Nick Gillam avatar image
Nick Gillam asked

AC Coupled ESS minimum battery size


We are installing a AC Coupled 3 Phase ESS (Grid Connected) system with 15kW 3P Fronius connected on Output of 3 x 5kVa MultiPlus II. Were planning on installing approx 30kWh of BYD lithium, however i noticed on the Victron AC Coupling & Factor 1.0 rule page there is also a minimum battery size required. (in this instance it is 48kWh)

Can anyone explain this?

30kWh of BYD will deliver 30kW of power & accept up to 20kW of charge.
If the PV inverter was on the Input of the Multi the minimum battery size 3 x 2.5kWh modules acording to the Victron/BYD guide.



2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

6 Answers
Paul B avatar image
Paul B answered ·

Hi Nick< Not a expert here but I would suspect they are refering to the Lead Acid Battery cemistry there,

As the Lithium Phosphate batteries can deliver HIGH instant load current (within reason) I would suspect that this figure if lithium is used would be reduced by 2/3rds (best Guess)

so dont worry about the battery side

I do stand to be corected in relation to the above statement

just keep the max solar at or under the 1 to 1 rule

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

Nick Gillam avatar image Nick Gillam commented ·

Thanks Paul,

But the AC coupling guide still specifies lithium 100A/H @48v per 1.5kW of PV/inverting power so I'm still at a loss.

0 Likes 0 ·
klim8skeptic avatar image klim8skeptic ♦ Nick Gillam commented ·

That works out to a 29kwh battery?

0 Likes 0 ·
Nick Gillam avatar image Nick Gillam klim8skeptic ♦ commented ·

100ah x 48 = 4.8kwh / 1.5kw of solar x 10 for 15kw of solar =48kwh

0 Likes 0 ·
JohnC avatar image
JohnC answered ·

Yeh, guidelines are guidelines. if you want BYD, talk to BYD. There may be all sorts of factors that could bob up to affect a particular supplier's recommendations - and warranty cover.

@BYD_service may summon the forum's local BYD man..

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

byd-service avatar image
byd-service answered ·

Thank you for 'summoning' me. For grid connected systems don't mind the minimum configuration. Only in the case of backup make sure that the battery surge power can support the systems potential surge power.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

puitl avatar image
puitl answered ·

I have 3x Multiplus 2 3000 and a 14,8kWh Lifepo4-Battery with 200A BMS. (16S / 48V)

Now i would connect 9kWp on AC-Out 1.
According the minimal battery-size from Victron i would need a 28,8kWh battery.

But: 9000W / 48V = 190A
So in this case everything would be fine.

An example:
Battery is fully charged, 9kW coming from the inverter on AC-Out 1 and grid is shutting down:

I think the only problem could be that single cell-voltages can rise to fast (because of the too small battery-size) and the BMS cut off the battery befor the Multiplus and Symo reduce the power.

BUT: How risky is that with an CVL at 55V (= 3,44V per cell)?

How fast reduce the system the power? 0,5 seconds? 2 seconds?

Next question: I can connect 3000W on the AC-Out 1 of one MP2, charge-power is only about 2000W (35A) - what happens to the remaining 1000W in the moment the grid is shutting down?

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

baxter avatar image baxter commented ·
In case the grid is lost, most likely the Symo will also shutdown and restart/re-sync to the offgrid net provided by the MP2. AFAIK there is no seamless failover of an ACcoupled PV Inverter in case of grid loss.
0 Likes 0 ·
puitl avatar image puitl baxter commented ·

Hi baxter!

Here at the end someone said that the Symo does not always a shutdown:

I am interested on more answers about my question...i think that the minimum battery-size recommended from Victron is very high to be always on the "save-side" in all setup´s out there...

0 Likes 0 ·
puitl avatar image puitl commented ·
@Daniël Boekel (Victron Energy Staff)

Do you have any explanation for this?

0 Likes 0 ·
Daniël Boekel (Victron Energy Staff) avatar image Daniël Boekel (Victron Energy Staff) ♦♦ puitl commented ·
If you make a system that is outside the minimum advised size: it's your own risk. Maybe it works fine, if not: move PV to input side of inverters.
0 Likes 0 ·
leighandchris1 avatar image
leighandchris1 answered ·

Hi people, I am new to forum so please excuse if silly suggestion, but I also want to know about minimum battery sizing when using grid tied inverter in ESS off grid mode.

