Sunny Island Sunny Boy AC Coupling Grid Tie System

[OzSolar]

Another guy with a DC Power Solar Trailer thread, sorry.

I've spent hours searching this forum, the web, SMA's website and SMA's manuals and keeping striking out on find detailed instructions or even basic SLD's.

I've even did a "control f" in both the Sunny Island and the Sunny Boy manual for "AC Coupling". The only thing I found was in the SI manual said "The AC side connection between loads, generators and storage devices.". Perhaps I'm not using the correct search terms?

The main thing I've leaned is @Hedges and others are both very knowledge and patient. But where did they learn this? Anything I'm finding on SMA's website is 5 to 10 years old and not very helpful.

I did an off grid AC Coupled Sunny Island Sunny Boy system a long time ago so I feel like I get the general concept. It was simple compared this grid tied with a existing auto start generator/ automatic transfer switch set up that I'm trying to do now. Probably overthinking it as usual.

I already have a few SLD's put together but I don't want to waste any ones time before I ask where is the secret vault of AC Coupling knowledge kept?

Thank you

 

[BentleyJ]

Your post brings back memories. I had the same problem about 4 years ago. I knew what I needed/wanted to do as far as adding battery storage to an existing grid-tie system but lacked knowing the difference between DC coupled and AC coupled so my searches didn't bring up anything relevant. I finally found (not sure how) some white papers from Schneider and Outback that were fairly detailed. Don't know if it will help, I attached the Schneider doc.

 

[OzSolar]

Thank you BentlyJ. It's nice to know that I'm not the only one who is needing more clarity with the finer details of AC Coupling.

That's first sample diagram I've that mentions the need to interlock generator. I knew there had to be someway to keep the AC coupled solar from seeing the AC Input while the generator was running but had hoped for something more elegant. Not totally sure what that would be of course....

Still sort of a head a scratcher. That interlock makes sense to me for an inverter with only one AC input (like my Sunny Islands) but it seems odd that the Schneider XW+ with it's two AC inputs would need this. What scenario would you ever want to sell back to AC Input 2 which in my limited experience always has belonged to the generator?

Thanks again for unlocking one more piece of the puzzle for me!

Or did I just mange to prove I have no idea what going on?

 

[BentleyJ]

Or did I just mange to prove I have no idea what going on?

You proved that you are smart enough to know when you need to do more research.

 

[Hedges]

Unfortunately, "@Hedges" doesn't send me an email, the way responding to a thread I've posted on does. I tripped over this thread looking at "What's new"

I'm not sure what the problem is you're trying to describe. I guess that's the problem.
Now that it is an adopted standard, many inverters meeting UL-1741-SA with "frequency-watts" option should work on AC1 output of Sunny Island. That's AC coupling.
SMA's latest Sunny Boys can be used that way either on or off grid. If strictly off-grid, the can also be set to an "off grid" mode which tolerates wider voltage and frequency variation, better for use with a generator.
SMA's earlier Sunny Boys use RS-485 and Sunny Island tells them when to obey tighter grid specs and when grid is disconnected and they can relax specs.

If you use both grid and generator, need a transfer switch (or generator interlock) of course. Each source must be off or disconnected for 5 seconds before connecting the other, so if Sunny Boy or Sunny Island is backfeeding the grid, it doesn't suddenly find itself driving out-of-phase into the new source. You also need a signaling contact to tell Sunny island which, so it avoids backfeeding generator and relaxes specs.

The only definition of "SLD" I can find is "Specific Learning Disorder". Did I guess correctly?

Because I'm using 10 year old Sunny Island and Sunny Boy, the 10 year old documents are fine for me. But I do wish they put out better descriptions & diagrams of how things connect and operate. There are new models and new monitoring devices.

 

[OzSolar]

I'm not sure what the problem is you're trying to describe. I guess that's the problem.

Thanks for responding. Not a problem per se rather a (personal) challenge in that I'm trying to find clear factory documentation (white papers, cases studies, something...) on the finer details of AC coupling SMA gear to SMA gear.

After re-reading and really digging in the SI manual more of it appears to be there than I originally captured. Whether or not its easy to understand or logically laid out is my own personal "SLD ax to bear".

FWIW, now that I've got this thread started I'm seeing "similar threads" on the bottom of the page that the search function wasn't revealing before and they are helping to fill in blanks so that's nice. Kudos to the robot the reads my messages.

The only definition of "SLD" I can find is "Specific Learning Disorder". Did I guess correctly?

Because I'm using 10 year old Sunny Island and Sunny Boy, the 10 year old documents are fine for me. But I do wish they put out better descriptions & diagrams of how things connect and operate. There are new models and new monitoring devices.

