LFP Available Short Circuit Current

[CharlieJ]

Since the OCPD ampere interrupt capacity (AIC) has to be matched to the battery available short circuit current (ASCC), does anybody know what the ASCC is for some of the more popular cells: Thundersky; CALB, GBS? How about for the Victron 330Ahr Smart LFP or the Lithionics family of batteries?

[cr180]

Interesting question. Most US battery vendors, Trojan for example, publish available short circuit values. Have researched CALB (thought about building my own bank), and did not find available short circuit values. They do publish battery cell resistance, but when you divide into cell voltage, the current value is small (on the order of 600 to 700 amps for 200 Ah cell). I would think might be prudent to use US numbers which are on the order of 3000 amps for 200 Ah.

[CaptainRivet]

Originally Posted by CharlieJ

Since the OCPD ampere interrupt capacity (AIC) has to be matched to the battery available short circuit current (ASCC), does anybody know what the ASCC is for some of the more popular cells: Thundersky; CALB, GBS? How about for the Victron 330Ahr Smart LFP or the Lithionics family of batteries?

Simply calculate the short curcuit current of your LFP bank:
If you scan the code of your cells you should get the values you need.
I take my bank as example: 16 cells Lishen 272AH, internal resistance 0,12mohm per cell, voltage per cell when full (worst case) 3,65V. restistance of busbars connecting cells 16mohm.

the formula to calculate ASCC od your bank is:
(number of cell x cell voltage) / ((number of cells x internal resistance) + (resistance of connecting busbar) = avaliable short curcuit current


(16x0,012mohm)/((16x0,012mohm)+16mohm)=58V/(0,192mOhm+16mohm)=58V/0,00352ohm=16477,27 A hupsi :-)


This means my main battery bank(!) fuse must withstand 16477A interrupt current.
What you also need to know that the short curcuit current of shortend LFP bank is raising quite slowly and the bank fuse won't interrupt it. The fuse just protects the whole installation upstream by disconnecting it from the bank.
That means your 400A class T or NH fuse will blow when your bank reached 2-3000A after several milliseconds but eg will continue to raise up to 16400A in my case. And these 16400A will try to arc the blown fuse and given the distance over the air from around 7mm of a class T that can physically happen even they have an IR of 25kAa. With 7cm distance of a NH3 fuse thats impossible.
Really surprised Nigel is not taken this into account. As i said i had an arced class T in a car with a shortened LFP bank and burned the car due to this. Since then i am only using NH fuses as battery fuse.

[HeywoodJ]

Originally Posted by CaptainRivet

(16x0,012mohm3.65V)/((16x0,012mohm)+16mohm)=58V/(0,192mOhm+16mohm)=58V/0,003520,016192ohm=16477,273607 A hupsi :-)

Easy to calculate, but the numbers have to be right On top of the typos, CALB and Thundersky, at least, list internal resistance at 0.1-0.5mohm/cell, or an order of magnitude greater than the 0.012 used in the sample. Can’t say I’ve ever seen a cell with 0.012mohm internal resistance, but possibly it exists in a high Ah cell. Also, internal resistance goes up with increasing and decreasing state of charge, a generalization is that lowest resistance is at ~70% SOC, ~3.4V and the short circuit number calculated at that SOC will be 10-20% higher than that calculated at 100% SOC.

[T1 Terry]

Originally Posted by HeywoodJ

Easy to calculate, but the numbers have to be right On top of the typos, CALB and Thundersky, at least, list internal resistance at 0.1-0.5mohm/cell, or an order of magnitude greater than the 0.012 used in the sample. Can’t say I’ve ever seen a cell with 0.012mohm internal resistance, but possibly it exists in a high Ah cell. Also, internal resistance goes up with increasing and decreasing state of charge, a generalization is that lowest resistance is at ~70% SOC, ~3.4V and the short circuit number calculated at that SOC will be 10-20% higher than that calculated at 100% SOC.

First tech info to be sorted, LFP and LYP cells are at 100% SOC at a rested 3.4v, any voltage above that value is surfacecharge held in the electrolyte and has virtually zero capacity.
Now for the blurry bits that turn this into a dark art.

1)The number of times the cell has been pushed outside to comfort zone (2.8v -3.6v) and length of time it remained there the internal resistance, it will increase above any new product value.

