Expansion of the System - SMA vs Victron and AC coupling

[SilverbackMP]

I'm trying to plan out Phase II of my system.
Here is Phase I https://diysolarforum.com/threads/my-sma-system-install-so-far-and-precharge-questions.54557/page-2 consisting of 72x 370 watt pannels, three SMA 7.7 Sunny Boy Inverters. 4x SMA Sunny Islands, around 100 KWH of battery (may expand this) and REC BMS in M/S configuration. Offgrid

This currently powers my mom's house and my shop and a few other outbuildings.

I don't think I'll have enough amps (currently at just shy of 100 amp at 240v) to push to a future house build or perhaps future shop additions.

In the future, I would like to "waterfall" the existing power to other inverter setups. I am debating between trying this with more Sunny Islands which have two AC inputs (grid and generator) or Victron inverters of which some have two AC inputs. I also like the fact that I can stack more Victron inverters than four.

I also have eight (yes, 8) spare SMA Sunny Boys so I would like a setup that can AC couple with additional panels and control these with frequency shifting.

I have read that Victron can AC couple. Has anyone used them in this way? Can I limit the AC draw from the "grid" (aka my existing main Sunny Island setup up)? Will they frequency shift when needed? I know less than shit about Victron's interface (trying to figure it out via YouTube and forum searches).

Same question of Sunny Island. Can I limit their draw from the "grid?"

So that any of the sub systems only draws a max of 30 amps from the main Sunny Island setup?

That would give me 30 extra amps for immediate use or battery charging at the shop's sub system, 30 amps for my future house's sub system, and the remainder left for my mom's house and a few other outbuildings.

Does this make sense?

Would also be interested in Schneider if they are a better choice (I think they can only be stacked four deep).

Or I might bite the bullet and buy that SMA Multicluster Box, but its $15k and I am screwed if something goes wrong with it or lightning finds its way in. I'd rather have semi independent systems like @toms has hinted at.

Or if this shit is too complex, can cause too many issues, I'll just install completely independent systems (would mitigate some of the lightning threat this way too). Downside is more gensets required. And I don't mean Harbor Freight ones either. Continuous duty rated ones and they ain't cheap...even on the used market. Those damned mosquito farms (Wetland Reclamation Program) have been driving used pricing up in my area as people are buying them to drive electric pumps (I don't know why they just don't buy diesel pumps, but that's another topic).

Mainly trying to decide if I want to start collecting more SMA Sunny Islands as deals pop up or build a Victron (or Schneider) system when I need it.

Open to any additional input.

 

[StartledP]

>I have read that Victron can AC couple. Has anyone used them in this way? Can I limit the AC draw from the "grid" (aka my existing main Sunny Island setup up)? Will they frequency shift when needed?

The Victron MP will frequency shift to prevent overcharging of the batteries as long as all grid tie inverters are connected to one of the Multiplus inputs, creating a sort of micro grid, and the grid is connected via one of the other Multiplus inputs. In this config the Multiplus then controls how much power is taken from, or delivered to, the grid.

This config has an efficiency cost as you need to convert from DC, to AC and back to DC again when charging the batteries. The charger of the Victron MP is also less than its rated inverter capacity, so beware.

I would suggest reading this as it answers a lot of your questions

AC-coupling and the Factor 1.0 rule [Victron Energy]

www.victronenergy.com

 

[Hedges]

Same question of Sunny Island. Can I limit their draw from the "grid?"

Yes.

The AC2 relay can handle up to 56A continuous. You could program for 24A max (e.g. if fed from a 30A breaker.) You could program to limit battery charging to 50A DC.

You could let downstream SI charge battery from "grid" (upstream SI), also backfeed from SB to charge upstream SI's battery, or not.

If you've told it AC2 is grid and allow backfeed from Sunny Boy, you can't limit that, would exceed the 24A if too much excess power. It does not disconnect to prevent excess backfeed. I've tried that, but have not tried exceeding the max 56A limit.

If you told it not to backfeed (not sure if that is "grid as generator" or a separate no-backfeed setting), then you can have SB in excess of 56A per phase; SI will disconnect to prevent backfeed and frequency shift. Supports SB wattage up to 2x SI wattage according to manual. The current draw limits would still apply. I think so long as power is present on AC2, SI will draw what it wants to recharge batteries.

Seems to me you can have a "no backfeed" system which drains upstream SI dry to keep downstream SI recharged, does not backfeed, and each SI can have SB size 2x wattage. This appears to offer the most PV, with every SI uptream or downstream allowed 2x wattage of SB.

Or you can have one which does backfeed upstream if each SI's battery doesn't want all the current, downstream SI must have SB sized no larger than 56A per relay and the upstream "grid" SI can have no more than 2x SB as the sum of all which feed it. I think that limits a 4x SI "grid" to 48kW of SB. All dowstream SI limited to 1x SB, 4x SI island limited to 24kW.

Of course, you can also go the "Chargeverter" route to transfer power from any AC system to a battery, except better not push power upstream or it will feed itself in a loop.

You can add DC coupling, with battery shunt or with Midnight Classic + interface (if not too new a firmware in SI, no longer supported.)

I think that transformer isolated and rectified AC could be fed to SB for AC coupling. I would set frequency-watts lower in such an SB so it curtails before PV does. Should have capacitor precharge circuit to prevent inrush.

If you programmed a supervisory device, you could use relays to pick any SI cluster as "grid", connecting it's AC2 from a busbar and connecting it's AC1. That would let one high SoC cluster recharge the others, or just a low cluster. Always have "grid" off for 5 seconds to ensure all SI disconnect before connecting a new out of phase source. This could be the easiest. Without any data link programming, if one SI closes its "generator request" relay, then select one with "load shed" relay indicating > 70% SoC to play grid and give it a transfusion.