Sorry about that. SLD refers to a "single line diagram", also called "one line diagram".

Once I have specific questions I'll report back.

 

[OzSolar]

If you use both grid and generator, need a transfer switch (or generator interlock) of course. Each source must be off or disconnected for 5 seconds before connecting the other, so if Sunny Boy or Sunny Island is backfeeding the grid, it doesn't suddenly find itself driving out-of-phase into the new source. You also need a signaling contact to tell Sunny island which, so it avoids backfeeding generator and relaxes specs.

Here's the line diagram from the SMA data sheet showing the DIGIN.



With that I built the below diagram for my site. Any critique will be appreciated. It's worth mentioning that my main goal for the additional 13kW of PV is to save money, not to beef up my back up power systems. Certainly don't want to ignore the opportunity though!!

I already have the SB6048's backing up the most critical loads on top of the propane generator backing up the entire site. Since we rarely lose power for more than a few hours for a few times per year it easier to ignore AC coupling but oh darn should that big ice storm, etc knock us out for weeks I'd really be kicking myself.

There's considerably more trenching and wire involved in a AC coupled vs a straight grid tie system due to the distances between the house, the solar array and the utility entrance.

What things am I missing/should be considering? Is their an appreciable difference between the total energy produced in a AC Coupled system vs straight tied tie system? I would think they'd be pretty close.

Thank you

View attachment 88329

 

[Hedges]

The link didn't open for me: https://diysolarforum.com/attachments/88329/

AC coupled on output of Sunny Island and AC coupled on grid side should be essentially the same. Only difference how much resistance and voltage drop in the way. Not much unless really long wires. What you do need is low enough resistance that voltage rise on top of what voltage utility gives you doesn't exceed max voltage, cause inverter to disconnect.

Lots of excess PV is good to have for backup situation, because it makes up for less sunshine. Also keeps battery full later in the day. PV is cheaper than batteries. Try to turn off loads like water heater at night, keep draw absolute minimum.

 

[OzSolar]

The link didn't open for me: https://diysolarforum.com/attachments/88329/

AC coupled on output of Sunny Island and AC coupled on grid side should be essentially the same. Only difference how much resistance and voltage drop in the way. Not much unless really long wires. What you do need is low enough resistance that voltage rise on top of what voltage utility gives you doesn't exceed max voltage, cause inverter to disconnect.

Lots of excess PV is good to have for backup situation, because it makes up for less sunshine. Also keeps battery full later in the day. PV is cheaper than batteries. Try to turn off loads like water heater at night, keep draw absolute minimum.

Well that's odd about that attachment. Sorry about that. Let's try it again.

 

[Hedges]

Is "Required Utility Disconnect" required to be between GT PV inverter and critical loads panel? or between Sunny Island and main panel?
Your Sunny Island may be able to backfeed utility from DC coupled PV, and I'd expect a disconnect to prevent any backfeed. If required; it was for me, but later became optional because utility could yank meter to isolate.

The trouble with your auxiliary relay is it appears to tell Sunny Island grid is connected any time grid has voltage. But ATS might still be switched to generator, so could lead to backfeeding of generator.

Alternatives could be auxiliary contacts actually switched by ATS switch. or, coil of NO relay fed by generator so any time generator is running you tell SI you're on generator. That should be safe in terms of not enabling backfeed, even if actually on grid. But it is not safe in terms of fail-safe; if wire breaks then contacts stay open for "on grid" even though off. I'd rather mechanical linkage to ATS mechanism.

That diagram is much like SMA SI manual, except mine doesn't say "120V to coil"

That "critical loads" is truly critical, will remain powered even if battery > 70% DoD, until it shuts down at 80% DoD.
You should add a load-shed relay and an "important loads" panel which sill get disconnected at 70% DoD.

I suggest also an interlocked "generator" breaker allowing main panel to be fed from important loads panel. With some manual flipping of breakers, this lets you power loads on main panel from inverter if desired.

 

[OzSolar]

Your Sunny Island may be able to backfeed utility from DC coupled PV

Yes it would! I had caught that at one point then failed circle back to address it. Good catch. I REALLY appreciate this!

The trouble with your auxiliary relay is it appears to tell Sunny Island grid is connected any time grid has voltage. But ATS might still be switched to generator, so could lead to backfeeding of generator.

Alternatives could be auxiliary contacts actually switched by ATS switch. or, coil of NO relay fed by generator so any time generator is running you tell SI you're on generator. That should be safe in terms of not enabling backfeed, even if actually on grid. But it is not safe in terms of fail-safe; if wire breaks then contacts stay open for "on grid" even though off. I'd rather mechanical linkage to ATS mechanism.