2)At best the internal resistance when it leaves the factory is give or take half a brick due to mass production variables.

3) the internal resistance is internal cell temperature dependant ..... the closer to 60*C, the lower the internal resistance.

How do you calculate the internal resistance? Rule of thumb is the difference between unloaded voltage and loaded voltage divided by the load current.

LFP cells throw a spanner in the works here because a fully discharged cell will drop a greater voltage to a given load as will a cell holding a rested voltage greater than 3.4v. The closer to the 3.3v mark rested the lower the voltage drop will be, but the cell must have been saturation charged first, in other words, the cell must hold better than 3.5v after an over night rest.

The higher the C factor ( load current compared to actual capacity) the higher the resistance .... that makes the ASCC virtually impossible to calculate, as the internal temp goes up the lower the internal resistance .... till around 65*C, then it goes up rapidly and a short circuit is the ultimate C value and will generate the most heat, but the mass of the battery will slow the heat build up .....

Then there is the chemical make up and actual capacity, plate material thickness and construction of the cells to consider.

New cylindrical cells have a lower internal resistance because they only have a single cathode and single anode plate with a separator matrix between them, but only when new, the internal resistance increases faster than a prismatic cell due to the problems with heat rejection from the inner windings

A very thin coating of the active material on the aluminium plate and graphite on the copper plate will reduce the internal resistance. The thinner the coating, the closer it's characteristics mimic an electrolytic capacitor and the lower the storage capacity and cycle life, add more plates in a prismatic cell or length of the copper and aluminium sheets in a cylindrical cell will make up for the loss of capacity, but not the cycle life.

Add yttrium to the mix, the difference between LFP and LYP chemistries, and the internal resistance can be be almost halved and the operating temp range increased.

Basically, you can quote what the manufacturer says, Winston LYP 0.4m ohms as an example, the general consensus of the interweb geniuses of 100m ohms, or test each battery and see what figure you come up with .....

There is a bit of interesting reading here
https://m.littelfuse.com/~/media/aut..._fuseology.pdf
and here

https://www.cableorganizer.com/learn...otection.php#R

The AIC of a mega fuse is 2000amps @ 32v, but I would only use them on 12v and 24v systems, just to be on the safe side.

For those with a higher voltage system, Eaton NH 00 to 1 size fuses in a knife style fuse holder will offer the protection you are looking for to avoid arc over in the fuse

T1 Terry

[CaptainRivet]

Originally Posted by HeywoodJ

Easy to calculate, but the numbers have to be right On top of the typos, CALB and Thundersky, at least, list internal resistance at 0.1-0.5mohm/cell, or an order of magnitude greater than the 0.012 used in the sample. Can’t say I’ve ever seen a cell with 0.012mohm internal resistance, but possibly it exists in a high Ah cell. Also, internal resistance goes up with increasing and decreasing state of charge, a generalization is that lowest resistance is at ~70% SOC, ~3.4V and the short circuit number calculated at that SOC will be 10-20% higher than that calculated at 100% SOC.

Right it was 4 am here...
The resistance of busbars are 0,016mohm not 16mohm, 25x10mm tinned cooper busbar.
I took whats printed on the cell and the report i got with them.
Terry is correct with 3,4V is 100% and yes lowest os at 70% SoC, well thats overhead then as i don't have thst value given.
So i calculate with what i got from manufacturer, if reality is lower, good but i spec to theoretical max.

(16x3,4V)/((16x0,012mohm)+0,016mohm)=54,4V/(0,192mOhm+0,016mohm)=54,4V/0,208mohm=54,4/0,0208=26153,85A

[CaptainRivet]

We need a current number, so lets try to find a way to get that:

Lets take winston with 0,4mohm as Terry said for 400AH cells and you have a 24V bank so 8 cells. Busbars have the same resistance then 1 cell, fits roughly from my experience, we wanna have ballpark numbers to get max current:
(8×3,4V)/((8×0,4mohm)+0,4mohm)=27,2V/(3,2mohm+0,4mohm)=27,2V/3,6mohm=27,2V/0,0036ohm=7555A or roughly 20C.
Given a peak of 4C delivery of the winston cells factor 5 of that is definitly in the right ballpark for a shortend bank, would see factor 10 as top limit so 15000A is not an unrealistic theoretical max current number to spec your fuse.
Learned in 30 years in top management to work with crosschecks and rough over the thump rules to checks and crosscheck if you are in a sensible bandwith and what i got presented seams to be correct or bs. That worked independent of branch or industry.

