AC Coupled mini-grid (micro-grid)

I'm sorry that I'm slow on the uptick here. I have been reading up on AC coupling to create a "mini grid" or "micro grid" depending on the literature. I see that SMA has written dozens of white papers pushing this technology, but one area that I can't figure out is how they disconnect the "mini grid" from the grid. Have I missed something? Xantrex has an application note demonstrating their technology using their latest firmware in the XW hybrid charger/inverters (also called "bi-directional inverters" in other literature) "AC Coupling of Inverters: Forming Mini-grids with XantrexTM XW Hybrid Inverter/Chargers and GT Inverters" document 976-0240-01-01 Rev A.

In the Xantrex architecture, the XW hybrid controls the system. The method used to control the output of the GT inverter is called “droop-mode-control” which is how rotating generators synchronize to each other. This involves leading or lagging in phase to control power transfer. As the GT inverter will also use phase control to push power on to the AC bus, the XW must integrate the phase difference to a frequency delta. In the literature, the figure of +/- 3% frequency is used as the max frequency delta, which at 60 Hz is +/- 1.8Hz. In this discussion about Sunny Island Inverters, stephendv mentions +/- 2Hz. However, in the Xantrex whitepaper, they drag the line frequency down by 5Hz, which is about 8%. Which is a lot. Is this OK? Does it work OK? Once the GT inverter shuts off, the XW seems to jump line frequency back to 60 Hz - it looks like quite a discontinuity in frequency. From that point on, how is the GT inverter restarted? Or does it keep trying to reconnect every 5 minutes and the XW hybrid inverter slews frequency to shut it down again and again?

The Xantrex suggested solution to the grid disconnect problem is to use an external AC transfer switch on the GT inverter to transfer from the grid to the XW output bus. However, there seems like there may be some troublesome transient behavior when the grid goes away. As I understand it, when the grid drops, the GT inverter will disconnect for 5 minutes, as required by UL. So the XW hybrid inverter must shoulder the entire load until the GT checks for line voltage again 5 minutes later. This seems like a very long time - can it be shortened - for example, can newer XW firmware control the GT via the Xanbus? Can this operational state be externally determined from the XW's Modbus interface so that it could be managed by central controller? Secondly, Xantrex disavows this configuration with a generator connected to the AC2 input of the XW hybrid. I can't understand how this would be any different from having several XW hybrids in parallel while running a back-up inverter generator. Have I missed something?

Finally, it occurred to me that I don't understand how the XW system handles restoration of grid power and synchronization when the GT is online or the back up generator is online. I can only assume that the XW will first charge the batteries from the grid if necessary, then switch off the generator, slew inverter phase to match the grid, and then switch to grid power. Does that sound right?

SMA/Sunny seems to be the leader in this area, and they seem to be the tightest integrated solution, with the exception of the grid disconnect. However, their hardware seems really cost prohibitive. "Julie in Texas" seems to accomplish control of a mini grid using software with Outback gear. I hope she can offer her experience.

Sorry about all the questions, I am learning that power system design is not as straight forward as I thought. I'd appreciate any thoughts or experience that anyone might have.

Andrew
«1

Comments

  • solarix
    solarix Solar Expert Posts: 713 ✭✭
    Re: AC Coupled mini-grid (micro-grid)

    Does not the island inverter drive a dedicated load (or micro grid) seperate from the utility grid? A different connection hooks the island inverter to the utility grid in order to charge or export. The GT inverter's power is available to the island loads directly, then used to charge the batteries, then any excess is exported by the island inverter to the utility grid. I've not done the SMA system but this is how the XW is.
  • Dr. Strangelove
    Dr. Strangelove Solar Expert Posts: 49
    Re: AC Coupled mini-grid (micro-grid)
    solarix wrote: »
    Does not the island inverter drive a dedicated load (or micro grid) seperate from the utility grid? A different connection hooks the island inverter to the utility grid in order to charge or export. The GT inverter's power is available to the island loads directly, then used to charge the batteries, then any excess is exported by the island inverter to the utility grid. I've not done the SMA system but this is how the XW is.

    That's the AC coupling, at the grid interface node. The XW inverter has an AC transfer switch that disconnects from the grid AC and to the AC2 (genset) input. The GT can only disconnect from the grid. If they are both connected at the grid node, when grid power fails, they are become disconnected from each other (or, worse, I suppose they could fool each other into believing the grid is still up). That's why Xantrex recommends putting the output of the GT inverter on a transfer switch between the grid and the "island" power, controlled by the AUX signal from the XW inverter.

    One SMA paper that describes their AC coupling design does show an external AC isolation switch to the grid is: "New V/f–Statics controlled Battery Inverter: Sunny Island - the key component for AC-Coupled Hybrid Systems and Mini Grids". As far as I can tell, it is because the Sunny Island is an off grid inverter, so it doesn't have an internal transfer switch. But the control of the external switch is not described.

