LifeP04 meltdown in a Sprinter van ?

[hipringles]

This is the f’ng problem….
I have been careful to not put metals or conductors “on” the pack. There is always a piece of plywood above the hole (build is not yet completed) and nothing touching the metal. The night before I was doing some work on the Shoilet and I needed to move the plywood out of the way.

What we figure happened - I f’ed up the top balance packs. Essentially, the supplier top balanced in packs of four, I got in packs of four, I kept all packs together for the most part….. until I put the final assembly together and mixed up two cells…..
from what we can tell - preliminary results - indicating that cell 1 that went up was supposed to be be in jar 2 and not 1…. And the supplier is going to let me know which others I mixed.

Point for me is to put a permanent S/N and group number on the cells. The imbalance was enough

View attachment 82540

So are you saying you took cells which were at different SOC and paralleled them? Ie, in the first "parallel" group you had 2 at a higher SOC than others?

OR are you saying you took 4 cells at a higher SOC, and series connected them to the lower SOC group?

The first should not cause issues, minus a few minutes of very high amperage between cells when you parallel them. The cells will balance to each other as they charge (basically a top balance), and will not cause a short or any issues.

The second would result in crap capacity in the pack. The only way it results in letting out the magic smoke is by the BMS not cutting off charging when the first cell group hits ~3.65.

If your BMS is not monitoring per cell, but rather for the whole pack voltage, then #2 could also be a problem.

Part of it is the BMS and part of it is me. I am going to assume I did something incorrectly (caused a short somehow)

"Caused a short somehow" when building the battery is immediately obvious. It is not something that will sneak up on you later.
Causing a short over time is likely due to either: connections getting loose, or something changing in the "environment" (aka container) that the batteries are in, such as someone dropping a metalic object on the studs, the batteries wiggling against a metal surface and drawing a short, etc

 

[noenegdod]

No - I know I know- f’ng rookie why the f did you not….

Considering the outside of the packs are all negative, even if over duration they wore out…

Honestly I am going over the whole system and building out for robustness.

I have a Simarine Pico system to monitor.
I will be going with 4 batteries- 4 cells each with 150A BMS and balance board on each battery and a Simarine Shunt with TS.
The Simarine solution has an ability to trip a relay (call into factory) and can trigger an alarm if a Temp sensor goes high.

Also going to build compression plates with FR4 pads to hold the batteries in place, locked.
This is all going to be in a box under the van where I will build an ejection solution to eject the pack if it decides to behave badly

I just wanted to give you a huge pat on the back for posting up, exposing everything you did, acknowledging mistakes, looking for solutions. and making the community a better place. ?

 

[TorqueMonster72]

…..saying you took 4 cells at a higher SOC, and series connected them to the lower SOC group?
The second would result in crap capacity in the pack. The only way it results in letting out the magic smoke is by the BMS not cutting off charging when the first cell group hits ~3.65.

If your BMS is not monitoring per cell, but rather for the whole pack voltage, then #2 could also be a problem.

Thanks for clarification on my words. This is what happened. So with four cells in parallel, B1 would have the 1120Ah capacity. B2, B3 and B4…
The BMS was “watching” 4 parallel 3.2V cells as one cell. B1 has always been low post assembly/install compared to the others.

Causing a short over time is likely due to either: connections getting loose, or something changing in the "environment" (aka container) that the batteries are in, such as someone dropping a metalic object on the studs, the batteries wiggling against a metal surface and drawing a short, etc

I never put locktite on the studs. I also in my stacks on top of the studs never placed loctite on the lock flange nuts.

I also used stick on zip tie holders to hold the wires down. I never inspected the wires post hold down and there is a possibility of moving and chafing through the insulation.
Nor did I inspect the pads held.

I also did not insulate the box for sub 0C temps, nor did I put heat pads to keep things warm.
I also kept shore power on and topping up the unit at 100% in sub 0C temps.
Also was drawing power from the cells whilst driving to use the 12V fridge..

Overall - my first mistake was to assume LiFePo4 can operate as well as AGM in the same environment with same procedures.

