption is that with a 40KA short circuit fault, the 400A fuse arcs and welds itself closed, then the 3000A fuse blows. Much easier to find 3000A fuses in a high AIC rating.

I thought that only breakers have the possibility of welding themselves closed. That's why the high AIC DC breakers have such strong switch springs. I had to bolt mine to the bulkhead.

With a first low AIC 400A fuse, the fusable link melts but the high amperage arcs over the gap (that's why the fuse has a low AIC). The arc is still passing amps. Question is how much amperage? We don't know exactly what the battery is producing and what the upstream short is drawing. I would assume that the arc, in generating a unknown amount of heat and light, is using some unknown amperage and passing some unknown amount. Let's say the arc is still passing 3,000 amps. The second fuse (3,000A with the higher AIC) won't blow, but something will. Probably the insulation on the cable or the equipment it is feeding.

How could one compute that the arc will always pass enough amperage for the second 3000A higher AIC fuse to blow?

 

[Nick F]

Sorry, but no. Whatever is upstream from that 3000A fuse needs to be able to take 3000A of current. Like the wire connected to it.

The diagram in post #65 is misleading. Note it shows bus bars on either side of the Class T. So in that diagram alone those had better be 1000A rated bus bars. Secondly, how are the batteries connected to the bus bar? With wire, that’s rated at 1000A? Or is the bus bar somehow bolted right to the battery terminals. And what size bolts are connecting it; M6, M8? Don’t know off the top of my head and don’t have time to look it up, but I’d bet they are not rated for 1000A. So there’s that.

And this discussion is using 3000A as its number? I’ve seen 1000A busbar (Blue Sea makes them) but I’ve not heard of a 3000A one in the marine world. Maybe/probably in the industrial world, but that’s a whole ‘nother thing.

I’m experiencing a nagging feeling that I’m missing some fundamental element of this discussion, because what I’m saying is not novel or radical. It’s pretty well established science and practice.

I believe that the fact that you are missing is that the very high short circuit current is cut off very quickly by the high AIC fuse. Because of this, the wiring upstream of it does not need to have a CONTINUOUS rating equal to the high AIC fuse's rating.

 

[DDW]

The whole premise of E-11 and the high AIC requirement is that all of the stuff upstream and downstream can withstand 40KA for at least a little while. And that the fuse would be the first thing to blow. And that the 300A fuse will arc and weld. And that this is the most likely thing to start a fire.

I'd say none of those assumptions are generally true. If something else in the circuit cannot withstand 3000A, then you will never get to 40KA.

In my experience, electric fires usually start because a short occurs that draws less than the amount required to blow the fuse. With 400A fuses you can generate all kinds of heat. A 400A class T will pass 600 indefinitely, at 12V that is 7000 watts, considerably more than a stove burner. None of this whole discussion contemplates that possibility. It is the reason GFI type breakers were developed.

If you believe the premises, then the failure cascade will be as I said: 300A fuse welds, 3000A fuse blows.

As an aside, E-11 does not acknowledge the presence of a BMS or that it will cut off the current. I know from experiments that at least on smaller batteries, it will. We worried about the effect of a short circuit on the batteries we use for instruments in sailplanes, these are typically 12 - 20 AH AGMs, now mostly replaced by LFP. We took examples of each and dropped them, terminal side down, on a heavy copper plate. On the LFP, no visible effect, BMS just disconnects. On an AGM it can be somewhat dramatic.

 

[NoRain]

I think the presumption is that with a 40KA short circuit fault, the 400A fuse arcs and welds itself closed, then the 3000A fuse blows. Much easier to find 3000A fuses in a high AIC rating.

But the whole point of a Class T fuse is so that the 400A fuse *does not* weld itself closed.

 

[Barking Sands]

So, acknowledging that there is far more complexity than is displayed here (charging, inverter, etc.) my take away is that for my two-bank (single battery each) solution, these are my options. One, sacrifice redundancy and shared capacity and keep each bank isolated and in exchange gain simplicity, or two, combine them and add additional protection. See attached. Is that correct?