I know short curcuit current must be high as i had a blown 300A class T got arced by the shortened bank, didn't know the actual capacity as they where cells from a scrapped military vehicle without AH numbers and any documentation but from battery monitor the bank was in the 400AH ballpark range in 4P4S config.

Terry never ever use a megafuse bs as main battery fuse for LFP, they get easly arced but before that they melt the cables around as it transferes enormous heat into them when running at >70% of their rating. Only class T and NH as main battery fuse, emergency or temporarilly a ANL is better then nothing but no megafuse.

[T1 Terry]

Terry never ever use a megafuse bs as main battery fuse for LFP, they get easly arced but before that they melt the cables around as it transferes enormous heat into them when running at >70% of their rating. Only class T and NH as main battery fuse, emergency or temporarilly a ANL is better then nothing but no megafuse.

From over confident stupidity, I can assure you a 300 amp mega fuse breaks a 400Ah 24v LYP battery supply very quickly, about the same length of time it took to think of the words to use when it went zap/pop as the other end of the fuse touched the terminal. Turned out it wasn't an overload of the inverter that took out the fuse, but rather the heat shrink a previous installer had put over a battery positive busbar where it crossed over a battery negative cross bar. Bolting down the replacement heavier cable was sufficient to create a weakness in the heat shrink and when the load went up on the inverter, the path through the weakened heat shrink became the path of least resistance .... a dead short at the inverter.

As far as ANL fuses, both maxi and midi ones, had one 300 amp unit on a 700Ah 12v battery smoke enough to set off the fire alarm.
With every ANL fuse holder, both maxi and midi type, we have had to replace due to heating issues at the clamping ends, one was enough to melt through the plastic case and the plastic lid on the electrics box and only saved the day by shorting out the blade fuse at the battery for system supply (5 amps) and that shut the whole system down via the Gigavac.

The same sort of issues with the round 5AG fuses because they use a wafer spring retained in each end and that creates resistance, maxi blade fuses because the contact area of the fuse blade into the female spade terminal is so small and standard size blade fuses over 15 amps, because they melt the whole lot, fuse case as well as fuse holder .....
Those stupid round small fuses Victron use in their power wire for the BMV has caused me more grief than any other fuse type due to the pitiful contact the soldered wire contacts make against the fuse, high mid pack voltage alarms where the Junsi said everything was fine, complete system shut downs because the BMV lost supply power so the relay contacts opened, simply low voltage alarms even though the rest of the system said the voltage was fine but the resistance at the fuse caused a voltage drop. They get cut out now and a 5 amp blade fuse and holder does the job fine with no problems.

I assume you are refering to these type of fuses https://www.swe-check.com.au/editorials/nh_fuses.php but they also must have a high enough AIR (AIC) and be DC rated. I have 3 on the bench in front of me that are 120kA rated, 500v capacity but it doesn't say DC so I'd have to assume they are DC rated. The 160 amp rating @ 500v would be fine for AC use, but I doubt I'd feel safe were it 500vdc, I'd be looking for a class R fuse, DC rated and 300kA AIR capacity ...... nobody gets dirty about over capacity when it comes to the AIR ........

T1 Terry

[CaptainRivet]

Originally Posted by T1 Terry

From over confident stupidity, I can assure you a 300 amp mega fuse breaks a 400Ah 24v LYP battery supply very quickly, about the same length of time it took to think of the words to use when it went zap/pop as the other end of the fuse touched the terminal. Turned out it wasn't an overload of the inverter that took out the fuse, but rather the heat shrink a previous installer had put over a battery positive busbar where it crossed over a battery negative cross bar. Bolting down the replacement heavier cable was sufficient to create a weakness in the heat shrink and when the load went up on the inverter, the path through the weakened heat shrink became the path of least resistance .... a dead short at the inverter.

