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

[Checkswrecks]

Practical fire fighting -

Problem: For those reports about EVs which occasionally catch fire, we've seen numerous cases where li-ion powered cars will smoke or burn for more than 5 hours and re-ignite numerous times. Any movement can cause an internal short circuit and the process re-starts. I've also seen LiFePo pouch cells go for a couple of hours. One car I looked at re-ignited 6 days after the initial fire (3rd ignition), because the heat melts out protective insulation in cells surrounding those initially damaged. Flight attendants put out small fires and smoke almost daily on airplanes these days and have gotten quite good at it, because so many batteries are in use. With so many million of them a bad or damaged cell is not a matter of "if" there will be a thermal runaway, it's a matter of when.

What is happening is that the plastic separator develops a small hole between the foils (anode and cathode), allowing current to pass, which creates a localized heat. The heat melts more separator and more current is dumped in, growing the hole, so the cycle and heat just gets worse.

So ever wonder how to put out a battery fire on a boat? The answer is that you need water to reach and cool the individual cells. The typical emergency response instructions for large systems state outright that the most efficient firefighting is to get the water into the racks to the cells. The ferry boat linked above actually has plumbing to each rack of battery cells to flood them with sea water if need be.

The plastic separator material melts at roughly 130C and the water of course boils off at 100C. With so many E-vehicles in Europe, they have come up with a simple and highly effective solution.

[mike_kelly]

ChecksWrecks, Thanks for the info. I found one interesting unit but there is just not enough information. Balancer/Monitor module with display, the actual balancer and the charger all tied together with serial communition. The label is Chargery but it appears to only deal with one bank of cells. It would be nice if you had say 4 100ah bundles to have a network connection, like Canbus-I2C-SPI, to allow one Monitor DIsplay to control multiple balancers with one central charger. They appear to be designed for high voltage systems for cars primarily. But not much info on the net as far as quality etc.

[mike_kelly]

Originally Posted by Checkswrecks

Practical fire fighting -

Problem: For those reports about EVs which occasionally catch fire, we've seen numerous cases where li-ion powered cars will smoke or burn for more than 5 hours and re-ignite numerous times. Any movement can cause an internal short circuit and the process re-starts. I've also seen LiFePo pouch cells go for a couple of hours. One car I looked at re-ignited 6 days after the initial fire (3rd ignition), because the heat melts out protective insulation in cells surrounding those initially damaged. Flight attendants put out small fires and smoke almost daily on airplanes these days and have gotten quite good at it, because so many batteries are in use. With so many million of them a bad or damaged cell is not a matter of "if" there will be a thermal runaway, it's a matter of when.

What is happening is that the plastic separator develops a small hole between the foils (anode and cathode), allowing current to pass, which creates a localized heat. The heat melts more separator and more current is dumped in, growing the hole, so the cycle and heat just gets worse.

So ever wonder how to put out a battery fire on a boat? The answer is that you need water to reach and cool the individual cells. The typical emergency response instructions for large systems state outright that the most efficient firefighting is to get the water into the racks to the cells. The ferry boat linked above actually has plumbing to each rack of battery cells to flood them with sea water if need be.

The plastic separator material melts at roughly 130C and the water of course boils off at 100C. With so many E-vehicles in Europe, they have come up with a simple and highly effective solution.

So if you get a LiFePo4 fire on your boat the best thing to do is open the seacocks?

[newhaul]

Originally Posted by mike_kelly

So if you get a LiFePo4 fire on your boat the best thing to do is open the seacocks?

that statement is out and out ridiculous.

In case you missed the glaring point that Lfp is a significantly different battery than those that are shown in the car fire videos. There is little difference between the likelihood of an Lfp fire and a Fla battery fire . The connections on the main distribution panel is actually the most vulnerable point . So don't forget the main system battery fuse.

[mike_kelly]

To quote Star Trek
Saavik:
Humor. It is a difficult concept. It is not logical.

[john61ct]

The whole issue of fires, and LI chemistries other than LFP are really off topic here.

[nebster]

Originally Posted by CatNewBee

Regarding BMS. You need one, definitely! No matter what you read here about shoulders and PSOC.

This is incorrect. There are many very happy packs running with no active balancing, including some much larger than his.

