Wow, basen rack.mounts batteries destroy a home in germany

Partimewages said:

In Texas maybe not going to work in Ohio

Click to expand...

Just install a gas boiler to keep it warm in there!

hwy17 said:

Not in my area. If it's a fixed installation it's supposed to be permitted, no tricks to get out of it. Other than just not asking and doing it anyway.

Click to expand...

May be I just make it into a trailer then.

hwy17 said:

Not in my area. If it's a fixed installation it's supposed to be permitted, no tricks to get out of it. Other than just not asking and doing it anyway.

Click to expand...

Forgot to mention, it's farm use. No permits required. I just need to report the size of the building for taxes to the county Auditor.

SeaGal said:

My 2p worth... As there was a gas boiler in the same room, it seems highly probably that the LFP's were not the cause of the explosion. This is for all of the reasons mentioned in this and the other thread and the fact this (as far as we know) is the only potentially reported incident compared to the, presumably, millions of LFP battery packs in use around the world.

However, if it is _possible_ that LFPs can vent hydrogen under normal operating conditions _and_ there is a quick fix to the design of our DIY battery packs to vent the hydrogen, then I guess it would make sense to remove that miniscule chance of this happening again by improved design?

Click to expand...

Well put.
I wonder the following:
If the failure modes are addressed one by one, as follows
1. over voltage on cells could cause gas venting to occur. Most of us have both a charging control (inverter or dedicated SCC) and a BMS on each pack. For an over-voltage to occur under this set up, both of these would need to fail together. This seems highly unlikely. A cautionary approach could be a third level voltage monitor, at <set point voltage> two pole relay opens and all current in or out of the ESS is shut down. Other options for over-voltage control? Lets hear it.

2. over current on cells (in or out) could cause venting of gas. Again typical installations everyone here is setting up will limit current flows with SCC/inverter charge limits, as well as BMS settings equipment limited max current (in and out). Most of us also employ Class-T fuses to limit max current, and DC breakers on each batter pack. This would seem like plenty of back ups to the back ups. I wondered as I added to my packs, what if all but one pack is offline - ie during charging, one by one the packs stop accepting charge, until only one pack it online, but the Inverters are still trying to push 200Amps - each pack is built with 125A breaker, and 125A Class-T, and the BMS is set to max 100A. Seems sufficient to prevent excessive current. Other thoughts? Other back up measures?

3. overheating cells could cause venting to occur. A heat detector alarm (like a smoke alarm) could alert you to the danger, various types of venting as noted by others earlier could be employed. Perhaps power assisted venting. I am not sure how the connects would be done (simply) but we could consider a dedicated vent attched to each pack case (for those of us using cases). Or for those that have all the LFP packs in a big steel box - add a venting system just to the box to direct the gases produced to the exterior under cases of excessive temperature to ensure any vented gases leave the room safely. The big steel box with battery packs in the box would seem the best option for this, rather than a server rack with lots of holes and slots and mesh panels.

OffGridForGood said:

Most of us have both a charging control (inverter or dedicated SCC) and a BMS on each pack. For an over-voltage to occur under this set up, both of these would need to fail together.

Click to expand...

The inverter / SCC overvoltage will not apply at a cell level, so just the BMS failure _could_ cause a cell overvoltage if the cells are not well balanced.

True, I missed that mode of failure.
If just the BMS fails, and the cells are not well balanced/we have a runner, it would be possible to have a cell voltage too high, leading to venting.
What back up system could we add, such that the BMS is NOT the sole protection (per cell) for over-voltage?

Edit: however if the BMS fails, does it not cut off charge/discharge ?

Two BMS's! Only kidding. Seriously though... it's all about probabilities and levels of protection vs. the risk.

Unfortunately MOSFETs do fail short circuit - I had to replace one recently on a switched mode PSU (not solar related) that was shorted in all directions and blew the mains fuse.

pvgirl said:

I've attached the full test report, while the test included multiple batteries, the report seems to indicate only one rack battery vented in the test.

Click to expand...

Thanks. From a risk perspective UL seems to think a single case is the failure mode.

Partimewages said:

Small fan in a large room will not cut it. Some type of sensor to start a fan of proper size might work. Continuous venting might not be good depending on climate.

Click to expand...

IIRC the LEL of H2 is over 4% of room volume. Your exhaust is a function of the maximum off-gassing rate of the batteries. Honestly, a think a small HRV unit is all that is needed; 2CFM would be plenty for a large room, especially if you are pulling from the batteries.

SeaGal said:

Two BMS's! Only kidding. Seriously though... it's all about probabilities and levels of protection vs. the risk.

Unfortunately MOSFETs do fail short circuit - I had to replace one recently on a switched mode PSU (not solar related) that was shorted in all directions and blew the mains fuse.

Click to expand...

Maybe that's not too far from practical. A second cell monitoring device that doesn't participate in balancing and only controls on/off charging current? It could also be interlocked with the BMS such that if either fails, power is cut.