    Or am I misunderstanding things?

    Andrew
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    One SMA paper that describes their AC coupling design does show an external AC isolation switch to the grid is: "New V/f–Statics controlled Battery Inverter: Sunny Island - the key component for AC-Coupled Hybrid Systems and Mini Grids". As far as I can tell, it is because the Sunny Island is an off grid inverter, so it doesn't have an internal transfer switch. But the control of the external switch is not described.

    That's pretty much it.
    The SI was designed as an off-grid inverter and since it's not permitted to grid feed from battery based systems in Germany and Spain (at least), so the SI on it's own doesn't support this. SMA have developed the Sunny Backup system, which is basically a sunny island with different firmware and an external automatic transfer switch, which allows you to switch the connection from the SB to the grid or to the mini-grid.
    From what I understand, they sell this as 1 integrated product - so you can't buy the transfer switch on it's own, although if your regs allow, I can't see why you couldn't use a different one.
    Victron also do an external transfer switch which does the same thing:
    http://www.victronenergy.com/solar/solar-switch/
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    That's the AC coupling, at the grid interface node. The XW inverter has an AC transfer switch that disconnects from the grid AC and to the AC2 (genset) input. The GT can only disconnect from the grid. If they are both connected at the grid node, when grid power fails, they are become disconnected from each other (or, worse, I suppose they could fool each other into believing the grid is still up). That's why Xantrex recommends putting the output of the GT inverter on a transfer switch between the grid and the "island" power, controlled by the AUX signal from the XW inverter.

    One SMA paper that describes their AC coupling design does show an external AC isolation switch to the grid is: "New V/f–Statics controlled Battery Inverter: Sunny Island - the key component for AC-Coupled Hybrid Systems and Mini Grids". As far as I can tell, it is because the Sunny Island is an off grid inverter, so it doesn't have an internal transfer switch. But the control of the external switch is not described.

    Or am I misunderstanding things?

    Andrew
    The Sunny Island does have a transfer switch. When the grid is up, the SI is in bidirectional passthrough mode and the inverter section is asleep. The microgrid is between the SI and and the solar generator, if there is one, with the protected loads panel in the middle. When the grid goes down, the SI disconnects itself from the grid and powers up its inverter from batteries and feeds the microgrid. The solar generator keeps producing because it does not know that the grid is down.

    There are a couple of drawbacks. The SI is only a 120VAC unit; to power a 240VAC microgrid you either have to have 2 (or 4, 6...) of them or put an autotransformer on its output.

    Also, the only way that the protected loads get power is through the SI (either from passing grid power through it or from its inverter from batteries), and if you open the switch to the batteries on the SI, it doesn't do either. Some sort of bypass switch is really needed so that you can route grid power around the SI to your protected loads so they don't go down whenever you turn off the SI.

    And to get the SI to play well with a solar generator, you need to use Sunny Boy inverters. The SI and SB have a communication line that throttles down the output of the SB when the grid is down, the loads are less than the output of the PV inverter, and the batteries are fully charged.

    The upside is that the SI is a very smart machine with tons of programming flexibility to cover a multitude of circumstances.
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    However, in the Xantrex whitepaper, they drag the line frequency down by 5Hz, which is about 8%. Which is a lot. Is this OK? Does it work OK?

    I know of at least one user of an AC connected sunny island that had issues with a fan on a pellet boiler that kept blowing because of the frequency changes. Don't know whether you can really blame the SI for that, but nevertheless frequency sensitive devices do exist.
    Once the GT inverter shuts off, the XW seems to jump line frequency back to 60 Hz - it looks like quite a discontinuity in frequency. From that point on, how is the GT inverter restarted? Or does it keep trying to reconnect every 5 minutes and the XW hybrid inverter slews frequency to shut it down again and again?

    Ah. I'm not familiar with the Xantrex system, but what you're describing sounds like it's an On or Off situation, not variable control like the SI. There's another interesting paper (non-SMA) about using this on-off behaviour which the authors say is bad for battery life because of "micro cycling". I don't pretend to understand it, but you might :D

    http://currentgeneration.co.nz/site/current/files/Partial%20AC-coupling%20in%20Minigrids.pdf
    SMA/Sunny seems to be the leader in this area, and they seem to be the tightest integrated solution, with the exception of the grid disconnect. However, their hardware seems really cost prohibitive. "Julie in Texas" seems to accomplish control of a mini grid using software with Outback gear. I hope she can offer her experience.

    I think for the US market the Xantrex offers split phase, is cheaper and now also offers some form of AC coupling is hard to beat.

