Evade 4.8kWh battery per 1.5kW Peak for Inverter on AC-Out

Hi,

i want to couple my Fronius Symo 10 kW inverter on the AC out.

I will get 3xMuliplus2 6k5 and 16kWh LiFePo4 Battery.

This violates the victron rule in the title.

As far is i understood this is to prevent battery overload when the net disconnects with a fully working inverter.

So i want to work around this limitation.

My plan to connect the inverter to the AC-Out2. When the net disconnects it should be disconnected. Inverter shuts down.

Then i reconnect ac-out2 and throttle the inverter to obey the 4.8/1.5 rule or better just work with the SoC.

Would this be possible?

Greetings Markus

The factor 1 rule is that the PV inverter rating must not exceed the inverter rating.
The victron is 6000KVA, 5000W. The fronius must not be capable of exceeding this.
The rule exists not just for grid failure but also for the conditions when loads switch at near full charge.
AC2 is also restricted in the power it can carry due to its relay.
Don’t bend the rules if you value your equipment and warranty.

Unless I’m misunderstanding something: the Fronius Symo 10.3 (I suppose that’s the one?) does 10kW in total, with about 3300W per phase - and that should be perfectly in line for a MP2 6k5 three phase setup, no?

As long as it can’t shunt more that 5000W into the inverter on a single phase and the batteries can deal with that then it would comply.
Your battery would be facing the full brunt of the PV inverter so must not be undersized.

@TigerQuasare

So you are correct i totally forgot the symo is three phase.

so why is there a need for reprogramming @Mad_One; frequency shifting would take care of battery overcharge when the grid is disconnected. If it is set up correctly.

Ps add a dc.mppt as.well if grid outages are expected.

The inverter ratio is 1:1 not PV to inverter ratio. So even though the Fronius is loaded it cannot produce more than its rating. The PV to PV inverter ratio is programmed into ess so its not an issue.

I am sorry for being unclear:

The factor 1 rule is obeyed.

I’m talking about the min battery capacity in 3.2.

I will obey the rule if necessary, but finding a way to get around this would be preferred.

I have a lot of ideas how to get around this, but first of all i think i have to understand why this rule really exists.

@lxonline : That is what i want to find out. As far as i have in mind it is about the following scenario:

Battery is full, Inverter runs at max, grid fails: Frequency shifting / Inverter adapting is to slow => stress for the multi &battery.

But this is all from my memory of posts i read about it. Some people said to just ignore it.

I found no good resource why this rule exists and if there are ways around it.

Another example: If this is only a problem with full battery i could normally connect the inverter on ac-in and in case of a power loss with an SoC < 90% i could could connect it to ac out. This could be done with some external relays

edit: I Just found a interesting thread: Minimum battery size for fronius AC coupled system - #14 by lxonline

There on the last post the author said it is about the charging amps.

The battery i use is this one: https://www.nkon.nl/de/amfile/file/download/file/1589/product/5836/

with 157A * 48V = 7536

This would be ok for me. The inverter currently runs at 9600W * 70% = 6720W to obey German law. I was planing to increase it in order to not cap the inverter at this limit and rather cap the feed in. But capping it to 7500W (my base load is > 500W) seems acceptable for me.

Once the grid fails, at least for me the Symo instantly drops to zero, then slowly syncs back to the MP2 offgrid and ramps up production again.

To me, the only question I don’t have a definitive answer to is what happens if production is more than AC loads while the batteries are full. I’d expect for the Symo to be dialed back (Modbus TCP?) or to be kicked offline by ramping the frequency up to 53hz?

Hi @Mad_One

I recommend adding at least one battery.
Why?

  • With 16kWh your battery will be charged before lunch time. You will waste majority of your energy when selling at negative day ahead price.
  • When charging a single 300Ah battery with 150A, your balancer will not keep the cells equalized.
  • If grid fails while battery is at 100% SoC and you are charging 7,5kW to grid, you will experience AC and DC spikes.
  • With 3xMP2 6k5 you can discharge the battery at its limit maybe even getting DC ripple warnings or shutdown. You should limit the DC current.

I think with current battery prices and no “Einspeisevergütung” while negative day ahead prices, having a battery covering 2-3 times your daily consumption is optimum, considering some 30-40% SOC buffer for blackout.

With 500W base load I guess your daily consumption is 20kWh. Though 2-3 batteries 32-48kWh) should be perfect. Than batteries also fits your 7,5 kW AC-PV and the 3xMP2 6k5.