 

[Rocksnsalt]

wow.
I have a feeling some comments are coming…?

?

 

[time2roll]

Sounds like the DALY did not prevent charging at low temperature. This may not be the first time.

 

[TorqueMonster72]

Wife has decided that I will not build the solution but take to an installer. I am reluctant but “happy wife happy life” and all….
I will insist however to run 4 12V 280A packs and it will be under van. Will mean heat pads for sure.

As well - the Xantrex XC and Nations/Wakespeed has been problematic so hopefully the installer will be able to manage to make this all work together.

 

[TorqueMonster72]

I just wanted to give you a huge pat on the back for posting up, exposing everything you did, acknowledging mistakes, looking for solutions. and making the community a better place. ?

No good deed goes unpunished….

I am sure that there will be those on the board who would want to have me drawn and quartered publicly after being flogged….

To those I say - yes I should have done more homework and should be an SME - and you also did not have to look at an erupting cell - into a space where your children would be sleeping, who could have been suffocated with toxic gas and then burned - and have to put the dammed thing out not knowing if any of the other cells are going to grenade…..

If they still want their pound of flesh, send me a PM.

 

[Browneye]

Sometimes we just don't know what we don't know. And really, that's why we're all here.
Sometimes you think you can get away with not doing things perfectly, or ignore inclinations that maybe what you've done isn't exactly top notch.
Sometimes good enough just isn't.
Sometimes we simply cannot foresee the ramifications for these failures or shortsightedness.

Some people just run away and hide, or lie about it, instead of facing the music. Or blame someone else.
Some are smug in their knowledge or experience and think they know better. Sometimes they do, and sometimes they don't.

I have no ill-will against anyone for not knowing everything about what they're doing. If anyone does then that is on them.

 

[hipringles]

Thanks for clarification on my words. This is what happened. So with four cells in parallel, B1 would have the 1120Ah capacity. B2, B3 and B4…
The BMS was “watching” 4 parallel 3.2V cells as one cell. B1 has always been low post assembly/install compared to the others.

I never put locktite on the studs. I also in my stacks on top of the studs never placed loctite on the lock flange nuts.

I also used stick on zip tie holders to hold the wires down. I never inspected the wires post hold down and there is a possibility of moving and chafing through the insulation.
Nor did I inspect the pads held.

I also did not insulate the box for sub 0C temps, nor did I put heat pads to keep things warm.
I also kept shore power on and topping up the unit at 100% in sub 0C temps.
Also was drawing power from the cells whilst driving to use the 12V fridge..

Overall - my first mistake was to assume LiFePo4 can operate as well as AGM in the same environment with same procedures.

Oh no. You wired all the BMS leads to one cell? That is unfortunate.

Wife has decided that I will not build the solution but take to an installer. I am reluctant but “happy wife happy life” and all….
I will insist however to run 4 12V 280A packs and it will be under van. Will mean heat pads for sure.

As well - the Xantrex XC and Nations/Wakespeed has been problematic so hopefully the installer will be able to manage to make this all work together.

Please be careful with under van mounting... This can have it's own issues

 

[time2roll]

I would be more inclined to have an interior steel box vented to the outside to reduce the chance of fire or fumes in the van.
Subjecting the battery to road hazards seems to add some risk.

 

[S Davis]

Even if you have someone build it, learn how and make sure they do it right. I have two 12 volt 280ah batteries in the back of my crew cab pickup, I researched for months before I came up with a plan I was comfortable with.

In my setup I am driving off pavement with rough roads so felt it a good idea to secure the cells as tightly as possible (11-14psi) with insulating material between. I used flexible buss bars and they are mounted on vibration dampers. I also scan them with a thermal camera regularly when drawing high amps.

And I’ll say it too, thanks for sharing.

 

[Mike Mas]

Hello - Sorry for your loss!

This just confirms that there is always a potential danger using using lithium - the entire safety of the battery is only as good as the BMS which protects it.

Mike

 

[TorqueMonster72]

Hey all. So…. The culprit - short to the chassis of the van. The corner rubbed on a piece of the frame holding the cells together. I thought I was being smart with how i was securing the cells, but….. not so much on bumpy roads.