Great points. We're on the same page. I'm in the "big picture" phase and hadn't gotten down the the finer details yet. My rough plan was to use the auxiliary contacts on the xfer switch or install a new NO relay that was closed when the generator was running. Hmnnn... either way that will result in the solar shutting down during the weekly exercise cycle, won't it? I suppose that's easily manageable by changing the timer on the ATS to run just before sunrise?

Good point about a wire breaking on a NO control circuit. It's going to stay open regardless if the generator is running.

That diagram is much like SMA SI manual, except mine doesn't say "120V to coil"

Here is where I found that diagram. Sunny Island 4548-US, 5048-US, 6048-US :: Grid Backup with Generator :: Digital Input "Supplemental information to SI4548-US, SI5048-US and SI6048-US Technical Description section 14.3 Generator and Grid. North American/US models only."

That "critical loads" is truly critical, will remain powered even if battery > 70% DoD, until it shuts down at 80% DoD.
You should add a load-shed relay and an "important loads" panel which sill get disconnected at 70% DoD.

I suggest also an interlocked "generator" breaker allowing main panel to be fed from important loads panel. With some manual flipping of breakers, this lets you power loads on main panel from inverter if desired.

I'm trying to figure out how far down the rabbit hole I want to go with AC Coupling for back up power. The costs and efforts of manual transfer switches/interlocks and all the wire and trenching are going to quickly add up compared to the ~small~ improvements the overall system uptime. It's the old 80-20 rule. 80%+ the best possible results will be captured in the first 20% of effort (and cost). Of course I'm not suggesting doing anything unsafe.

So here's a V2 diagram that feels a like a good balance of cost to benefit ratio and ticks all of the safety boxes and most of the "convenience" boxes.

I will look forward to input.

 

[Hedges]

From the SMA note,

"However, if an auxilary relay is not available, the installer will need to purchase an external relay/contactor with a 120V coil or 240V coil. The voltage for the coil must be provided from the grid side of the transfer switch. In the event of a grid outage, the voltage will no longer be applied to the coil, and the contact will close."

I would be concerned brownout in the grid might have sufficient voltage to close the external relay (or keep it closed) even though ATS sensed grid failure and switched to generator. Then PV backfeeds into generator.

Auxiliary contacts on transfer switch would signal SI based on switch position, not presence of voltage from grid (or generator). That should allow generator exercise without disconnecting from grid (so long as ATS priority is grid when both sources present.)

In your diagram, I would have had SI and battery at the house. Probably SB at house to, with PV wires to solar array at edge of field.
(batteries of course far enough from house to burn harmlessly, if lithium.)

I assume you have a breaker at meter. Yes, if disconnect switch ties in there, no backfeeding of generator can occur. You can't send generator power to SI.


Different placement: How about SI (and maybe SB) by generator and utility meter? Send power through wire to house to critical loads panel, and load-shed relay to main panel.

An interlock breaker in critical loads panel could let it be fed from main panel as you say.
An interlock breaker in main load panel could let it be fed from critical loads panel (good to have load-shed relay there.)

I would like to be able to take a generator (main panel in your case) and feed it to PV input of a Sunny Boy. That would let generator backfeed AC1 output of SI, just like any PV input. Galvanic isolation required (transformer) and maybe inrush limit (precharge resistor.) Some inverters have a "turbine" or constant-voltage mode.

 

[OzSolar]

@Hedges It will take me a bit to "chew" on all of that. There's not much real estate near the house or the meter which is a combo meter with breakers.

The current set up is one of those DC Solar trailers that have been on the market thanks to whatever shenanigans happened there. It's currently intact with it's flooded lead acid batteries.

 

[Hedges]

You'll have to see what shape the batteries are in. Some people found dry cells and either reduced capacity or different SoC.
The 2400W of panels included probably aren't enough to equalize (might have a better chance one battery at a time.)
Hopefully they can be brought to a good state and get another decade's use.

The DC trailer enclosure takes care of weatherproofing SI. But spacing around them didn't follow installation manual, and DC added a small fan for the enclosure. You might find a way to operate with doors open and extra roofing to keep weather out. Of course, heat generation would be from serious power inverted or charging. In a SI/SB system, most power is PV to loads or to grid.

My pallet of new-in-the-box SI was presumably from those same shenanigans. $1,000,000,000!

 

[Hedges]

Sunny Island has air inlets at the bottom. Cutting holes in bottom of enclosure and ducting to bottom of Sunny Island would ensure fresh air, rather than recirculating hot air. Which would happen in the box even when SI's fans turn on.
SMA designed their "OptiCool" system to work mostly by convection, with variable speed fans run when necessary. Clearance was to allow air escape by convection.