The BMS industry would like you to believe that you need a complex, active system to run a successful battery. Their incentives are not necessarily aligned with ours.

LFP cells and connectors have slightly different resistances, what leads to imbalances when cycling. There is no way to avoid that, there are no ideal pre-selested packs, they will drift.

The evidence shows that this is not the case in real-world use. Slight capacity variations cell to cell do change the individual voltages per unit energy flux (which is one reason we bother to end balance in the first place), but whatever delta/drift exists is too small to matter in healthy cells. Unless you insist on charging to 100%. Then you do need such a management system, and you should also plan on the concomitant loss of cycle life.

[john61ct]

+1 to all that ^^^^

If starting off with secondhand or otherwise sub-par cells, or as a bank ages and SoH is reduced (tested capacity falls below nameplate rated)

and especially if the owner has regularly been pushing stop-charge SoC up past the shoulders

or using bottom balancing and letting SoC fall regularly below (resting) 3.2Vpc

then more drastic imbalances may arise more regularly, further reducing effective bank capacity and / or requiring regular corrective maintenance

or even "live" per-cell balancing gear for safety, rather than simply monitoring.

[Q Xopa]

Originally Posted by newhaul

john how do you do any of this ??

Did you suddenly wheel out and buy Lfp cells ?
You should have told the house . Now you need to post all about it .
Size and brand of cells . What bms are you using . Also how much is invested in your house bank for your cruiser type boat?
If you did
Welcome to the group. If not please stop misrepresenting yourself on these threads.

As you commonly request- please post all you results, links and costs.

Thanks.

[newhaul]

Originally Posted by Q Xopa

As you commonly request- please post all you results, links and costs.

Thanks.

no problem its posted in the build and cost thread
http://www.cruisersforum.com/forums/...ml#post2864236

[Checkswrecks]

Originally Posted by john61ct

The whole issue of fires, and LI chemistries other than LFP are really off topic here.

John -
For the most part, but not totally. As mentioned, I do come across things that most people don't and I've been doing a lot of battery cases for the last 6+ years. For example, the results of hard-shorted LFP which went up in intense smoke and left behind an impressively burnt out stainless battery box. Somebody simply dropped a tool which rolled across the terminals. I agree with you that LFP is pretty darned safe but it still has done impressive damage simply from negligence.

Mike -
While opening the seacocks would be a bit extreme it'd be a lot simpler to be able to flood the battery box from a sink hose, deck wash-down, etc.

[newhaul]

Originally Posted by Checkswrecks

John -
For the most part, but not totally. As mentioned, I do come across things that most people don't and I've been doing a lot of battery cases for the last 6+ years. For example, the results of hard-shorted LFP which went up in intense smoke and left behind an impressively burnt out stainless battery box. Somebody simply dropped a tool which rolled across the terminals. I agree with you that LFP is pretty darned safe but it still has done impressive damage simply from negligence.
:

that sounds real similar to what would happen in the same scenario with lead based batteries.

Hence the reason its always best to have any battery covered to prevent accidental hard shorts such as you just described.

[senormechanico]

Originally Posted by newhaul

your correct about having other options but you would carry more weight here if you actually were using Lfp for a house bank like the title of the thread implies .

To paraphrase, John is saying,
" I DID NOT, have experience with LiFePo4's". Not one time. Now, I need to get back to the work of educating the American people who actually are using, or thinking of using them."

[Q Xopa]

Originally Posted by newhaul

no problem its posted in the build and cost thread
http://www.cruisersforum.com/forums/...ml#post2864236

Sorry that request was for John.

[Oceanride007]

Would like to put a smallish pack together say 180AHrs, Intend having largish AGM battery in parallel but isolated, have read 165 pages here, takes forever. Intend to charge LFP with MPPT mostly, maybe a DC2DC charger as standby.
Aware that any MPPT charge voltage can overcharge. Suppose Float voltage can be wound down to safe "No charge" voltage, (don't know?).
Is that done is it possible, how do you avoid mini charge cycles, if vessel layed up. Have thought of applying Logic with timers, relays etc (yuck?). And keeping safe margins, but capacity is being lost.
How is this done without baby sitting, esp those without cell level BMS? Thks.