Wouldn't a bank of reasonably priced super capacitors placed accross the battery be capable of absorbing the initial spike from load switching until frequency shifting kicked in?

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

pierrick-ausseresse avatar image
pierrick-ausseresse answered ·

it's unlikely that super capacitors will do what you expect them to do.

LFP cells impedance is not only low, it's also relatively flat versus frequency.

The high current rating of the supercap (SC) won't help if they have higher impedance.

let's start DC (sub 1Hz) :

My LFP battery (230Ah, 16cells) has a drop of 1V between 100A load and no load (impedance = 10mΩ). That's about 2% of the nominal voltage between no current and full current.

This same 2% will be the votlage change across the SC before the battery enter protection. Thing is, 2% voltage change across a capacitor is 3.7% change of internal energy. Less than 4% of your capacitor is actually "used" during a single event load jump. Not exactly cost effective.

you would end up with an insanly large SC to actually increase the power rating of your LFP battery when you could reach the same effect at lower cost with simply more battery. As a side effect you'd get also more energy stored.

In the frequency domain it gets more complicated. There you could reduce the AC 50Hz current ripple from the battery. Due to the (once again) low impedance of LFP and the abscence of need to provide a ripple free DC current most inverter provide only little filtering of the battery current. My own Victron multi RS has about 10% ripple (Idc/Iac) at 50Hz, measured during discharge without solar with a Hall current clamp.

this current mean some added aging for the battery. Exactly how much is hard to tell but it's certainly not 10% reduced lifetime. 50Hz current is less problematic then DC current as it doens't cause much diffusion (AKA charging).

There, you'd need 1F at 60V rating to reduce the AC current by 50%. 1F of capacitance is not that expensive. For a load that takes 200ms to start you'd then need 10F and 100F for 2s. 100F is not cheap but 10F would still do for many loads. Maybe / why not ?

Last warnings :

- in the frequency domain you will care a lot about the SC ESR. reaching 10mΩ (Zc + ESR) impedance seems easier on paper than it is in reality.

- you seriously do NOT want your BMS to re-engage on a discharged super capacitor. That would call for some fancy currents.

- I wouldn't spend money on reducing 50Hz AC currents without more understanding of the Lithium chemistery. No one know for sure which kind of charge profile best extend LFP life and not all AC currents are bad :

all of that can change if you use lead batteries.

TLDR : don't use Super Capacitor with LFP. save you money for something else.

2 |3000

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

janieronen avatar image janieronen commented ·
@Pierrick Ausseresse Much appreciate your educational response to the topic.

In real life have been so much easier working with LFP than lead-based systems, but as like to learn more about LFP and switching power electronics do you know does higher switching frequency have measurable implications to battery?

Heard that RS-series inverters have higher frequency than transformer-generation(20kHz), but could not find data to prove that. Any case I am about to purchase RS multis as they can be connected in three phase configuration and will buy also differential probes for scope to learn more.

Three-phase setup in my mind is better for battery as 50Hz ripple will be balanced more evenly and no zero current times between sine waves. I could be wrong so like to test that as well. AC-side waveform looks good with energy-quality meters but that DC-bus and specially LFP side reaction to PWM swathing is interesting topic.

0 Likes 0 ·
pierrick-ausseresse avatar image pierrick-ausseresse janieronen commented ·
I do not know exactly the effect of HF current on a LFP battery.

but I do know one thing, it is extremely unlikely that the "low frequency" victrons have no switching regulator on the DC side. You need something to get the DC current into an AC wave.

In my opinion the difference between the multi RS and the 50Hz transformer variant is the frequency "across" the isolation barrier. Older multi must be creating the 50Hz waveform on the DC side and have a 50Hz transformer sitting across the isolation. The multi RS must have 20kHz across the isolation and demodulate the 50Hz wave on the AC side. This requiere switching elements on both side of the insulation instead of the DC side only : more complex.

what does it means ? from the DC side you're likely not to see any difference. If you do, you will see it on the AC output.

disclaimer : I am NOT a victron engineer nor did I open a multi to see the insides. it's just my own understanding of power electronics.

for 3 phases : the multi RS do not have 3 phase operation yet (should arrive Q4 2023 on the dual tracker multi). Balanced 3 phase is better for 50Hz ripple (there is none) but that is, IF you have balanced 3 phase. You never have balanced power in a regular house, even if you're still better that with single phase.

0 Likes 0 ·