Paralleling LifePO4 banks with...

[sunshine_eggo]

I'm torn on that. I like using an external shunt and shunt trip breaker (vs running current through the BMS) it's got WiFi and I can get data into my other stuff.

I don't have any Victon equipment, so I don't know if that part would help me. If you've got links, I'll read about the set up.

It was just about complete battery monitoring and some control...

 

[400bird]

Did you have other concerns except that what happened to LFP full charge voltage when connect to NMC?

Not really, but I'm all ears if you have anything that I've missed.

 

[OffGridInTheCity]

@400bird - I was thinking about using Batrium with mixed NMC and LifePo4 packs/cells....

I've used Batrium for 5+ years now and I'm up to 126 longmons so I have un understanding of how it works. I particularly like one central monitor, safety, and balance focus. Plus I've learned how to extract the data for my custom dashboard and software monitoring.


Monitoring. Batrium w/longmons (not sure about K9 approach) could easily monitor the different types of chemistries. So what if the blue bars are different in groups of 14 and 16. You know what you have and can interpret that OK. I know that each 14 blue bars above is 1 battery in my powerwall from left to right because I set it up that way.

Shunt-trip/safety. Due to the planned min/max operating range it seems like you could do 1 set of hi/low shunt-trip points that would be safe enough for the combined battery bank.

Balance? I find that my (healthy) packs do not need active (daily or top) balancing. They will drift ~40mv further apart from the last balance over 9 months and then I do a touch-up rebalance. You can see I'm running 50mv max difference at this moment in the pic above but so what with a 3.5v low and 4.0v hi operating range.

Assuming a touch up balance every 6 months... I could imagine temporarily changing the longmon network by unplugging/plugging the longmon chain + change the number of packs/cells in the software the duration of the touch-up balance of that specific chemistry. With a little prep - this change could be done in <5mins. Downside, the 'other' packs/cells would be unmonitored for a couple of weeks but it's low risk with a healthy setup.

Going further - if you did active balancing on each battery (cheap BMs or active balancer), then is seems like you'd only need to balance the overall system IF the LifePo4(s) vs NMCs drifted relative to each other but I'm not sure that would happen. IF it did.... I'm imagining the Batrium auto-level feature as capable of tightening this up and you'd have still have monitoring of all the cells/packs.

In any case - just thinking out load

 

[yabert]

Not really, but I'm all ears if you have anything that I've missed.

I was asking myself if I had correctly understand your question/concerns.
Maybe your question could have been is the the LFP will degrade faster if they abnormally stay at higher voltage/floating voltage for long time?

 

[Zwy]

@400bird - I was thinking about using Batrium with mixed NMC and LifePo4 packs/cells....

I've used Batrium for 5+ years now and I'm up to 126 longmons so I have un understanding of how it works. I particularly like one central monitor, safety, and balance focus. Plus I've learned how to extract the data for my custom dashboard and software monitoring.
View attachment 199246

Monitoring. Batrium w/longmons (not sure about K9 approach) could easily monitor the different types of chemistries. So what if the blue bars are different in groups of 14 and 16. You know what you have and can interpret that OK. I know that each 14 blue bars above is 1 battery in my powerwall from left to right because I set it up that way

K9 will look the same, it is one continuous graph. Mine is 64 cells wide.

I considered using Batrium again on the next build but the cost of the extras with the cost of the Batrium was too much for the low budget build. Plus, I have some cells coming out of the house pack that need an active balancer.

This is when I worked the settings to get balance but the cells still drifted.

 

[400bird]

I was asking myself if I had correctly understand your question/concerns.
Maybe your question could have been is the the LFP will degrade faster if they abnormally stay at higher voltage/floating voltage for long time?

Well said, that's exactly my concern. Or would anything in this set up cause faster degradation of either cell type?

 

[400bird]

Thanks for the input on the Batrium.

The NMC pack has WM5 (?) main supervisor and MM8 for the 3 (identical) sub packs.

For the LifePO4 packs, I'm pretty sure I'll need to run a completely separate Batrium set up with a Watchmon core and 3* K9's, and a dedicated Batrium shunt for that battery.

Based on my research, I can't run separate cell voltages ranges on one WM5, so I can't run all 104 cells on one monitoring platform.

I have already ordered a 3 phase, DC rated ABB shunt trip breaker, that will allow the Batrium to disconnect the LifePO4 packs from the rest of the system and also isolate them from each other.

 

[400bird]

Plus I've learned how to extract the data for my custom dashboard and software monitoring.

What are you using to extract the data? I'm using the UPD listener from Daromer, but his website is down. A quick Google makes it look like Batrium is now is hosting this on GitHub.

 

[yabert]

Well said, that's exactly my concern. Or would anything in this set up cause faster degradation of either cell type?

I don't know, but I bet you shouldn't be consern much if you don't plan to ''float charge'' the LFP cells for days.
I mean, if you discharge from 100% SOC each night, the potential damage are probably marginal.

 

[Zwy]

Thanks for the input on the Batrium.

The NMC pack has WM5 (?) main supervisor and MM8 for the 3 (identical) sub packs.

For the LifePO4 packs, I'm pretty sure I'll need to run a completely separate Batrium set up with a Watchmon core and 3* K9's, and a dedicated Batrium shunt for that battery.

I highly recommend the Core with K9's. Balance current will be lower unless you add a fan to the K9's.

Based on my research, I can't run separate cell voltages ranges on one WM5, so I can't run all 104 cells on one monitoring platform.

That would be correct.

