LiFePO4 Batteries: Discussion Thread for Those Using Them as House Banks

[goboatingnow]

The rated voltage and interrupt rating are closely related, because ultimately, the fuse's survival is based on the power it can readily dissipate, which is a function only of its internal resistance and the square of the current passing through it.

Interrupt ability is a combination of several issues, the key one being the prevention of Arc-over. The second one being the containment of the disintegrating element and the quenching of any potential flammable element.

SO creepage , airgap, containment materials are all important , not just Watts. interrupting a high voltage medium current requires different design parameters then a low voltage , high current application.

A fuse is tested at a voltage an current , and the manufacturer has to stand over that claim. In practice interrupt capability is significantly greater then rated and lower voltage have a significant effect as the airgap issue and the maintenance of high voltage integrity becomes less.

Real life tests like Mainsail are entirely appropriate, given that he is operating the fuses at a significantly different operating point. It will not be Ul approved etc. But what works works,

Adequate only for sufficiently small banks where maximum current is sufficiently limited to 6000A by total circuit resistance.

Thats entirely acceptable in a LI setup as being discussed, Also have a look at the short circuit current rise times for Li, that is a factor as well.

Dave

[ColdEH]

One question guys, Isn't battery bulging due to high C rates of charging and discharging ?

So would it really be a problem in a decent sized house bank , say around 600ah. If you never exceed .5C would you really have to worry about bulging or heat ?

Regards

[deckofficer]

Originally Posted by typhoon

One question guys, Isn't battery bulging due to high C rates of charging and discharging ?

So would it really be a problem in a decent sized house bank , say around 600ah. If you never exceed .5C would you really have to worry about bulging or heat ?

Regards

That is what I've read too. My LiFePO4 are charged at 0.5C, but I banded anyway.

[ColdEH]

Originally Posted by deckofficer

That is what I've read too. My LiFePO4 are charged at 0.5C, but I banded anyway.

So no problem as yet ? No heat , no bulging?

Regards

[Maine Sail]

Originally Posted by diugo

I said simply that LFPs can easily exceed 2700A. Do you really dispute this? It is less than 7C for a 400Ah bank. This amount of current is only accomplishable by a 3.3V cell with an internal resistance of no more than 1.2 millohms. Again, not unreasonable, as large cells are spec'd by Winston to be well under 0.5 millohm---and you yourself measured one of your cells at 0.7 millohm IIRC.

I don't dispute at all that Li banks can easily exceed 2700A of short circuit current, Heck Enersys/Odyssey claim 5000A for a single group 31 12V TPPL AGM battery. I would have to guess these banks are capable of 5 figures into a dead short. This especially if a simple group 31 AGM can do 5000A..

After all it ws you who said this:

Originally Posted by diugo

All circuit breakers have a high interrupt capability already built in, to handle a direct short circuit situation. To be specific, the 100A shunt trip breaker I'm looking at is rated at something like 10,000 amps. Now, that may seem adequate---except that 10kA is only 14C to a 700Ah bank, and LFPs have been known to produce 20C or even more, when shorted. Therefore a "catastrophe fuse"---one with an interrupt rating much higher than 10kA---would also be prudent. The nominal fuse rating should be based on the ampacity of the interconnects and main lines (300A in my case).

back on Dec 20, 2012 I posted this:

Originally Posted by Maine Sail

Charlie,

Check out the L17T series fuses. These are 100,000 AIC rated even for DC use, come in sizes from 70A to 1200A, and are used in the telecom industry. A L17T 400A fuse runs about $89.00 here... Still have not found a suitable fuse holder for them but that can easily be over come..

I then went on to have a long conversation with Cooper Bussamann when I decided on Class T as opposed to the L17T series....

Originally Posted by diugo

I chose 2700A because I honestly believed that it was the correct DC interrupt rating for ANL fuses---based on the same datasheet downloaded from several sources.

On the other hand, you quickly claimed the 6000A rating as gospel,

]

I did not "quote anything as "gospel" I simply stated actual facts that are easily obtainable from Cooper Bussmann. I can't explain why teh particualr data sheet you found omitted the 32Vdc rating but they do sell ANL's with a 6000 AIC. I have physically spoken with their engineering department on this very issue, well before this came up last night, to see if they had a 12V AIC rating for their ANL's, which they do not.

Originally Posted by diugo

then in the same post, contrived an experiment to "test" this. Did any of your tests truly succeed at sending 6000A through an ANL?

Your attempts at insults are simply rude. The FACT is that these experiments were NOT contrived in response to yesterday's posts I did the ANL experiments over a month ago and the MRBF tests just last week. They were done in response to NOTHING but my own curiosity because I have sub circuits protected by ANL's. All of these tests were conducted WELL BEFORE this topic came up last night.

I physically uploaded the 8GA / 300A test to Youtube on March 8th. This is upload is dated by Youtube and took place WELL before this topic came up here yesterday. Unless I am a psychic then your misleading statements about my "contrived" experiments are simply incorrect...

Originally Posted by diugo

If not, then the encouraging results you obtained are not earth shattering---they simply indicate the fuse behaved as advertised.

Based on everything we've been led to believe about AIC, and that Odyssey suggested a Class T to protect their AGM batteries, for a 400Ah bank, I am relieved to know that these ANL's and MRBF fuses had no issues tripping with a Li bank. I am still going to use a Class T as my main OCP for the bank..

[diugo]

Originally Posted by goboatingnow

You need far more study firstly on DC and secondly on batteries to make such statements. Have a read of first and second order modelling of batteries for some illumination. The DC system on your boat is not a static voltage, load dumps, transients, glitches, spikes etc are all there and dynamic response is essentially a form of AC ( there is in reality no AC or DC anyway, everything is dynamic).