 

[OzSolar]

You'll have to see what shape the batteries are in. Some people found dry cells and either reduced capacity or different SoC.
The 2400W of panels included probably aren't enough to equalize (might have a better chance one battery at a time.)
Hopefully they can be brought to a good state and get another decade's use.

I found the cells topped off and the batteries in ~good~ shape. I've equalized them several times and did a few rudimentary capacity tests which takes a bit when you've got ~48kWH. I got ~36kWh each time which seemed pretty reasonable. ~.95 inverter efficiency x ~.85 battery efficiency = ~39kWh's. All the cell SG's are very close to each other as well.

The DC trailer enclosure takes care of weatherproofing SI. But spacing around them didn't follow installation manual, and DC added a small fan for the enclosure. You might find a way to operate with doors open and extra roofing to keep weather out. Of course, heat generation would be from serious power inverted or charging. In a SI/SB system, most power is PV to loads or to grid.

Hmnn.... more food for thought.

 

[OzSolar]

Of course, heat generation would be from serious power inverted or charging. In a SI/SB system, most power is PV to loads or to grid.

Please make sure I understanding this correctly. You're saying when using the SI/SB combination in a grid tie set up they will generate less heat than vs being in off grid mode? That's what I was thinking and it seems logical to me.

On that subject. Any idea much less efficient is a SI/SB combo at producing usable AC KWH's? IE: SB by itself would generate 15,000 usable AC KWHs/year. SB/SI combo would generate X % less?

I'm assuming the SI's don't sleep at night so those tare losses plus keeping the batteries float charged would add up.

I can't get any answers from SMA. The phone support people actually told me "that's above the scope of thier knowledge that and I'd have to send an email to the engineering" plus warned me that they would likely take weeks to respond if ever. Yikes.

 

[Hedges]

What I mean is the Sunny Islands provide starting surge, then settle back to just charging batteries. During the day, PV into Sunny Boy powers the big loads. So Sunny Island (which has restricted and recirculating airflow in the DC solar box) won't be generating as much heat as it would in a DC coupled system.

SB will give peak efficiency generating A/C.
When SI is running, it consumes 25W. If in sleep mode, 6W. I don't think there is a way to have it sleep and pass through grid to SB.

It can sleep off-grid when no-load, or with several, one or more can run as needed depending on power demand. But I have heard from another forum member that in a split-phase setup, when they sleep you lose the second phase, so I plan to experiment with an auto-transformer.

Float I think is at a voltage where battery doesn't draw current. My 4x SI report 0.1kW into battery, which doesn't have enough digits to be sure but sounds like 4 x 25W = 100W.

Welcome to support! At least it is US based! And manned by homo sapiens, not by AI!

I know what I know from RTFM, plus engineering understanding and experience with them.

 

[OzSolar]

SB will give peak efficiency generating A/C.
When SI is running, it consumes 25W. If in sleep mode, 6W. I don't think there is a way to have it sleep and pass through grid to SB.

Again you've given me more to think about than I was asking for. Ha It's seems to be getting clearer and clear that for me using the SI/SB is quickly getting to be more hassle and expense than it's worth. My grid is super reliable (today at least) and there's already a big generator backing up the entire farm. For a full time off grid property, particularly those that could use the SI as the back bone of a central power system with SB's scattered around at various buildings (cottages, work shops, etc) is really where the SI/SB system would shine.

Thanks again for letting me bounce my rambling ideas off of you!

 

[Hedges]

GT PV is quite cost effective. If I buy all hardware and provide my labor for free, my estimate is $0.025/kWh amortized over 20 years.

Battery backup is not cost effective on a per-kWh basis. It is cheaper than paying utility to extend power lines if not already at the grid.
Just they batteries themselves will cost $0.50/kWh for power cycled through them.

That isn't even considering the cost and complexity of diagnosis and repair. If GT PV fails, utility power to your house still works. If battery/inverter system fails, downstream loads shut down. I put in manual transfer switch (interlocked breakers) to bypass and go straight to grid, but no assurance anyone else would know what to do if I was no longer around.

Battery backup is worthwhile if you're willing to pay a premium for power to keep it working during grid failures. Not all power has to go through it (at $0.50/kWh wear-out cost for battery), just night-time consumption. My PV is a few times larger than what battery could store, and most power is use-it-or-lose-it during the day, runs A/C and everything else. (I'd rather throw away unneeded $0.025/kWh power than pay $0.50/kWh to store it.)

I put off installing Sunny Island for 17 years because it didn't make financial sense for me. If you have a large expensive system including diesel generator, you might want a small Sunny Island with some PV on it and maybe a transfer switch to generator. That would power some loads 24/7 with less generator usage. (Transfer switch feeding Sunny Island grid vs. generator must be "slow", 5 second minimum delay on switchover. E.g. manual interlocked breakers.)