Edit: the 4/0 cables would not be required in the right side image between the buss and fuse and switch, since these are all Pro Installer series. The connect directly.

The battery that was placed in this diagram was the Epoch V2-T. How about just changing out the included 400A Class T for a 225A Class T inside the Epoch V2-T under the user serviceable window. The Epoch 12V 460AH is rated at 230A continuous anyways. And you can probably eliminate the second cut off switch. I would put that cut off switch on the outlet to the inverter.

 

[Delta Riverat]

When the 400A fuse "welds itself closed" the 3000A fuse will just sit there and get warm while the fire starts.

 

[NoRain]

I believe that the fact that you are missing is that the very high short circuit current is cut off very quickly by the high AIC fuse. Because of this, the wiring upstream of it does not need to have a CONTINUOUS rating equal to the high AIC fuse's rating.

No I hear what you’re saying but disagree. You don’t get to choose the size of the disaster you may encounter. That’s why there are standards and practices.

Said another way, you need to protect the wire from a continuous current flow that it is not designed to carry. Not just an instantaneous short circuit current load. In the example being tossed around here, if you had a partial short, like a partially chafed through 4/0 that could still carry 2000A, you are hooped. Your 3000A fuse will not blow, and that 4/0 (or what’s left of it) will glow cherry and you got yourself a fire.

But why in the world would you use two fuses to do the job that one properly spec’d fuse can and will do. This adds additional connections and unnecessary complexity for no good reason. And why are the 400A fuses in this discussion “welding”? A 400A Class T (BlueSea) with its AIC of 20kA (actually higher at 12 or 24V) will blow without welding or plasma arcing if you get a full short. I don’t think it is appropriate to be using anecdotal “evidence” in this kind of discussion. Like, “I’ve never seen this happen” or “I know a guy that did this for years and nothing bad happened” or “in my experience what you are describing never happens, it’s this other thing”. Rules and standards are based on many things and one biggie is the laws physics. A wire of a given gauge and construction (etc etc more specs) can handle x amount of current for y amount of time and so on. Nowhere in there is an option for you to say “well I did this in my installation so that can’t happen”. Many disasters DO happen for these unseen reasons and that’s why the safety mechanisms that are designed to prevent the catastrophes are spec’d the way they are.

 

[BDofMSP]

The battery that was placed in this diagram was the Epoch V2-T. How about just changing out the included 400A Class T for a 225A Class T inside the Epoch V2-T under the user serviceable window. The Epoch 12V 460AH is rated at 230A continuous anyways. And you can probably eliminate the second cut off switch. I would put that cut off switch on the outlet to the inverter.


View attachment 172152




View attachment 172151

Thank you. I was aware of that fuse but wasn't sure it met the ABYC Standards. I was under the impression that I needed to protect against the whole 460ah, but of course that's capacity not current. That works great!
In that case the 650a bus bars is fine too, not to mention the 4/0 cable. I was wondering how people were managing 920a but they're not - it's 460.
I have a full diagram with the inverter, Orion's, Cerbo, etc. I'll update that now. Thanks again.
BD

 

[twistedtree]

And yet I never saw this recommendation before LFP batteries. And if I had dropped a wrench across my big AGM bank I'm sure it would not have been pretty. They each had like 4,500 amps of short circuit current (so I had a class T even then, but just one for the bank, which also protected the downstream wire (so it had the AIC plus amp rating for the wire size).

@Mac2 I edited this just as you were "liking" it, so things may not match now. At first I read @twistedtree as explaining why we have "normal" downstream fuses, when "the thing" from my thread was specifically why do we now fuse individual batteries IN ADDITION to the bank as we always have.

