They aren't dangerous per se, but demand respect. The failure modes are extremely rare if you use quality cells and treat them well. Proper BMS is absolutely necessary! Sadly I often find BMS-deniers that think it's just not needed, you can read about a lot of fires from the "DIY powerwall" people running w/o BMS.
Now that you can get a cheap and easy BMS, there's no excuse. I think I'd be comfortable installing an LFP pack in a crawlspace, but I'd build a box to house them. You also want to constrain them on the large faces so they can't swell anyway. Constructing a box with a dryer vent tube that leads outside would allow me to sleep well at night. On Wed, Feb 25, 2026 at 12:15 AM DOOLEY PHILIP G JR <[email protected]> wrote: > LFP sounds much more dangerous than I had been led to believe. I had been > planning to put a used Calb LFP house backup battery in the insulated > crawlspace under the airlock entryway. However the floor above the cells > is not airtight possibly gassing the escapees. I could just get another 8 > golf cart batteries to replace the old ones there that came from my first > EV. I had hoped to soon avoid crawling in there with a watering can. > Connecticut climate spends a lot of time below freezing, do you have more > info about charging LFP cells at or below freezing ? possibly slowly ? Is > water involved internally or do other chemicals have the same freezing > point ? > > My first EV (a '93 escort electricar) now has a Trimetal Li battery giving > a 30 mi range when above freezing and the '98 Solectria now has used coda > LFP batteries. Both have Orion BMS. > > I am working on assembling 2 ea 72AH 12V LFP batteries to replace the lead > batteries for lighting. > > Phil Dooley > > > On Tuesday, February 24, 2026 at 01:33:31 PM EST, (-Phil-) via EV < > [email protected]> wrote: > > > To be clear LFP is a shortened abbreviation of LiFePO₄. > > Yes, That was my conclusion too. I have earlier video on it with more > details: https://www.youtube.com/watch?v=i27lApNWkyA (Note: This was > made > 2 years ago before Tesla introduced the LFP version they are rolling out > now) > > Indeed, Tesla is now using LFP in some models for the "12v" system. LFP > is "safer" than ternary; No doubt, and more tolerant of abuse. If you are > building any kind of home backup or battery storage that will be in or near > your house, it's my opinion you should stick to LFP. Those cells can still > go into thermal runaway internally, but it usually doesn't spread to > adjacent cells. (USUALLY!) > > HOWEVER, LFP can still have nasty failure modes that emit huge volumes > (many orders of magnitude of the cell volume) of toxic gases that can > easily kill you and/or cause an explosion if they are inside your house. > So a large pack in your attached garage could easily kill you in your > sleep even if just one large format cell goes off! It should be in a > sealed enclosure with a vent to the outside. I've seen dryer vents used > for this with the little metal flap, so it stays closed unless there is > overpressure in the enclosure and then safely vents outside. Another > technique used is a foil seal. You see this on the top of many large LFP > cells. The foil will burst and allow venting, but ensures the cell stays > sealed in normal operation. Since you cannot charge these cells below > freezing, and they have a longer life if kept cool (i.e. They prefer to > live in the same temperatures we do), it's tempting to install them inside. > > I personally have a ~9kWh LFP battery in my house using 16 of the Calb > 180Ah cells in a 16S configuration. (~48V) It's in an air-sealed enclosure > with a vent fan that pulling air from outside through a filter and venting > out the top to a vent stack on the roof. I live in CA where I enjoy mild > temperatures, so the pack stays pretty comfortable. The fan is on a > thermostat set for 80°F and this is located at the top near where the fan > is. I specifically chose a brushless fan and solid state thermostat to > ensure there's no chance of ignition. This fan/thermostat arrangement > stops the air from getting too cold in the winter. My BMS keeps the cell > voltages from ever exceeding 3.4v and pack at or under 54v which is > conservative. This reduces danger somewhat and greatly extends the life of > the pack with only