House burned down

[Brucey]

Water cooled diesel definately seem like the way to go for long term durability and fuel efficiency.

So far so good with my open frame inverter Predator 8750W, but its massively loud and obnoxious.

 

[SilverbackMP]

Water cooled diesel definately seem like the way to go for long term durability and fuel efficiency.

So far so good with my open frame inverter Predator 8750W, but it’s massively loud and obnoxious.

Our primary generator is a 30KW Ingersol Cummins powered unit I picked up used six or seven years ago.

 

[Shimmy]

High or low voltage cell limit (or any other critical fault you program such as battery temp) is hit, the Batrium shuts down the whole system using the shunt trip ABB breaker. Some might complain, "Well, it shuts down the whole system".

That is what I want, it forces me to go and find the problem. It has happened when I played with charging voltage limits and had a runner. It works exactly as designed
...No need, one shunt trip breaker on the whole battery bank with a Class T fuse on each 16S (or whatever series you are using) battery. Each 16S battery in my case has a K9 that monitors and balances each battery and relays cell, balancing and temp data to the CORE.
...Let's get realistic here. Attempting to find fault with using only a Class T fuse for overcurrent protection is pointless, just like having redundant overcurrent protection in your house electrical panel would be.

I wish I could find a more polite way to say this, because you have a lot of great insight and experience... but you are incorrect.

Once a cell goes out of tolerance tripping the bank will not stop it from charging via the parallel strings. It may not lead to a catastrophic failure, but the BMS is unable to protect the battery from damage.

Your example of fuses in series is poor; if the two fuses provide the same protection then they are very much redundant. Circuit breakers in series with fuses does actually happen in high fault current situations-- the fuse is used for instantaneous protection in the maximum fault current range, but the circuit breaker maintains coordination of the system short of a bolted fault. The difference is those are UL-listed series-rated combinations.

Which can provide more current to a damaged cell, a charge controller or 3+ strings of battery in parallel?

 

[Zwy]

I wish I could find a more polite way to say this, because you have a lot of great insight and experience... but you are incorrect.

I use the Batrium, I know what it can do and can't do.

I can't say the same for you. Maybe you have used it or maybe you haven't.

Once a cell goes out of tolerance tripping the bank will not stop it from charging via the parallel strings. It may not lead to a catastrophic failure, but the BMS is unable to protect the battery from damage.

The BMS doesn't need to, that is the purpose of the Class T fuse if current is high. Other than that, either that cell hits the high voltage disconnect setting and the Batrium trips the shunt trip. If the battery cell fails and goes below the low voltage disconnect, it will trip the shunt trip breaker.

Your example of fuses in series is poor; if the two fuses provide the same protection then they are very much redundant. Circuit breakers in series with fuses does actually happen in high fault current situations-- the fuse is used for instantaneous protection in the maximum fault current range, but the circuit breaker maintains coordination of the system short of a bolted fault. The difference is those are UL-listed series-rated combinations.

So you have fuses after your electrical panel for the same circuit? Show us the photos, I'd like to see it.

Which can provide more current to a damaged cell, a charge controller or 3+ strings of battery in parallel?

The 3 strings and high current will blow the Class T. My busbars and cable are rated way higher than the Class T. This is on purpose, the Class T is the weak link and will blow before any wire melting. Proper design will mitigate failures.

 

[Hedges]

I have a 100A fused disconnect for PV, after 200A service disconnect breaker. The smaller wire gauges require lower amperage OCP, but I think the fuse will be faster with a dead short.

I considered coordinating fuses with breaker for the service.
I settled on 400A class-T fuse in series with 200A breaker, because curves indicate breaker is guaranteed to go first for moderate overload (1000A or so), while fuse is guaranteed to go first for extreme current events.
The idea was to protect breaker, not have to deal with PG&E it it needed replacement.

I ended up not installing it due to cramped space inside the box.
I also rediscovered that the main breaker, which is something like Challenger/Zinsco, is simply plugged on and retained with a screw. It can be replaced hot.



As for Batrium, if there is a shunt-trip breaker or other disconnect per battery string, that would let it prevent further damage.

I would not rely on just a fuse or regular breaker per string. One bad string could be seriously over-charged by neighboring strings dumping current into it, without blowing the fuse, if some cells shorted.

If BMS can't interrupt the current, it can't do anything about it.

 

[robbob2112]

I've read the batrium docs and design examples and looked at the different builds. I haven't run it, but I might in future.

I think there is validity to what Hedges and Shimmy are trying to say.

If the batrium detects a faulted cell inside a single battery string there should be a way to shut off the single string without letting the other strings dump current into it. If that means a contactor per string or a shunt trip per string in addition to a class T per string I think it would be worth the effort.

With that you could also compare current per string and if any went out of bounds from the others as an early warning sign you could trip the out of bounds string offline and send a warning out.