The requirement was baked into the AIC requirements and was sufficient for flooded and AGM lead batteries. 20kA covers you. But it's not sufficient for Lithium Ion, and apparently TPPL lead batteries (I don't know any details on them). That's why it changed - because LPF and TPPL can exceed 20kA. E-13 (the Lithium Ion standard) originally talked about fusing to meet the short circuit current specified by the battery manufacturer, but it was acknowledged that very few battery manufacturers were providing that number. So one guy set out to run some simulations and came up with the 5kA per 100Ah formula for cases where the short circuit current isn't specified.

 

[twistedtree]

Regarding the model on the left, yes I get that there are (probably unacceptable) limitations. But I wanted to make sure I really understood what was being said. Regarding the right, I corrected myself that the 4/0 cable wouldn't be there (although I haven't looked up the capacity of the connector bars that the Pro Installer system uses). But the 650 amp buss bar is undersized. And it's the biggest one they make in that series I believe. So now we're back to some other "roll your own" solution I guess.

Or you could reduce the 450A fused down to 300A.

 

[twistedtree]

The whole premise of E-11 and the high AIC requirement is that all of the stuff upstream and downstream can withstand 40KA for at least a little while. And that the fuse would be the first thing to blow. And that the 300A fuse will arc and weld. And that this is the most likely thing to start a fire.

Exactly. The 300A fuse fails to open and current continues to flow where the circuit expect it to have stopped within the trip curve of the 300A fuse

I'd say none of those assumptions are generally true. If something else in the circuit cannot withstand 3000A, then you will never get to 40KA.

In my experience, electric fires usually start because a short occurs that draws less than the amount required to blow the fuse. With 400A fuses you can generate all kinds of heat. A 400A class T will pass 600 indefinitely, at 12V that is 7000 watts, considerably more than a stove burner. None of this whole discussion contemplates that possibility. It is the reason GFI type breakers were developed.

Yes, but this is why downstream conductors must be able to handle current within the trip curve of the fusing.

I'm not sure about indefinitely, but a Class T fuse can pass 600A for 15-20 minutes before tripping.

I think GFI breakers address different issues specific to AC circuits, though I suppose you could do it for a DC circuit as well.

If you believe the premises, then the failure cascade will be as I said: 300A fuse welds, 3000A fuse blows.

I'm not sure where this 300A followed by a 3000A circuit came from, so can't comment.

As an aside, E-11 does not acknowledge the presence of a BMS or that it will cut off the current. I know from experiments that at least on smaller batteries, it will. We worried about the effect of a short circuit on the batteries we use for instruments in sailplanes, these are typically 12 - 20 AH AGMs, now mostly replaced by LFP. We took examples of each and dropped them, terminal side down, on a heavy copper plate. On the LFP, no visible effect, BMS just disconnects. On an AGM it can be somewhat dramatic.

E-11 and E-13 specific exclude the BMS when considering fusing and max short circuit current. This is because a BMS's design and performance is completely unknown, so can't be relied on for anything in a standard. If at some point there are BMS performance standards, then this aspect of over current protection may very well change. I submitted and had approved a similar change for the required battery disconnect switch. Now a BMS disconnect switch can double for the battery switch under a specific set of circumstances. But a solid state BMS switch is NOT one of them.

 

[SteveK]

The whole premise of E-11 and the high AIC requirement is that all of the stuff upstream and downstream can withstand 40KA for at least a little while. And that the fuse would be the first thing to blow. And that the 300A fuse will arc and weld. And that this is the most likely thing to start a fire.

Exactly. The 300A fuse fails to open and current continues to flow where the circuit expect it to have stopped within the trip curve of the 300A fuse

Is this 300A fuse a separate fuse like MRBF or is it a class T 300A. If it is a 300A MRBF and intended to handle accidental shorts downstream, a cheap remedy to protect the class T which is more costly to replace then I think it is a good idea.
Does not matter if it will fuse on a AIC current as it will be protected by the class T.