I had butyl backing on the case and the road vibrating wore out the butyl at that point over time. I never expected this as an outcome as automotive glass is held in place with Butyl tape….

F*** me! I am a F’ng moran!
Still going to hire someone to overview the battery build and installation.

 

[TorqueMonster72]

Root Cause Analysis.

Primary cause
Chafing of the Cell housing exposing ground plane to the van chassis - When I put the assembly together, I thought it was secure. driving the van caused the cells to rub the box that it was enclosed in. Over time, the cell protection wore away and caused a short.

Secondary contributory
- BMS did not balance cells - The BMS was having problems keeping all the cells balanced. Jar2 was always higher than the other 3 jars, which restricted some flow to Jar 1.
- No Cell warmth - the location of the batteries was under the van and not insulated, nor heated. LiFePo4 do not behave well in 0C and should not be charged/discharged. I was using the batteries while I was driving on the highway in -3C conditions.
- BMS and Xantrex were not properly aligned - The Xantrex XC Pro 3000 has always given me problems. As indicated here - and other forums - the settings on the DALY BMS was much higher then the Xantrex was set to do. Aligning the Xantrex to BMS caused further issues.
- Keeping SOC @ 100% - I have always kept the inverter in float mode and the BMS keeps accepting. LiFePo4 like to stay in 40% SOC only, unless it is actively being used.
-

Actions to mitigate the risk
1) properly securing the batteries from movement - I have purchased from Sunfun a kit that has a battery case that the cells can be put into.
2) BMS with balancing per cell - with the Sunfun kit, they have a 150A BMS with Temp sensor and will monitor and manage only 1 cell per group
3) Addition of Active Balancing board - understanding the variability in cell balancing
4) Unless I am boondocking - while I am home, or at a place where there is shore power, the batteries will be stored at 40% SOC only
5) Adding additional monitoring - while we have the monitoring of the system, we should enable alarms and keeping the system in a very tight window.

So the wife and I will be working with an installer to verify and validate the building of the cells, how it will be tied in, and set up.
I have the SunFun kit on order and will build 4 individual batteries. we also agreed to bring the cells into the van, and with the Simarine solution, I will place a shunt on each battery, with its own Temp sensor. will set up an alarm to trip my car alarm if it goes outside parameters.
we also will set the BMS to float at 40% and stop taking charge past 50% - and will no longer plug the shore power in to charge - use only Solar to keep this up (and honestly may change)
the batteries will also come into the van. Will use 3M to "Blanket" the batteries regardless to make sure they are always warm. Since I have a hot water tank that also uses the coolant system, I may just also add a coolant radiator to blow warm air over the packs.

I will keep everyone updated on how things progress.

SFK-260 Basic Kit - Built your own Lifepo4 Batteries for RV, Back up Power / Stand by power or power walls using Eve 280 Cells / RCPT / Explor cells. Sunfunkits.com​

SFK-260 Basic Kit - Built your own Lifepo4 Batteries for RV, Back up Power / Stand by power or power walls using Eve 280 Cells / RCPT / Explor cells. Sunfunkits.com
www.sunfunkits.com

Quick Nautic Boiler B3 2012SL 5 Gal 1200W w/Heat Exchanger​

Quick Nautic Boiler B3 2012SL 5 Gallon 1200W w/Heat Exchanger**Casing in composite material and Tank in Aisi 316 stainless steelBenefit * High quality materials. * Heat exchanger with wide exchange surface. * Produces hot water from the electric...

www.marine.com

 

[bgflyguy]

Do you have a picture of that corner before the failure? I am having trouble picturing what you are saying.

 

[Orion Designs]

 

[TorqueMonster72]

Top right corner and all down the side there.

 

[TorqueMonster72]

?

WTF dude…. Popcorn…. Pass it over with my belt and shoelaces FFS

 

[Orion Designs]

WTF dude…. Popcorn…. Pass it over with my belt and shoelaces FFS

 

[Rocksnsalt]

News to me that LifeP04 should be stored at 40% state of charge?

That seems unrealistic and near worthless for home standby, van use, anything.