marginal reduction of capacity. I have a monitor on > the BMS that will send me a text if it ever goes too far off track, here's > the current state: > Overall pack: 54.0 V > Highest Cell: 3.384 V (02) > Lowest Cell: 3.375 V (10) > Cell Delta: 0.009 V > Battery Temp: 28 °C > > Note that regardless of cell quality, you should always expect an internal > cell failure, it's rare, but does happen. You should build your battery > system as I did, with this failure in mind! > > Here's a copy/paste from an LFP safety analysis: > "When a lithium iron phosphate (LFP or LiFePO₄) cell undergoes internal > failure (e.g., due to an internal short circuit) or is externally shorted, > it can lead to rapid heating, potentially triggering thermal runaway. This > process causes the cell to vent gases through its safety valve or rupture > points. These gases arise primarily from the decomposition of the organic > electrolyte (e.g., carbonates like ethylene carbonate), breakdown of the > solid electrolyte interphase (SEI) layer, and other internal reactions. > The emitted gases include both flammable and toxic components, posing risks > of fire, explosion (if accumulated gases reach ignition conditions in a > confined space), and health hazards from inhalation. > > Main Dangerous Gases Emitted: > > 1. Hydrogen (H₂): Often the most abundant flammable gas (frequently 30–55% > or more of the total vent gas volume, depending on conditions like state of > charge). Highly flammable and explosive when mixed with air; contributes > significantly to explosion risks. > > 2. Carbon monoxide (CO): Toxic and flammable (typically 8–28% range in > studies). Colorless, odorless gas that causes asphyxiation by binding to > hemoglobin; a major toxicity concern. > > 3. Carbon dioxide (CO₂): Non-flammable but can displace oxygen and > contribute to asphyxiation in high concentrations (often 15–36%). > > 4. Hydrocarbons (e.g., methane CH₄, ethylene C₂H₄, ethane C₂H₆, propylene > C₃H₆): Flammable and present in smaller but significant amounts (combined > often 5–15%). These lower the ignition energy and widen the flammability > range of the gas mixture. > > 5. Hydrogen fluoride (HF): Extremely toxic and corrosive gas (produced from > decomposition of the LiPF₆ salt in the electrolyte and fluorinated > binders). Forms hydrofluoric acid in moist air or when water is applied > (e.g., during firefighting); causes severe respiratory damage, skin burns, > and systemic toxicity even at low concentrations. LFP cells can produce > notable amounts of HF, sometimes more than other chemistries under certain > conditions." > > On Tue, Feb 24, 2026 at 7:55 AM Mark Hanson via EV <[email protected]> > wrote: > > > Hi Phil etc > > That’s very interesting that just normal charging ternary NMC cells that > > are dead at/below 2V per cell can cause a fire due to dendrite growth > > shorting out the plates. > > Your video showed Tesla using these cells in their 12v aux batteries on > > new EVs. Sounds like Tesla screwed up and should use LFP cells for 12v > > battery replacements like everyone else > > Do LFP batteries have the same problem if a dead cell <2V per cell is > > normally charged? I’m sure I’ve done this in the past with LiFePo4 cells > > without thinking about it (mostly small battlebots). Don’t recall > running > > into this when I used to convert EVs using Calb or Thundersky LFP cells > > (since EVs didn’t sit to long without being charged) > > Stay Charged, > > Mark > > Sent from my iPhone > > _______________________________________________ > > Address messages to [email protected] > > No other addresses in TO and CC fields > > HELP: http://www.evdl.org/help/ > > > > > -------------- next part -------------- > An HTML attachment was scrubbed... > URL: < > http://lists.evdl.org/private.cgi/ev-evdl.org/attachments/20260224/a9cb8b19/attachment.htm > > > > _______________________________________________ > Address messages to [email protected] > No other addresses in TO and CC fields > HELP: http://www.evdl.org/help/ > > -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.evdl.org/private.cgi/ev-evdl.org/attachments/20260225/24018c98/attachment.htm> _______________________________________________ Address messages to [email protected] No other addresses in TO and CC fields HELP: http://www.evdl.org/help/