As far as ANL fuses, both maxi and midi ones, had one 300 amp unit on a 700Ah 12v battery smoke enough to set off the fire alarm.
With every ANL fuse holder, both maxi and midi type, we have had to replace due to heating issues at the clamping ends, one was enough to melt through the plastic case and the plastic lid on the electrics box and only saved the day by shorting out the blade fuse at the battery for system supply (5 amps) and that shut the whole system down via the Gigavac.

The same sort of issues with the round 5AG fuses because they use a wafer spring retained in each end and that creates resistance, maxi blade fuses because the contact area of the fuse blade into the female spade terminal is so small and standard size blade fuses over 15 amps, because they melt the whole lot, fuse case as well as fuse holder .....
Those stupid round small fuses Victron use in their power wire for the BMV has caused me more grief than any other fuse type due to the pitiful contact the soldered wire contacts make against the fuse, high mid pack voltage alarms where the Junsi said everything was fine, complete system shut downs because the BMV lost supply power so the relay contacts opened, simply low voltage alarms even though the rest of the system said the voltage was fine but the resistance at the fuse caused a voltage drop. They get cut out now and a 5 amp blade fuse and holder does the job fine with no problems.

I assume you are refering to these type of fuses https://www.swe-check.com.au/editorials/nh_fuses.php but they also must have a high enough AIR (AIC) and be DC rated. I have 3 on the bench in front of me that are 120kA rated, 500v capacity but it doesn't say DC so I'd have to assume they are DC rated. The 160 amp rating @ 500v would be fine for AC use, but I doubt I'd feel safe were it 500vdc, I'd be looking for a class R fuse, DC rated and 300kA AIR capacity ...... nobody gets dirty about over capacity when it comes to the AIR ........

T1 Terry

What you had with ANL i had with megafuse several time, never with ANL.
ANL contact surface and material thickness is bit bigger then megafuse in the same rating.
All fuses blow at their rating, doesn't matter if ANL,mega,Class T, NH or whatever.
The problem i have experienced with megafuse from different manufacturers is that they emit enormous heat into the fuse holder and connecting cables when run >70% rating eg 250A with a 300A fuse.
clear to me why just look at the surface of a 200A fuses i have lying here:
Megafuse:13mm x 1mm thick=13sqmm
ANL:17mmx1.5mm=25,5sqmm
NH1: 15mmx6mmthick=90sqmm
Little test i did was running the 200A at 180A for 15min:
Megafuse: >70 C cable and fuse holder starts to get soft, stopped after 10min
ANL: 54 C
NH1: 32 C
Ambient was 25C.
With the NH1 i could have done that 7x24x365 and they would stay at 32C. Which one do you think has the lowest voltage drop after 15min...clear NH1.

All NH fuses in GG or GL are DC rated at 80-85% of their AC rating eg the max 500V AC or 440V DC, you just need to look at spec sheet. They are primarly developed for grid, thats why only AC is printed on them.

Agree the round fuse with spring are devils work, replace them right away with the automotive blade ones.

[s/v Jedi]

You need to perform a test with the actual house bank, a good multimeter and a power resistor (like 100W 8 Ohm from Amazon).

Connect the multimeter to the 12V or 24V bank terminals and note voltage. Now connect the resistor and note voltage again. Repeat a couple times and take average of readings (drop highest and lowest readings and average the rest)

Now set the multimeter to resistance mode and measure resistance value of the power resistor a couple of times and average the readings.

Now the math:

Voltage drop = bank voltage without resistor - bank voltage with resistor

Current = bank voltage with resistor divided by measured resistance in Ohms

Battery bank resistance = Voltage drop divided by Current

Short circuit current = bank voltage without resistor divided by Battery bank resistance.

[CaptainRivet]

Originally Posted by s/v Jedi

You need to perform a test with the actual house bank, a good multimeter and a power resistor (like 100W 8 Ohm from Amazon).

Connect the multimeter to the 12V or 24V bank terminals and note voltage. Now connect the resistor and note voltage again. Repeat a couple times and take average of readings (drop highest and lowest readings and average the rest)

Now set the multimeter to resistance mode and measure resistance value of the power resistor a couple of times and average the readings.

Now the math:

Voltage drop = bank voltage without resistor - bank voltage with resistor

Current = bank voltage with resistor divided by measured resistance in Ohms

Battery bank resistance = Voltage drop divided by Current

Short circuit current = bank voltage without resistor divided by Battery bank resistance.