    Regarding the issues you raise about the frequency control of the Xantrex, yes the SI does this smoother, but then if your system is primarily a grid connect system, the battery based usage should be very minimal, so the niggles with the Xantrex are really only a concern for a tiny portion of the usage time. The rest of the time it would behave as a normal GT system.
    If on the other hand, the system is primarily off-grid, then I wouldn't go with AC coupling at all.
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    ggunn wrote: »
    And to get the SI to play well with a solar generator, you need to use Sunny Boy inverters. The SI and SB have a communication line that throttles down the output of the SB when the grid is down, the loads are less than the output of the PV inverter, and the batteries are fully charged.

    There's no external comms line, the SI and SB just communicate through grid frequency which the SI controls. 60Hz = send 100% of your power, 62Hz = send 0% power, and everything in between in a linear scale.
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    stephendv wrote: »
    There's no external comms line, the SI and SB just communicate through grid frequency which the SI controls. 60Hz = send 100% of your power, 62Hz = send 0% power, and everything in between in a linear scale.
    That's not correct. There is a comm line that runs between the SI and SB. You install daughter boards on both units and run CAT5 between them. You are correct about the frequency munging to control the output of the SB, but the comm line tells the SB that it is the SI that is doing it. I'm not sure what other information is exchanged over the comm line, but it is definitely there in SI-SB installations.
  • Solar Guppy
    Solar Guppy Solar Expert Posts: 1,989 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)

    From a practical stand point the XW/GT and SB/SI systems are identical.

    To run the XW like the SB, the GT inverters and connected on the Output of the XW, just like how the SB/SI work.

    The only difference is the XW can also sell back to the grid and auto drop off with the built in disconnect.

    You can use any UL1741 GT inverter in the XW "mini island" and yes the XW, like the SI will drift the frequency to knock the GT inverters offline and they will auto retry every 5 minutes per the UL1741 spec

    Both are proven configurations and work as described, the XW has more features and a lower price point
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    ggunn wrote: »
    That's not correct. There is a comm line that runs between the SI and SB. You install daughter boards on both units and run CAT5 between them. You are correct about the frequency munging to control the output of the SB, but the comm line tells the SB that it is the SI that is doing it.

    Hehe, nope. The comms line is purely for monitoring the systems (and it's optional), there's no communication between the SI and SB over it. You have to manually program the SB to be in "off-grid" mode, and once that's done then it knows that it should vary it's output based on the frequency.

    You can have a look at the SI installation manual which explains the process.
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    From a practical stand point the XW/GT and SB/SI systems are identical.

    [snip] and yes the XW, like the SI will drift the frequency to knock the GT inverters offline and they will auto retry every 5 minutes per the UL1741 spec

    Just to be clear, the SI doesn't knock the SB's offline, it varies their output proportional to the loads. E.g. if there's 3kW available from the solar array, and the battery is fully charged, but there's only a 1kW load, then the SI will dynamically adjust the frequency so that the SB outputs exactly 1kW. And if that 1kW load suddenly disappears, then the SI adjusts the frequency again so that the SB stops it's output, but it's by no means offline. If a new load should be switched on, the frequency adjusts again within seconds and the SB's respond immediately with the required power.
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    in the Xantrex whitepaper, they drag the line frequency down by 5Hz, which is about 8%. Which is a lot. Is this OK? Does it work OK?

    The GT inverter will likely disconnect long before the XW ramps the freq down to 5 Hz. The XW must have a wide range just in case the random model of GT inverter is more liberal on grid freq spec (-0.7 Hz to +0.5 Hz). XW will stop its downward freq slew when it detects the GT inverter has shut down.
    Once the GT inverter shuts off, the XW seems to jump line frequency back to 60 Hz - it looks like quite a discontinuity in frequency.

    It looks that way from the drawing but I doubt it is abrupt. It would stress the MOSFET's and transformer to do an instantaneous freq step. It probably takes a few seconds to slew back to 60 Hz.
    From that point on, how is the GT inverter restarted? Or does it keep trying to reconnect every 5 minutes and the XW hybrid inverter slews frequency to shut it down again and again?

    When the XW forces the GT to shut down, at that point battery power is being used to power ACout loads. In the 5 mins it takes for the GT to reconnect there will be some discharging of batteries. After the GT reconnect, charging will occur and when the batteries again exceed absorb voltage the XW will again force the GT to disconnect. (If ACout loads cannot consume the PV power).

    This is a disadvantage of the XW over the Sunny Boy system which appearantly can modulate output control on its GT inverter by communicating via a linear change in line frequency. They probably have a patent on the freq slew control function.