I can assure you that I have done far more studying on impedance than you have or ever will. I reread your hilarious joke a few pages back about engineers. With your mail-order "electronics" degree, you may be more fun at parties---but please leave the real science to others.

First off, the purely reactive transients you describe are infinitesimal in comparison a large battery's internal resistance. More to the point, do external reactive devices (like motors) in a circuit influence the battery's internal resistance? Of course not. Reactive devices only influence the reactive components of the impedance---resistance is not affected.

I get the distinct impression that you have hooked up an impedance meter to a boat battery and then observed wildly varying impedance while the boat was in operation. Thus immediately concluding that the internal resistance of a battery varies wildly. I guess this is why we engineers make the big bucks.

I'll tell you what. If you promise never to mention impedance again in this thread, we can have a more intelligent discourse on how internal resistance might actually change---and to what extent it would do so in the small time it takes a fuse to blow.

[deckofficer]

Originally Posted by typhoon

So no problem as yet ? No heat , no bulging?

Regards

No heat (well very little) and no bulging, so I probably didn't need to band them. Rather safe than sorry.

[ColdEH]

Originally Posted by deckofficer

No heat (well very little) and no bulging, so I probably didn't need to band them. Rather safe than sorry.

Whats your max discharge rate ?



Regards

[diugo]

Originally Posted by Maine Sail

Your attempts at insults are simply rude. The FACT is that these experiments were NOT contrived in response to yesterday's posts I did the ANL experiments over a month ago and the MRBF tests just last week. They were done in response to NOTHING but my own curiosity because I have sub circuits protected by ANL's. All of these tests were conducted WELL BEFORE this topic came up last night.

I physically uploaded the 8GA / 300A test to Youtube on March 8th. This is upload is dated by Youtube and took place WELL before this topic came up here yesterday. Unless I am a psychic then your misleading statements about my "contrived" experiments are simply incorrect...

Sorry, your post #2295 of yesterday clearly said "I just popped", strongly suggesting your experiments were more recently conducted. If you are upset about the word "contrive", it simply means "ingeniously devise"---no insult was intended. I actually applaud you for sacrificing the fuses. My only gripe was that it was not a valid test of the ANL's 6000A interrupt capacity---which was the topic being hotly debated at the time.

Based on everything we've been led to believe about AIC, and that Odyssey suggested a Class T to protect their AGM batteries, for a 400Ah bank, I am relieved to know that these ANL's and MRBF fuses had no issues tripping with a Li bank. I am still going to use a Class T as my main OCP for the bank..

I bought my four 260Ah cells with the expectation that their combined internal resistance would be around 2 millohms. Dividing 13.3 by 0.002, I calculated a maximum current of 6650A (25C), which exceeded the 2000A DC interrupt rating of my shunt trip breaker, and exceeded the 2700A rating (as well as the "new" 6000A rating) of an ANL fuse, so I too bought a Class T fuse and holder---before my cells arrived.

So I was quite disappointed to find out the battery's actual resistance exceeded 8 millohms, which limits maximum current to only 1663 amps (6C). The Class T fuse turned out to be redundant---and contributes additional voltage sag at high currents. Still, I trust the fuse more than the breaker, so am keeping it.

[deckofficer]

Originally Posted by typhoon

Whats your max discharge rate ?



Regards

Currently (no pun intended) 0.3C. Will be used later on inverter loads of 1.2C.

[senormechanico]

Originally Posted by deckofficer

No heat (well very little) and no bulging, so I probably didn't need to band them. Rather safe than sorry.

Similar to my experience.

I ran my fully charged 200 aH LiFePo4 bank through the sine wave inverter to power the hot water tank (47f water at the start).

The discharge rate was 108 amps for an hour.

While the water temp at the galley faucet went to 173f, the battery temp only rose a couple of degrees, and the battery charge was 46%.


Steve

[ebaugh]

I still want someone to explain what we might expect to see if the ANL failed the test. The only thing I saw posted to my original request was potential arcing. Even if it arcs, the Internet searches I did indicated it would not last, or at least that's how I read them for a 3 or 12V application. HV is different. But I have no claim to an engineering degree, so maybe one of you can help me out.

[ColdEH]

Originally Posted by deckofficer

Currently (no pun intended) 0.3C. Will be used later on inverter loads of 1.2C.

I have an order in for 600ah of Sinopoly cells I hope to get them by May.

My Victron inverter takes 300amps max , and I will be charging at 240 amp so I should be staying in the 0.5C range .

But as you I will band them , just to be sure .

Regards

[bill good]

Impedance is only one part of ESR as it is given so you can apply it in parallel with the DC resistance to work with complex circuits. i.e. get both figures "Z" & "R" before you progress then you cal ESR. So?? The idea of protecting is to cal your load X a factor & the first disconnect at the battery is determined by this & not battery capacity. Also the devices are "current limiters" not fuses. The functions are different. Some installations do call for current limiters to be used instead of links between cells. I do see ESR being quoted with reference to frequency but that is only with ref to the AC component resistance "Z".

[deckofficer]

Originally Posted by typhoon

I have an order in for 600ah of Sinopoly cells I hope to get them by May.

My Victron inverter takes 300amps max , and I will be charging at 240 amp so I should be staying in the 0.5C range .

But as you I will band them , just to be sure .

Regards

As I've said to others, welcome to the lithium club. You will discover how conservative they are rated as to a-hr, and when you pull 300 amps it will be a fraction of the voltage sag of LA.