Thanks for the feedback… You should always make sure you have as many backup options as necessary… but it’s rare to have everything under control.
That’s why communication with the inverters is important… if a BMS ever interrupts the charging process… NO damage, just a cell with too high a voltage—the inverters MUST throttle their output or shut down… as with Victron… Otherwise, the remaining batteries will receive even more charge… That then becomes a real chain of errors.
Keep in mind that you now want to work with 3 battery banks instead of one.
What kind of BMS are you using?
Have you ever calculated your system’s maximum power output? … How much current would the batteries normally receive?
Fuses always have two characteristics: one thermal and one for short-circuit protection. If the 200A gFuse (or NH1 in Germany) trips, it means a truly critical situation has occurred. It won’t trigger a thermal overload shutdown due to excessive charging current—say, 250A—if the BMS doesn’t shut down the system. That’s what the fuse’s rating curves are for. Right now, the gFuse really serves more as a shutdown device for you as the user rather than protection against excessive charging current… except in the absolute worst-case scenario of a cell or BMS failure…
The BMS is integrated to the battery and to be honest the batteries came from China without any brand. What would the recommendation be for backup BMS communication between the inverter and the batteries considering that it is integrated into the battery. I saw that Victron has a BMS management but my understanding is that it only works with their batteries
This is a new property and there are no real load apart from a submersible 40W pump. I am new to all of this and enjoy learning so are you saying that I should let the integrated BMS be the primary failure point because it protect the battery more efficiently from overload and excessive charging?
I’ve attached a diagram showing the trip time of a gFuse (we use NH00 and NH02) to help you understand
For the 200A version, I’ve marked two scenarios for you:
Red indicates normal thermal load at 200A—trip time approx. 1000 sec!!! At 210A, the BMS would disconnect IMMEDIATELY.
Blue indicates a short circuit at 1300A – 0.01 sec
At 1300A, the fuse would trip immediately… could the battery and wiring handle that… maybe…
I will check the Diagramm for the Adler EF3 Fuse … this is for 100A !!! … you can see that at 200A would trip at 100sec !!! at 150A and less … trip time … never
Ohhhh wow … a 40W Pump and a Load of 3 huge Batteries …
Do you have a datasheet or know the manufacturer of the batteries and inverters? … If possible, I’ll take a look at them … Maybe the BMS can interrupt or reduce the charging process …
The property is under construction, mostly finishing up so no heavy equipment are in use currently but after we move in we there will be load.
The current battery wire size is 35mm2 which I doubt that this will handle the 1300A, the battery manual recommends 50mm2 but this is still undersize based on the NEC code. I didn’t know that it would take such high current to for the fuse to trip immediately, looks like I may have to reduce the fuse size but base on the first graph I can see the current is still very high for the fuse to trip immediately.
I have included the datasheet and manual for most of the equipment that I have on the DC side, from the panels, RSD, inverter, battery, PV rotary isolator, and the fuse. The SLD that I was provided is also included which shows the recommended wires and MCB/MCCB sizes. The math was not adding up to justify the 125A for the 200A max continuous load and the 225A for the parallel bus you can review it and share your thoughts.
In a LiFePo battery, a short-circuit current of over 1500 A can certainly flow… we’re talking about 0.01 seconds here… even with 35mm² …
If there’s a fault in one of the cells or between them, a fuse like that won’t trip anyway… this really only applies to a fault in the wiring UP TO the BMS… everything beyond that… Fingers crossed…
That’s definitely a very interesting setup and a great schematic… Impressive
So you only have a single-phase grid… that makes a lot of things easier, of course, but also requires larger cable cross-sections
Can you—and are you allowed to—feed power back into the grid at this high power level? …
What’s the rating of your home’s circuit breaker?
This comparison is always interesting, no matter what… It all looks like overhead lines with poles and drop wires to the house…
Do you need to install an emergency switch for the fire department?
Thanks, the single line diagarm (SLD) needs to be updated to include the Lynx DC Distribution with the gBat 200A fuse+disocnnect between the battery and the lynx DC distribution and the victron mega fuse 200A/80v in the lynx DC distribution for the inverter protection.
Yes, it is a single phase for residential unless requested but 3-phase is available if needed. We are allowed to send to the grid but the amount of paper work that is needed we didn’t bother.
The home circuit break is 30A, with regards to the emergency switch (RSD) this was implemented because of experience and we had no way to isolate the panel DC as the fire burnt through the breaker and had to cut the PV string with a pliars. Also, I saw that it is required in building in the Philippine electric code (PEC)
That all sounds very well thought out and professional… I’d be interested in how the batteries and inverters communicate… I’ll have to look into that again.
Is there a master and two slaves… who controls what and whom? What happens when the batteries are full… does everything go into the grid, or is there a “zero” feed-in with communication via the bidirectional meter?
Could you possibly use signal/dry contacts from the BMS to stop or throttle the inverter output…
As long as you’re connected to the grid, none of this is a problem because the inverters can feed power into the public grid… but in the event of a power outage, they’ll have to throttle or shut down their output depending on the battery’s SOC.
Currently, the batteries are not in parallel and this is the goal to have everything in parallel which is why I purchased the Victron lynx DC distribution x 2 for the DC busbar. Regarding the inverters there is a master and the other two are slaves, which manages the DC and throttles the PV input when the batteries are full for example, when the batteries are full at 1000hrs the PV output will drop to ~20W (because of a light bulb or two) and if we put load the inverter will allow the PV to deliver the consumption on demand. The inverter is configure to use PV, Battery, and grid (in that order), the grid only kicks in when the battery meets the threshold set and will charge the battery and supply the load. There is dry contact on the inverter but it uses the inverter communication with the BMS to trigger any action that is needed.