And i garuntee let 10 people do that with the identical bank and you have 10 different widespreaded results.
Now we see the problem, nice theory from in the guidelines...
Provide a clear table with Li chemistry, voltage and AH capacity and the specs you need for the fuse.
Everything else is bs as not practical and nobody knows what to do.

I use what i have, provided by manufacturer and described, maybe far too high but 1000% not too low so safe.

[s/v Jedi]

Originally Posted by CaptainRivet

And i garuntee let 10 people do that with the identical bank and you have 10 different widespreaded results.
Now we see the problem, nice theory from in the guidelines...
Provide a clear table with Li chemistry, voltage and AH capacity and the specs you need for the fuse.
Everything else is bs as not practical and nobody knows what to do.

I use what i have, provided by manufacturer and described, maybe far too high but 1000% not too low so safe.

Actually, my proposed test is accurate and what you write is imho… bs.

The reason results vary is because installations vary. Deal with it; such is life. Jumpers, crimps, torque, shunts, contactors, BMS’s etc. are all different and make up for the larger part of the total circuit resistance.

If you can’t do this properly then hire a pro

[T1 Terry]

Originally Posted by CaptainRivet

What you had with ANL i had with megafuse several time, never with ANL.
ANL contact surface and material thickness is bit bigger then megafuse in the same rating.
All fuses blow at their rating, doesn't matter if ANL,mega,Class T, NH or whatever.
The problem i have experienced with megafuse from different manufacturers is that they emit enormous heat into the fuse holder and connecting cables when run >70% rating eg 250A with a 300A fuse.
clear to me why just look at the surface of a 200A fuses i have lying here:
Megafuse:13mm x 1mm thick=13sqmm
ANL:17mmx1.5mm=25,5sqmm
NH1: 15mmx6mmthick=90sqmm
Little test i did was running the 200A at 180A for 15min:
Megafuse: >70 C cable and fuse holder starts to get soft, stopped after 10min
ANL: 54 C
NH1: 32 C
Ambient was 25C.
With the NH1 i could have done that 7x24x365 and they would stay at 32C. Which one do you think has the lowest voltage drop after 15min...clear NH1.

All NH fuses in GG or GL are DC rated at 80-85% of their AC rating eg the max 500V AC or 440V DC, you just need to look at spec sheet. They are primarly developed for grid, thats why only AC is printed on them.

Agree the round fuse with spring are devils work, replace them right away with the automotive blade ones.

Do you have a photo of the mega fuse type you have had issues with? We use Bussman and Littel brand and they have been ultra reliable for the last 12 yrs we have been using them.

I'll get a few photos of the melted midi ANL and Maxi ANL fuse holders, not just one, lots of them .....

T1 Terry

[CaptainRivet]

Originally Posted by s/v Jedi

Actually, my proposed test is accurate and what you write is imho… bs.

The reason results vary is because installations vary. Deal with it; such is life. Jumpers, crimps, torque, shunts, contactors, BMS’s etc. are all different and make up for the larger part of the total circuit resistance.

If you can’t do this properly then hire a pro

You are an engineer and me too and know who to measure this properly, 99% out there are not and will screw this up. Eg do they have a meter thats accurate on 0,xx mohm or mV? A lot out there watch 5 youtube how to videos and rhink they are a pro...

Every cable is different too, still we have a table with rating and how to fuse it.

Use a best case installation, measure it as reference,add some security margin and make a table with voltage and capacity of the bank and needed fuse in 50AH steps for Lifepo4.

That will be much more safe and accurate then a guideline like every law that has 1000 interpretations and excuses plus need a bunch of instruments and a pro to size it correctly. And about pro, just look at post from me above, do you think that pro installer whould have got that right, my guess not.

[CharlieJ]

@CaptainRivet #9:

All fuses blow at their rating, doesn't matter if ANL,mega,Class T, NH or whatever.

No, fuses do not "blow at their rating", they blow in accordance with their load/time curves.

clear to me why just look at the surface of a 200A fuses i have lying here:Megafuse:13mm x 1mm thick=13sqmm

"Area" is computed by multiplying length x width. You have measured either length or width and multiplied it by thickness. The mathematical result is not area.