75kwh of EG4 LifePo4 storage draining overnight with 1-2kw load.

[Shazbot]

I've had 3 incidents of almost completely full batteries draining sometime at 9-12 o'clock at night.
Twice when I had a Tesla 120v slow charge going, once when I had my small ceramic kiln on.
I woke up to drained batteries, until the point the batteries turned off all lights.

I'm looking for guidance on diagnosis steps.

System designed using Will's 48V system as a template.
* 8 BlueSun 455W Bifacial Solar Panels, plan to add 28 more panels once my structure is up.
* 2 EG4 6500-EX Inverters
* 3 Racks, 6+6+4 48v EG4 Batteries
* 240 volt panel.

Current thoughts on further diagnosis:
* As it seems to be occurring at night only, this may be correlated with battery DC=>AC conversion draw rather than PV DC=>AC conversion, but that's just a guess.
* I could add shunts to various locations to track volt/amp when the next event happens.
* I could setup the inverter monitoring app to see what's happening when the next event happens to get better insight into what's happening.
* I could check all electrical connections just in case there's some connection issue.

Any info on why this might be happening or guidance as to how to diagnose this would be much appreciated.

 

[42OhmsPA]

A shunt on the main battery negative would be extremely beneficial. A clamp meter that can read AC and DC is also a great investment if you don't have one already; make sure you get one capable of reading inrush current, it will come in handy.
Have you taken into account of the idle consumption of the 6500ex's? Iirc they will burn over 100w/h just being powered on.
I suspect with your current amount of limited panels your not getting anywhere near a full charge but it's hard to say without data.

Edit to say the watchpower app is pretty frustrating and worthless, from my experience. I'd recommend Solar Assistant but that has been quite the debate with EG4 inverters...

 

[Shazbot]

Thanks for the response, much appreciated.

I have a clamp meter, klein cl390, that I think should be able to capture inrush. That would help me capture a peak draw... The kiln does do periodic heating cycles, I'm not sure about the car charger. I'm guessing that a spike per cylic draw could be registered and perhaps attributed to an extreme battery draw?

I don't think the idle consumption of the 6500ex's should be impactful. They haven't caused a problem on other nights since it's been operating since June. Even 200w/h X 10 hours = 2kwh wouldn't impact the 70+kwh available... unless it's somehow contributing to a parasitic drain of some sort that would bring the batteries to ~0 charge.

The current panels definitely don't give me a full charge, it's about 10-15kw max per day (minus normal daily usage), so I should be back at full in a week and a half. I only use the kiln/car charge when I have full-ish batteries for the time being.

If it's helpful, the car charger uses 1.2kw/hr and the kiln uses 2.3kw/hr, so even if they're both on along with the fridge, 6500ex draw and misc over a 10 hour period, I can't imagine more than 47kwh being used and I've never had both running simultaneously.

I got a raspberry pi, but haven't gotten it setup quite yet.

I haven't setup a shunt before - you note on the "main" battery - I have 16 batteries in 3 racks... Can you clarify what the main battery in this scenario would be?

 

[rhino]

I haven't setup a shunt before - you note on the "main" battery - I have 16 batteries in 3 racks... Can you clarify what the main battery in this scenario would be?

What is the negatives to all 3 racks connected to? The shunt just needs to be inline with the negative of all 3 racks.

 

[Zapper77]

Are all 16 batteries working?

 

[Shazbot]

What is the negatives to all 3 racks connected to? The shunt just needs to be inline with the negative of all 3 racks.

They're all connected to a negative bus bar before splitting to the 2 inverters.

 

[Shazbot]

Are all 16 batteries working?

They appear to be, they each all charge back up to full 4 green dots.

 

[Zapper77]

They appear to be, they each all charge back up to full 4 green dots.

Can you pull up the data using BMS Test on your computer.

 

[millsan1]

How are the 3 racks wired to the inverters?

 

[Shazbot]

Each rack is connected to a pair of bus bars, negative and positive, via 4/0.

The bus bars each connect to the two inverters via 2/0 through fuses. Here’s a link to the design I used. https://www.mobile-solarpower.com/48v-complete-system-blueprint.html

 

[millsan1]

Each rack is connected to a pair of bus bars, negative and positive, via 4/0.

The bus bars each connect to the two inverters via 2/0 through fuses. Here’s a link to the design I used. https://www.mobile-solarpower.com/48v-complete-system-blueprint.html

On the racks themselves, do you have the positive on top or bottom, and the negative on the opposite?