CoolWill said:

Maybe that's not too far from practical. A second cell monitoring device that doesn't participate in balancing and only controls on/off charging current? It could also be interlocked with the BMS such that if either fails, power is cut.

Click to expand...

Setup for that would be fairly easy if you use terminal blocks with 3 ports on it.

I’m thinking a fire/heat resistant box vented at the top to the outside. I’m guessing the venting will create positive pressure. Not sure a bottom vent is required. In a positive pressure event, it might work against you. Not much volume in an enclosed battery rack. But still thinking about it.

Daddy Tanuki said:

Click to expand...

I noticed they use the plastic cased Wistron cells for the LFP test. Curious if the more common thin walled aluminium cells used in many of the diy and rack mounted batteries behave similarly in terms of containing expansion?

Daddy Tanuki said:

how many cells? they claimed a battery... too many people looking for an issue that does not exist at this point, the amount of gas that a 16s pack could put out is barely enough for the cubic meters... all of the articles did not specify how many cells just a "battery" or a "battery"'s cell output. sloppy data.

Click to expand...

Presumably 143 cubic meters is the volume needed to keep the hydrogen concentration below explosive levels, when one 5kWh battery vents. So for many who are running 20kWh+, they will be in that "danger zone" of too little air volume for a given capacity, albeit with a very small likelihood of occurrence.

rhino said:

Setup for that would be fairly easy if you use terminal blocks with 3 ports on it.

Click to expand...

So how do we build it?
Or it is better to ask JK to build it into the BMS - all one unit?
DIY - Arduino - monitor the cell voltages and if any exceed the set point the main relay (added) cuts off all charge and discharge.

this is an interesting thread, and also one that is always good for rejuvenating a second look at our current installs. I think its just safe to play it safe, I would have put my lfp batteries outside in a shed if it didn't get to -40*c here, and reading this makes me want to still do that!

myles said:

I would have put my lfp batteries outside in a shed if it didn't get to -40*c here, and reading this makes me want to still do that!

Click to expand...

Can always just put them in an insulated, temp controlled heated (and possibly now, vented) case. Don't need much power to keep them warm. A 20W heater and 2" of PIR kept mine at 20C+ when was -5c outside last winter and it was only on about 50% of the time.

Occams Razor. Let's put a relatively sealed up utility room in the bottom of a house and put a boiler in it. Then throw a PV system with some batteries in there along with it. When it blows up blame it on the batteries.
In another thread a guy was talking going off-grid and putting propane in to heat / cook along with a PV setup. I suggested running the propane to a generator in an outbuilding with all the PV system, use a heat pump for the house and a wood / pellet stove inside for brutal cold days. The reasoning should be obvious. You want to seal up your house pretty tight and well insulated against the cold. This means all the airflow inside is forced, any gas that leaks stays inside. No gas in the house no invisible combustibles that are easily ignitable. Wood generally will not ignite spontaneously from vapors and a spark created by lighting a candle in the living room. If you want to run gas appliances inside, more power to you, but not me, never again. I'm lucky to be alive, someone else in this thread also described a similar incendent. The problem with gas is you just can't see it, and noses are amazing about filtering out repeated smells. Sensors can fail. Dunno if this was gas or not, boiler could have been oil fired, but most of Germany uses natural gas not heating oil. It's just too easy to have a gas leak, and not know about it within a reasonable time frame.
Battery fumes from LifePO4 causing an explosion, while remotely possible, are highly improbable.

We have a fairly air tight house as well, and certainly you absolutely need to have HRV running on a schedule. Ours is on for alot of the day just exchanging the air in the house. The garage where the LFP batteries sit, is heated slab 12*C approx, and I have an exhuast fan that is run on a humidistat in the garage so there is always fresh air being pumped in and out of the garage. Air tight houses need more attention to air quality, and humidity than any other type of building

Groscout said:

Is it possible that the boiler t&p valve and boiler limits failed creating a super heated steam explosion event without fire at all? Corked water heaters blow up houses all the time. A water heater recently blew a house to little bits in Plum, Pennsylvania:

Click to expand...

yeah remember mythbusters:

water tank will yank the house off the foundation!

That blast had flames riding on top, so I don't believe it was a steam explosion. Gas wouldn't have been released until after the steam explosion, so no buildup.

I would think that was gas itself.

(based on reasoning, not past experience reviewing such things.)

Now as for the Mythbusters one, I think that was electric not gas, so no flames, can't use it to evaluate the house explosion.
But it was quite effective at "disresembling" the shed.

CoolWill said:

30 kWh of energy is less than the energy contained in a gallon of gasoline. Do we believe that a single gallon of gasoline could turn a house to ruble?

Click to expand...

And gallon on gasoline has higher energy content than 60 pounds of TNT or C4.
Takes helluva big mansion if you can't level it with 60 pounds of TNT.

Now 1 pound of TNT won't blow a house to ground but 1 pound of wheat flour will do it.