I never liked the installation fixed to the house. I case I like to move, and the new buyer would not want to have a rather expensive and complex home-battery, I just take it with me.
For that possibility I designed the system into a industrial cabinet.
Featuring:
Standing, and rolling on the floor, only 2 fixations (M8 wedge bolts) on the wall.
Optional EV charge station
DC PV charger
2 separate feed-in connections for my existing SolarEdge PV systems.
Backed up power supply for Cerbo en IP-network stuff.
Avoiding the crazy Victron suggestion to connect Cerbo with an unfused thin wire on the Lynx busbars.
Appying Victron energy meter VM-3P75CT in a nice way, a solution for the fragile coils and the thin connection blocks (1,5mm2) Victron wants to connect to 25Amp fused AC circuits.
With a decent solution to have a small 1Amp fuse in Lynx distributor.
SolarEdge are kept separate using grid-feed-in and rely on VDE4105 PV power reduction in case of grid failure and maximum SOC.
Making use of Victrons (and others) 3D models to design the whole assembly in SolidWorks.
Good point. I had my thoughts upfront, heat is always a concern.
Too bad indeed the Victron components do not have a great efficiency.
The cabinet is made out of metal and heat can be transferred to the outside. The rating of the cabinet is, according to manufacturers software (Rittal, RiTherm) is 36W/degree.
According to a Multiplus II datasheet the efficiency is 95%. I have 3x 5000W variant so this 5% is 250W per Multiplus meaning worst case scenario 750Watt heat is dissipated.
With 36W/degree this means a additional temperature of 750/36 = delta T = 21 degrees.
The environment is estimated max 29 degrees, the internal temperature is @this scenario 29+21 = 50 degrees.
This is worst case scenario leaving out little heat from cabling, additional components, but also heat transported out via cabling and metal contacts of the support beams.
Practically the temperature didn’t get this hot at all. In a regular household the individual fase load is the challenge, not the total of all fases. But one never knows what will happen in future.
End conclusion: Yes disposing of heat with the help of a fan or by natural convection is needed. For now I prefer a closed cabinet so everything stays clean and dust free. I measured once after cooking ans baking a pie on the battery, the temperature reached 39 degrees at the sealing.
I have a 3phase multiplus 2 3k ESS system with 6 mppts, and the room is 1,5mx1,5m and 2,4m high
I can tell you: it gets really warm in there fast in the spring/summer time with full PV production, without ventilation 35-40 degrees is easy, with active ventilation i have right now (300m3/h max) it never gets higher the 30
So in your enclosure heat will build up even faster,
it will circulate warm air even warmer
So my advice: get some good ventilation in there!
And maybe even in the room its installed in
Just as a note, there is no such suggestion, as that would indeed be crazy. The Cerbo OEM power supply lead is fused (3.15A, to be exact) and any aftermarket wiring would also need to be fused appropriately to the wire.
The wiring, supplied with the Cerbo, is indeed fused. But why equip the leads with the 8mm rings? The schematics I got from the Vicron dealer looks like this: (Part of it)
I also could not find a Mega-fuse-sized fuse of 3.15Amps. 30Amps was the smallest, only @ AliExpress.
Wondering how many people follow these suggestions.
Thank you for the suggestion and the information. Over here it’s getting winter. Focus on Victron performance now. I will install a recording thermometer to have some real results. I agree, with you I could become too warm.
