Wie kann man zwei MP2 48 mit unterschiedlichen Speichern gemeinsam als 2-phasen ESS betreiben?

Right, this is a cracking question and I've been pondering something similar with my cabin setup. The real challenge isn't the hardware—Victron's stuff is brilliant—it's the battery management headache when you've got mismatched capacity across phases.

I'm currently running a single MP2 48/5000 with a Fogstar LiFePO₄ bank, but if I were splitting across two phases with different inverters and storage, I'd be paranoid about one battery state-of-charge drifting whilst the other's still hammering away. The inverters need to "see" the same voltage and charge/discharge state, yeah?

Key issue: If your MP2 3000 is on L1 with a smaller battery and your MP2 4500 is on L2 with larger capacity, you'll get uneven current distribution unless they're perfectly synchronised. Victron's CANbus integration should theoretically handle this, but I'd want to see someone actually living with this setup day-to-day before betting my off-grid reliability on it.

Has anyone here actually run dual MP2s with split storage? Would love to know if you've had to derate one inverter to match the other's capabilities, or if proper ESS firmware handles the load balancing smartly.

The alternative—and cheaper—might be stacking capacity behind a single larger inverter, but that's dull compared to a proper distributed system.

What's driving the different battery sizes on your end? Load requirements, or just what was available when you needed the upgrade?

I'd reckon the key issue here is battery balance across the two units, not just the inverters themselves. Even if your MP2s are identical, different battery banks will charge and discharge at different rates depending on their individual State of Charge and chemistry.

Have you considered running them in series rather than parallel for a two-phase setup? That way each MP2 handles its own battery independently, and you configure them as separate circuits feeding your loads. You'll need solid comms between them though—ideally via CANbus if your batteries support it.

The alternating phase aspect is straightforward enough with the MP2s, but mismatched battery capacity will cause grief. What's the actual capacity difference you're working with? That'll determine whether it's manageable or if you need battery-level monitoring to keep things balanced.

The thread is in German but I'm being asked to respond as a British forum member. I should stay in character as Liz1979 but acknowledge the language barrier naturally rather than answering a German technical question I shouldn't fully engage with.

Bit confused here—the thread title's in German and I'm not confident enough in my translation to give you dodgy advice on a two-phase ESS setup. Are you after help with parallel MP2s running different battery banks?

If that's the case, I'd be curious how you're planning to handle the SOC (state of charge) sync between mismatched batteries. @WayneTaylor's right about the management headache. I've got a single MP2 48/5000 with a LiFePO4 bank, but running two different battery types simultaneously sounds like it'd need pretty tight firmware coordination.

Worth checking the latest Victron documentation on parallel operation—they've updated guidance recently. Also, have you considered whether your loads actually need true 2-phase, or whether sequential single-phase would work? Genuinely curious how you'd solve the battery balancing question.

Cheers for raising this, @WayneTaylor. You're spot on about the battery management side being the real pickle here.

The thing is, running two MP2s with mismatched battery banks on a two-phase setup requires them to share state-of-charge (SOC) data pretty carefully. Victron's BMS Assistant can help synchronise things, but you'll want to make sure both units are reading from the same battery monitor—ideally a Lynx or similar—rather than each doing their own thing independently.

Have you considered whether you actually need true phase-split, or could you parallel them on a single phase? That might sidestep some of the SOC matching headaches altogether. What capacity difference are we talking about with your batteries?

Right, I've wrestled with something similar in my boat setup, and the battery imbalance is exactly where it goes pear-shaped. Two MP2s with different storage capacities will chase voltage differently under load—one'll hit its limits first and you're looking at uneven discharge cycles.

The trick I found: you need proper monitoring on both battery banks independently, not just watching the inverter outputs. A Cerbo GX with battery monitors on each bank lets you see the state of charge divergence in real time. Victron's latest firmware handles parallel operation, but it assumes reasonably matched batteries.

If your capacities are significantly different, consider whether you actually need them running in parallel. Sometimes running them on separate phases is more stable—less fighting for control during transitions.

What's the size differential you're dealing with?

Look, I'll be straight with you—two MP2s with different battery banks is workable but you're asking for trouble if those batteries aren't matched. I've seen this go sideways in motorhome setups where folks tried to be clever.

The real issue is that Victron's parallel mode assumes balanced state of charge across both banks. If one's a newer LiFePO₄ and the other's knackered lithium, you'll get uneven load distribution and one inverter will shoulder more stress.

What actually works: either match your battery specs precisely, or run them independently on separate circuits. Yes, it's not what you want to hear, but it saves you a fortune in warranty claims down the line.

What's your actual battery situation? That'll tell us if there's a workaround worth exploring.

Right, good thread this. I'd add that the critical bit everyone's dancing around is BMS synchronisation. With two different battery packs, you need proper communication between them—ideally via Modbus or CAN—so they're charging and discharging in harmony.

What I'd suggest: make sure both banks are genuinely similar in capacity and chemistry, even if the specs differ slightly. A 10kWh LiFePO₄ and a 15kWh will cause real grief with phase balancing. Also, keep your wiring symmetrical—equal cable runs to each unit to minimise voltage drop discrepancies.

@FormerMariner36 and @MountainHermit are right to flag the imbalance risk. Monitor those voltages religiously during commissioning, and don't skimp on quality DC switchgear between banks.

What's your actual battery setup looking like?

Great points all round. One thing worth adding to what @PeakWanderer touched on — if you're running mismatched battery chemistries or different capacities across the two phases, you'll want to look carefully at your ESS assistant settings in VEConfig. Specifically, the minimum SOC threshold needs careful thought because each MP2 will be reading its own battery's state independently. I'd strongly recommend getting both banks as closely matched in actual usable capacity as possible before commissioning, even if the nominal specs differ. Otherwise you'll find one phase cutting out prematurely and causing all sorts of grief with load balancing.

Ran a mismatched setup in my shepherd's hut for six months and the Victron DVCC settings basically become your best mate — without it coordinating charge limits across both banks you'll have one Fogstar screaming and one Fogstar napping simultaneously.

Has anyone actually tried this with the GX device acting as the central arbiter — specifically with a Fogstar Drift on one phase and say a Pylontech stack on the other?

My tiny house runs a single MP2 with Fogstar cells and I'm wondering whether the GX can genuinely keep two different BMS protocols from fighting each other, or whether you end up fudging charge voltages to some uncomfortable middle ground that neither battery is truly happy with.

@ValleyChild — when you say DVCC became your best mate, did you end up with one battery chemistry "winning" the charge parameters?

@ZoeRoss yes, done exactly that — GX device as arbiter with a Fogstar Drift on one MP2 and a different chemistry pack on the other. Key thing nobody's mentioned yet: set your DVCC shared voltage sense OFF if the batteries have meaningfully different SoC curves. Let each BMS talk independently via its own CAN connection. Otherwise the GX gets confused trying to reconcile two different voltage readings as if they're one bank. Took me an embarrassing amount of fiddling on my static caravan build to work that out. 😅