I have already ordered a 3 phase, DC rated ABB shunt trip breaker, that will allow the Batrium to disconnect the LifePO4 packs from the rest of the system and also isolate them from each other.

If you already have the ABB then you might as well use Batrium.

One thing with the Core and Toolkit is it now has web integration with the latest firmware update if you want to view down to cell level from anywhere. I haven't enabled it as I don't need that data on my phone but some might find it useful for instance if they did not have Solar Assistant.

 

[OffGridInTheCity]

What are you using to extract the data? I'm using the UPD listener from Daromer, but his website is down. A quick Google makes it look like Batrium is now is hosting this on GitHub.

I see that Daromer's GitHub is gone! and it looks like Batrium has taken it over. Good that it's still published.

I wrote a custom PHP function to parse the snapshot report, set Batrium to export the report "Every 5 mins", then do a background process to parse the report and push to a DB for a web page dashboard.


I also did a code snippet to confirm I can pull data broadcast by Batrium on my local internet the way Daromer does, but since I already have the snapshot report parsing every value possible I continue to use that. Every 5 mins is good enough for me.

If you just want a few values, the code is not 'that bad' as the packet definitions are documented in GitHub docs - https://github.com/Batrium/WatchMonUdpListener/tree/master/docs.
For example, here's some PHP to get the System ID + Min Cell Voltage if someone wanted to trigger a shunt-trip.

A run just now shows my lowest blue bar on the powerwall Batrium "SYS8560" is 3.54v


FYI - Batrium broadcast data is *not secure* (no password or authentication needed to see it) on you're local internet.

 

[400bird]

I don't know, but I bet you shouldn't be consern much if you don't plan to ''float charge'' the LFP cells for days.
I mean, if you discharge from 100% SOC each night, the potential damage are probably marginal.

Yeah, I don't plan to float the LFP much at all. The system is charged by the sun every day and discharges some during the peak energy rate hours starting at 4 pm

 

[Semoreo]

Yeah, I don't plan to float the LFP much at all. The system is charged by the sun every day and discharges some during the peak energy rate hours starting at 4 pm

I am in the process of adding 30kwh lfp to my 21kwh nmc cells right now. My situation is the same as yours. Charge with solar, partial discharge each day during peak rates. I'm really interested in how this has gone for you.

Having the charge and discharge voltages here are really helpful. The constructive discussion was really nice too. I was worried about posting this same question that I'd get the nmc bad posts all day.

Any extra swelling? Unexpectedly large differences in load sharing? How do the lfp cells like floating up there? Does the nmc discharge much to float the lfp? Do the combined packs play well after full charge cycle?

Thanks!

 

[400bird]

I am in the process of adding 30kwh lfp to my 21kwh nmc cells right now. My situation is the same as yours. Charge with solar, partial discharge each day during peak rates. I'm really interested in how this has gone for you.

I build slowly, the LFP cells are assembled into modules of 4 and 6 (small enough for movement) and currently balancing. So, I don't have results yet, I'm still months out. I have to pour the concrete pad, build the racking for inside the battery box, buy and wire in the BMS, etc

Having the charge and discharge voltages here are really helpful. The constructive discussion was really nice too. I was worried about posting this same question that I'd get the nmc bad posts all day.

Shhh don't say NMC too loudly. I got lucky the first time

Any extra swelling? Unexpectedly large differences in load sharing? How do the lfp cells like floating up there? Does the nmc discharge much to float the lfp? Do the combined packs play well after full charge cycle?

Thanks!

Depending on your build speed, you might get answers before I do. I'm probably 6+ months out still.

 

[Semoreo]

I build slowly, the LFP cells are assembled into modules of 4 and 6 (small enough for movement) and currently balancing. So, I don't have results yet, I'm still months out. I have to pour the concrete pad, build the racking for inside the battery box, buy and wire in the BMS, etc

Shhh don't say NMC too loudly. I got lucky the first time

Depending on your build speed, you might get answers before I do. I'm probably 6+ months out still.

Ok, I guess I'll let you know . My cells have been paralleled for several months so should be fairly close to even and the top. I just assembled them into 16s in a box. Got the bms connected and need to figure out a way to connect to a jk bms with an Android to set it up. I have to add fusing too but that's pretty easy. I prefer to bms & fuse each string of cells since they're not 18650s.

Might have it connected this afternoon but I'm not sure the safest way to parallel both banks. Even a small difference in pack voltage might create a large balancing current. Don't need to start replacing fuses this soon.

 

[Ampster]

My cells have been paralleled for several months so should be fairly close to even and the top.

When you say "the top", do you mean they were charged to close to the top?

 

[Semoreo]

When you say "the top", do you mean they were charged to close to the top?

I paralleled all 16 cells then connected a cc/cv power supply to them set at 3.4v 5a for about a week.

Edit- Reason I said close to the top is because the draw after a week never went below 1a @3.4v and they settled down to 3.3v exact.

 

[400bird]

I'd go higher than 3.4 for the initial top balance, that's just the start of the knee where the cells voltage to SOC relationships starts to become apparent.

 

[Semoreo]

I'd go higher than 3.4 for the initial top balance, that's just the start of the knee where the cells voltage to SOC relationships starts to become apparent.

Well, they're assembled in their box with everything connected. So maybe I'll just hook a charger up to it and let the bms do its thing. It's supposed to be an active balancer so if i give it enough volts it should balance the cells the rest of the way if they aren't already there. I did have the cells hooked up in parallel for 3 months after the initial charge. Just no charger.