 

[Shazbot]

On the racks themselves, do you have the positive on top or bottom, and the negative on the opposite?

ah, they’re opposite, but there is a difference.

2 racks of 6 have negative connections coming out of the top (right), positive out of the bottom (left).

1 rack of 4 has the negative connection coming out of the bottom (right), positive out of the top (left).

That sounds like it shouldn’t be that way aesthetically, but I’m unclear why it’d be incorrect electrically.

Regardless, it seems reasonable to change the 4 rack to align with the 6 racks.

 

[42OhmsPA]

ah, they’re opposite, but there is a difference.

2 racks of 6 have negative connections coming out of the top (right), positive out of the bottom (left).

1 rack of 4 has the negative connection coming out of the bottom (right), positive out of the top (left).

That sounds like it shouldn’t be that way aesthetically, but I’m unclear why it’d be incorrect electrically.

Regardless, it seems reasonable to change the 4 rack to align with the 6 racks.

That should be fine, as long as the + and - are opposite of each other on each rack. This allows current to flow evenly through each battery.

Can you pull up the data using BMS Test on your computer.

Agreed, this would be a great starting point to make sure every pack is working.

 

[labeeman]

I did a little math and I think you are not fully charging your batteries with only 8 panels and on bad sun days it gets worse. On a good sun day I think you are only getting 3kw of power an hour and i think that is on the high side.

 

[Zwy]

I did a little math and I think you are not fully charging your batteries with only 8 panels and on bad sun days it gets worse. On a good sun day I think you are only getting 3kw of power an hour and i think that is on the high side.

While you are probably correct about not reaching full charge, a good sun day with 4 hours of sun would yield around 14Kwh of power. Some of that power would be used to power loads during the day. OP mentioned this here, "The current panels definitely don't give me a full charge, it's about 10-15kw max per day (minus normal daily usage)," so his yield is about what I calculated.

OP needs a shunt and possibly something like Solar Assistant to monitor loads/SOC.

 

[labeeman]

While you are probably correct about not reaching full charge, a good sun day with 4 hours of sun would yield around 14Kwh of power. Some of that power would be used to power loads during the day. OP mentioned this here, "The current panels definitely don't give me a full charge, it's about 10-15kw max per day (minus normal daily usage)," so his yield is about what I calculated.

OP needs a shunt and possibly something like Solar Assistant to monitor loads/SOC.

Yep he does need more instrumentation in order to know what is going on. I just went through 7 days where I only harvested 5kwh for the whole 7 days I have 80kwh of storage.

 

[hwy17]

What are the battery temps doing at this time? I had my Lifepower4 suprise shut off on me recently when I thought it was about 70% charged, and I suspect low temps were involved. Not temp sensor cutoff, but low cell temp dropping the voltage.

Otherwise inaccurate SOC (probably involved in my case too) and buy a generator so you can fully charge.

 

[Shazbot]

What are the battery temps doing at this time? I had my Lifepower4 suprise shut off on me recently when I thought it was about 70% charged, and I suspect low temps were involved. Not temp sensor cutoff, but low cell temp dropping the voltage.

Otherwise inaccurate SOC (probably involved in my case too) and buy a generator so you can fully charge.

I’m unclear on what the battery temps were. I agree with others I need more instrumentation.

Ambient temp would be around 60 F.
I harvested 12kw today, 5kw spent.

I’ll run the laptop to battery verification just to confirm that each is reading out something sane and figure out the app.

I’ll try to trigger the event again once charged.

Thanks all for the advisement.

 

[Skypower]

like others have mentioned, you need one big shunt. I’d probably go with a 500 amp shunt at one “choke point”. Victon Smart shunt has a nice stand alone app or optional display and/or it can integrate easily into into Solar Assistant. This way you Don’t have to worry about Solar Assistant vs EG4 potential connection issues. The Smart shunt Bluetooth range is short. The thing about a good shunt is that you know exactly what you got and anticipated time at present rate of charge/discharge to full or empty. No guessing.

 

[Zapper77]

I’m unclear on what the battery temps were. I agree with others I need more instrumentation.

Ambient temp would be around 60 F.
I harvested 12kw today, 5kw spent.

I’ll run the laptop to battery verification just to confirm that each is reading out something sane and figure out the app.

I’ll try to trigger the event again once charged.

Thanks all for the advisement.

If you want to get real nerdy...you can download BMS log data for the last 3-5 days. And error/alarm data for who know how long.