In the case of this issue there was a large current spike and voltage drop then 20 minutes later everything faulted. That from the Victron perspective. Were there a shunt trip per string it might have been able to trip the faulted string out and prevented the issue. Might have made no difference at all.

 

[Brucey]

I've read the batrium docs and design examples and looked at the different builds. I haven't run it, but I might in future.

I think there is validity to what Hedges and Shimmy are trying to say.

If the batrium detects a faulted cell inside a single battery string there should be a way to shut off the single string without letting the other strings dump current into it. If that means a contactor per string or a shunt trip per string in addition to a class T per string I think it would be worth the effort.

With that you could also compare current per string and if any went out of bounds from the others as an early warning sign you could trip the out of bounds string offline and send a warning out.

In the case of this issue there was a large current spike and voltage drop then 20 minutes later everything faulted. That from the Victron perspective. Were there a shunt trip per string it might have been able to trip the faulted string out and prevented the issue. Might have made no difference at all.

Basically the safe option would be to have a watchmon and shunt for each string. Obviously there's significant savings going with just one and multiple strings but also potentially some risk.

We don't know for but it's prob likely that the watchmon detected the imbalance in the cells and triggered it's shunt. However the system configuration didn't allow for current to be interrupted between the 7 strings. The shunt could only stop the current to the bus and inverters.

 

[SilverbackMP]

In the case of this issue there was a large current spike and voltage drop then 20 minutes later everything faulted. That from the Victron perspective. Were there a shunt trip per string it might have been able to trip the faulted string out and prevented the issue. Might have made no difference at all.

REC does this. For under or over. In M/S config.

 

[OffGridInTheCity]

If one wanted a single BMS to shut off an individual battery(s) within an overall powerwall..... I don't know of any that can do this out of the box.

In practice, if one has a serious enough problem with a battery within a powerwall that it needs to be shut-off, it's reasonable to shunt-trip the whole battery which forces manual action to see what's going on before continuing operations. Manual breakers or a wrench to disconnect a wire can be used to restore operations with the battery taken off line. This has the advantage requiring attention to make a decision as apposed to auto-disconnecting a battery within a powerwall and make it easy to have a problem expand

However, no judgements from me!

If one wants to use a single Batrium to manage individual batteries, here's some suggestions....

1) Parse the WiFi Data.....
Batrium data can be captured live on you're local internet - the specs are published as the "WatchMon - WiFi UDP Protocol" here - https://wiki.batrium.com/en/faq/native-canbus-protocols One could have 1 x Batrium Core and read the data and trigger an individual battery shunt-trip based on data for the cells in that particular battery.

Here's an example PHP to find my minimum pack voltage.... it's not *hugely* difficult.


Here's a run where it found the min voltage of 3.630v from my SYS8650 unit (I have multiple Batriums)


2) Parse the Snapshot Report
Alternately... (this is what I do for my integrated web based dashboard) one can auto-dump the snapshot report every 5 mins and parse the text file to extract data to make custom decisions. The data is easy to read - no byte level specs needed.


---- Side Affect of disconnecting a subset of cells for active charge/discharge ----
You now have a group of static cells and others that are actively changing due to charge/discharge. Off the top of my head, the main issue would be to turn off active balancing. I don't use Bypass or Auto Level for daily operations as it's not needed except for touch-ups every 9 months - so this wouldn't be an issue in my case.

 

[Zwy]

As for Batrium, if there is a shunt-trip breaker or other disconnect per battery string, that would let it prevent further damage.

I would not rely on just a fuse or regular breaker per string. One bad string could be seriously over-charged by neighboring strings dumping current into it, without blowing the fuse, if some cells shorted.

I don't have a problem with the design or the setup. I designed my system for when a cell might possibly short internally and sleep very well at night.

Someone made the comment the BMS is what caused the fire here and a bunch of members who never ran a Batrium jumped on it. Then when that didn't work, now it is attempting to say the Class T fuse won't interrupt the circuit, after I asked if the house electrical system has 2 circuit breaking devices .

Talk about constantly stretching.

If BMS can't interrupt the current, it can't do anything about it.

It doesn't need to. Does your house electrical system tell the power plant to shut down if you have a fault in your house?

Rhetrorical.

I think some here need 5 fuses and 3 BMS's per battery to ensure that this is shut down. Make it 6 fuses and 4 BMS's, an even 10 possible circuit disconnects because 5 fuses might not break the current and the 3 fet based BMS's didn't work either, but melted all together.

It isn't ridiculous now is it? Afterall, more is better and more after that is even better yet.

This is what you sound like reading the comments.

Rant over.

 

[Zwy]

I would not rely on just a fuse or regular breaker per string. One bad string could be seriously over-charged by neighboring strings dumping current into it, without blowing the fuse, if some cells shorted.