Pls mind a small MPPT for black start.

During offgrid operation the MP2 controlles AC-PV via frequency shifting.
Current PV-inverter shall implement this feature to get grid certified.

Starting at 50,2Hz they shall start reducing generation at 40%/Hz. Above 51,2Hz they shall drop to zero.
MP2 starts with save 53Hz and adjusts the frequency based on SOC and current loads.
This is configurated in ESS assistant or for pure offgrid systems in PV assistant.

Well i am on a small budget right now. I am planning to increase the battery and also the solar. But for now i have limited resources. More than one of these banks is not in the budget right now.

I get 0.07€ Einspeisevergütung for each kWh i send to the grid. The 16kWh battery will not be empty on most nights. So economically this is the go to.

I could as described add some relays to only connect the inverter at SoC < 80% while grid is dead.

The 3x MP2 6k5 are oversized for this battery. In the first step i only buy one. Then the next 2 for the 3 phase system. This is to be save for the future. Some batteries are added easily but the Multis are not as easily changed.

My daily consumption is indeed around 20kWh in summer.

edit: mppt for blackstart is planned but for later.

@Mad_One

I had to take same decision. Pls mind all MP2 in a cluster need to have same hardware revision.

You can read about my experience at:

If you like, we can talk. Pls write to bk-ve@gmx.de

There is a reaction or response time. In the ideal world this happens quickly.
At high production, your PV inverter does not have a battery, if a large load stops instantaneously this surplus power must go somewhere, and that somewhere is your multiplus and battery.

This is why the rule exists.

To understand the background, consider the following situation: the PV inverter is at full power, supplying a big load. The Multi is in inverter mode. Then, suddenly and at once, this load is switched off. At that moment the PV inverter will continue operating at full power until the AC frequency has been increased. Increasing this frequency will take a very short time, but during that time all power will be directed into the batteries as there is no other place for it to go. This causes the following:

  • When batteries are (nearly) full, the battery voltage will spike, possibly causing the Multi to switch off in DC over-voltage alarm.

  • The same spike will cause the AC output voltage of the Multi to spike, as these two are directly related, and when the spike on the battery voltage is high and fast enough, the Multi can never regulate its PWMs down fast enough to prevent the spike on AC. This spike can damage the PV inverter, the Multi and also any connected loads and other equipment.

  • Another problem is that the Multi starts charge current protection.

  • In the best case it might switch the grid inverter off immediately by setting the AC frequency to the disconnect frequency as configured in the assistant.

Also, undersized battery packs do not last. Constantly high cycling, rates of charge/discharge will wear it out. It is a false economy and a bad idea to have an imbalanced system.
The battery is the center of the universe, spend the money there and invest in the rest later.

I think i kind of understand the reason. But i have some more questions.

If the inverter is connected on ac in and the grid fails, there is no buffer the energy can go. What happens in this scenario?

If the ac-out2 is configured to drop in case of grid failture, would this be fast enough to guard the multiplus / battery? Then i could reactivate the acout2 with throttled inverter in case the SoC if the batter is is NOT full.

The transient window between grid loss and inverter shutdown is real and is not being dismissed.

However, a correctly designed system accounts for this — switchgear ratings, cable sizing, and inverter contactor specifications all need to reflect worst-case transient conditions.
The fronius inverter’s ROCOF detection also means response is faster than a simple relay trip. The concern is valid at the design stage, not an ongoing operational risk in a properly commissioned system.

No reprogramming would need to be done everytime the grid goes out.

Get a better IT network, it is only too slow when communicattions are slow.

What i can’t quite wrap my head around:

No matter the size of your battery, eventually it will be full. And eventually, the grid will or may fail.

Then what? Does that mean, unless you do only charge to 90% all the time, you’re „doing it wrong"?

Are there any installations that are „doing it right", considering above?!

Well this is the scenario i thought would be problematic. My IT network is totally fine.

As i think i understood the problem that causes this rule i think i have a solution.

Cap battery at 90% SoC and try to move that state to the evening when no full inverter power comes. This would be good for the battery anyways.

Cap the inverter at 7500W = 7500W / 48V = 156A. This should be ok for the multi as well as the Battery.

This scenario is likely to not happen very often where i live in Germany. I would guess probably less then once a year.

I think this is a bad situation for every battery, but a bigger battery just can handle it better.

High c rating charge and discharge is bad for batteries. Heat is the main killer of lifepo4.