Send me all your Class T fuses and I will test each one for you and send the fuses back when I'm done with the results.

 

[D71]

How many of you guys fall under

National Electrical Code (NEC) 2020​

2021 INTERNATIONAL SOLAR ENERGY PROVISIONS (ISEP) | ICC DIGITAL CODES

ICC Digital Codes is the largest provider of model codes, custom codes and standards used worldwide to construct safe, sustainable, affordable and resilient structures.

codes.iccsafe.org

Are your batteries UL compliant. Lead acid are allowed with rules all other batteries have to be UL compliant. IEEE 1145 nickel cadium

If live under newest code and your home burns down than your home made batteries and non UL system might mean you be SOL. With insurance. Just a warning. You might even be brought up on charges ….

Mine - tiny system are hobby purpose ….. emergency use ….
Most UL batteries are server rack style with bms and communication - built in BREAKER that covers the battery… the problem is with a rack of batteries - breakers need a final fuse or breaker at the end because of them being in parallel. Whoops. String have to be opened up because batteries in parallel are still passing power.—-leveling. Strings have to isolate.

OP had fuses for his batteries and FUSES for his inverters as recommended by Victron.

Does your local code follow

National Electrical Code (NEC) 2020​

Is Batrium UL?

 

[upnorthandpersonal]

Off topic, but I find it funny that it's called 'ICC' (International Code Council) and that it creates the IBC (International Building Code), when it only applies to the US.

 

[Partimewages]

We're special!

 

[robbob2112]

I don't have a problem with the design or the setup. I designed my system for when a cell might possibly short internally and sleep very well at night.

Someone made the comment the BMS is what caused the fire here and a bunch of members who never ran a Batrium jumped on it. Then when that didn't work, now it is attempting to say the Class T fuse won't interrupt the circuit, after I asked if the house electrical system has 2 circuit breaking devices .

Talk about constantly stretching.


It doesn't need to. Does your house electrical system tell the power plant to shut down if you have a fault in your house?

Rhetrorical.

I think some here need 5 fuses and 3 BMS's per battery to ensure that this is shut down. Make it 6 fuses and 4 BMS's, an even 10 possible circuit disconnects because 5 fuses might not break the current and the 3 fet based BMS's didn't work either, but melted all together.

It isn't ridiculous now is it? Afterall, more is better and more after that is even better yet.

This is what you sound like reading the comments.

Rant over.

Maybe you should rename yourself "The Batrium Defender" or "The Batrium Ranter" ...You and @OffGridInTheCity can have a cage match to fight over them... I'll sell tickets and buy a batrium for myself.... just as an idea

But seriously, what would be wrong with something that is controllable to shut down single strings? With regard to DIY batteries I imagine anyone that was serious enough to build them isn't going to ignore any sort of warning flag and will proactively check on things. Preferably an answer without the rant since it is a serious question.

Note - I have a background in network and equipment monitoring currently. Isolate a problem piece of gear at any sign of issue and either fix the issue or adjust the detection method to be more accurate. That doesn't mean being complacent and ignoring issues, it means automating to allow more time to devote to other issues.

I am not specifically talking about this particular issues. I am talking about large parallel battery banks be it batrium, traditional BMS or other types. Signals from the individual strings/cells to be interpreted and acted on. It is unlikely that the events will spread to other strings unless they happen so rapidly the overall shunt trip/breaker can't even trigger in which case the individual cutouts won't trip either.

 

[D71]

Maybe you should rename yourself "The Batrium Defender" or "The Batrium Ranter" ...You and @OffGridInTheCity can have a cage match to fight over them... I'll sell tickets and buy a batrium for myself.... just as an idea

But seriously, what would be wrong with something that is controllable to shut down single strings? With regard to DIY batteries I imagine anyone that was serious enough to build them isn't going to ignore any sort of warning flag and will proactively check on things. Preferably an answer without the rant since it is a serious question.

Note - I have a background in network and equipment monitoring currently. Isolate a problem piece of gear at any sign of issue and either fix the issue or adjust the detection method to be more accurate. That doesn't mean being complacent and ignoring issues, it means automating to allow more time to devote to other issues.

I am not specifically talking about this particular issues. I am talking about large parallel battery banks be it batrium, traditional BMS or other types. Signals from the individual strings/cells to be interpreted and acted on. It is unlikely that the events will spread to other strings unless they happen so rapidly the overall shunt trip/breaker can't even trigger in which case the individual cutouts won't trip either.

He has had good luck with Batrium - it works for him. The OP had it too. Wasn’t so fortunate.

I had bms go nuts then clear itself while tied in parallel. Just all on it’s own - was sent replacement. Had one same batteries that stopped a runner and opened the battery from charging via bms.

The victron smartsolar 150/35 is a bit to aggressive tripped it. While back a guy smoked his battle born batteries he is a youtuber on an island. Same charger think. So…. I throttled back the 3.65 to 3.45 and everything went great. VIP Bms saved the day

Battery was readjusted charge 3.45 been fine ….








Andy’s offgrid garage mentioned something about victron going to 3.45v to charge in a video but don’t remember which one of videos it was. It was a battery build video.

 

[D71]

Off topic, but I find it funny that it's called 'ICC' (International Code Council) and that it creates the IBC (International Building Code), when it only applies to the US.

They did that as part of a WEF GLOBAL DEAL to my understanding. Those are however the reasons I won’t put big money into a system.

I don’t want code enforcement breathing down my neck demanding entry into my home nor have my insurance denied. They tell on you in my area. So

OP was lucky in older post he stated his belgium insurance covered him. I wonder are they still covering him. Why it is important get things in writing for legal purposes…from insurance and such

 

[robbob2112]

He has had good luck with Batrium - it works for him. The OP had it too. Wasn’t so fortunate.

I had bms go nuts then clear itself while tied in parallel. Just all on it’s own - was sent replacement. Had one same batteries that stopped a runner and opened the battery from charging via bms.

The victron smartsolar 150/35 is a bit to aggressive tripped it. While back a guy smoked his battle born batteries he is a youtuber on an island. Same charger think. So…. I throttled back the 3.65 to 3.45 and everything went great. VIP Bms saved the day

Battery was readjusted charge 3.45 been fine ….

View attachment 214115


View attachment 214112
View attachment 214113

View attachment 214114
Andy’s offgrid garage mentioned something about victron going to 3.45v to charge in a video but don’t remember which one of videos it was. It was a battery build video.

I am still in the planning stages and intend a separate building for electrical and that to be sub-divided in to at least two compartments. One for the inverters and MPPT and the like. And one or two for battery banks. All three in as near to flame proof a building as I can afford. Which probably means steel studs verse wood, firerock, durock on all walls and ceilings. Baffled pass-thoughs for the cables to keep a fire from burning through the partitions. View ports to see what is going on inside without opening a door, an extenguiser port, a manual pull extenguisher inside the room, and relay controlled fans and vents to use the excess heat from the inverters to heat the batteries in winter and exhaust fans to cool everything in summer.


All that said - with a basical every precaution done if I wanted a large battery bank to provide say a week of power in a rainy/snowy season without a generator I wouldn't want the whole thing to trip off because a single battery string had an issue.

I really like the monitoring capability of the batrium system, but I also really like the ability to trip individual strings offline for maintenance and inspection.

 

[D71]

I am still in the planning stages and intend a separate building for electrical and that to be sub-divided in to at least two compartments. One for the inverters and MPPT and the like. And one or two for battery banks. All three in as near to flame proof a building as I can afford. Which probably means steel studs verse wood, firerock, durock on all walls and ceilings. Baffled pass-thoughs for the cables to keep a fire from burning through the partitions. View ports to see what is going on inside without opening a door, an extenguiser port, a manual pull extenguisher inside the room, and relay controlled fans and vents to use the excess heat from the inverters to heat the batteries in winter and exhaust fans to cool everything in summer.


All that said - with a basical every precaution done if I wanted a large battery bank to provide say a week of power in a rainy/snowy season without a generator I wouldn't want the whole thing to trip off because a single battery string had an issue.

I really like the monitoring capability of the batrium system, but I also really like the ability to trip individual strings offline for maintenance and inspection.

I think the batrium looks great for specs. 200+ cells monitoring is fantastic.

I considered this 10x10 shipping container but there are no bargains - wood floors.


Put some panels on sides and top.
The 10x10 cost is about 10x20 which is about cost of 10x40.

I’ve got breaker with lower rating then fuse as final fail safe for my batteries. My Batteries are manufactured packs in metal locker. Each is 12 volt heaters and blue tooth…4 bms. Not UL mine are emergency. If buying again would of bought UL everything.

 

[robbob2112]

I think the batrium looks great for specs. 200+ cells monitoring is fantastic.

I considered this 10x10 shipping container but there are no bargains - wood floors.
View attachment 214116

Put some panels on sides and top.
The 10x10 cost is about 10x20 which is about cost of 10x40.

I’ve got breaker with lower rating then fuse as final fail safe for my batteries. My Batteries are manufactured packs in metal locker. Each is 12 volt heaters and blue tooth…4 bms. Not UL mine are emergency. If buying again would of bought UL everything.

@Hedges mentioned some of the midnite breakers can be remote tripped. Those seemed really interesting for some of this sort of thing.

I really like the 10x10 shipping container... The wood floor could be removed then pour concrete into it from the outside, or leave the wood and put durock and ceramic tiles on the floor. So it would become part of the foundation. I may have to look into those when the time gets closer.