    Another important point on the XW system is there is no battery charging current regulation when GT inverter backfeed power comes in. If the PV system is producing 5kW it has to go somewhere or be shut down. If your loads are taking 1kW then 4 kW is pushed into batteries regardless of their AH size. I think Xantrex should state that if your battery bank is not large enough to take a bulk charge rate of full PV wattage then don't do AC coupling of GT inverter.
    The Xantrex suggested solution to the grid disconnect problem is to use an external AC transfer switch on the GT inverter to transfer from the grid to the XW output bus. However, there seems like there may be some troublesome transient behavior when the grid goes away. As I understand it, when the grid drops, the GT inverter will disconnect for 5 minutes, as required by UL. So the XW hybrid inverter must shoulder the entire load until the GT checks for line voltage again 5 minutes later. This seems like a very long time - can it be shortened - for example, can newer XW firmware control the GT via the Xanbus? Can this operational state be externally determined from the XW's Modbus interface so that it could be managed by central controller?

    Depending the grid pertubation, there will usually be a dip in AC voltage for a cycle to two until the XW opens the power relay to the grid. The GT inverter on ACout may react to this one or two cycle glitch and shut down or it may stay connected. If it stays connected then XW becomes the frequency master. If GT disconnect then 5 mins later it will reconnect. The time to reconnect is set by the GT inverter. XW has no control over this.

    The transfer switch is recommended for a grid feed that has a second backfeed utility meter, separate from incoming utility meter. This is only required where the utility pays you based on its separate backfeed meter flow. The diagram in the app note is not very clear on the metering arrangement.

    The second use of a transfer switch is to cut off a generator feed to prevent gen from getting a power backflow push.
    Secondly, Xantrex disavows this configuration with a generator connected to the AC2 input of the XW hybrid. I can't understand how this would be any different from having several XW hybrids in parallel while running a back-up inverter generator. Have I missed something?

    The final quirk to the XW system is if a generator is used. A normal generator frequency spec may not be stable enough for GT inverter to stay lock on to. If this happens the following is a don't care.

    The XW does not restrict a PV excess power push to AC2in so if loads or charging cannot consume PV power it will push the excess at the generator. They recommend putting a transfer switch to disconnect generator line when GT inverter is active (or it could be other way around). The problem is while on the generator the gen sets the frequency and the XW cannot slew the freq to cause the GT inverter to disconnect. Again, if PV is producing power it has to go somewhere or be shut down.
    Finally, it occurred to me that I don't understand how the XW system handles restoration of grid power and synchronization when the GT is online or the back up generator is online. I can only assume that the XW will first charge the batteries from the grid if necessary, then switch off the generator, slew inverter phase to match the grid, and then switch to grid power. Does that sound right?

    Remember the schematic. ACin -> power relay -> bidirectional inverter hard wired in parallel with ACout.

    When grid shows up, the XW must first syncronize to it then close power relay. If the GT inverter on ACout does not like the freq/phase slew rate change as the XW syncronized to grid it will disconnect and reconnect 5 minutes later.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)

    This thread is great reading and very informative, and even though I don't do anything with this sort of configuration I've found it interesting.

    But I have got to point out something for the sake of clarity which applies to all posts under ever thread:

    Dr. Strangelove said: "they drag the line frequency down by 5Hz"

    RCinFLA said: "ramps the freq down to 5 Hz"

    There is a big difference between these. The first is a shift in frequency from 60 Hz to 55 Hz. The second is a shift in frequency from 60 Hz to 5 Hz. I don't know which is the correct terminology for the operation of the inverter, but I'd suspect the former.

    I know: I'm a nit-picking old fool. But the point is it can be very important to get terminology right, as well as mathematical calculations. Often the one depends on the other. It could be the difference between "Yeah! It works!" and "So, have they discovered what the cause of the fire was?"

    Okay; end of admittedly unwarranted English lesson. You can all razz me now. ;)
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    This thread is great reading and very informative, and even though I don't do anything with this sort of configuration I've found it interesting.

    Dr. Strangelove said: "they drag the line frequency down by 5Hz"

    RCinFLA said: "ramps the freq down to 5 Hz"

    I did not say it ramps freq down to 5 Hz. I said it ramps frequency down until GT inverter disconnects. The disconnect will usually happen in less then 2 Hz offset if the GT inverter is operating per UL1741 specs. Per the app note it could go as far as 5 Hz down in attempt to shed the GT inverter.
  • Dr. Strangelove
    Dr. Strangelove Solar Expert Posts: 49
    Re: AC Coupled mini-grid (micro-grid)
    RCinFLA wrote: »
    ... a lot of useful information ...

    Thank you very much - this was instructive. I understand a lot more now.

    I don't think that the Xantrex GT inverters know how to limit power. If they did, I would expect that there would be a Modbus register to set the power limit or at least read it back. SMA/Sunny identifies registers 30835 as "Active Power Limitation Mode" and 30839 "Active Power Setpoint %" but they are "Read Only" (SUNNY WEBBOX MODBUS Interface Technical Description). The SMA/Sunny system power output seems to be only managed via the "droop mode" (AC coupling) line control.

    Have I missed anything? I'd like to learn as much as I can about finer grained control of these systems, as well as how the other manufacturers like Outback and Midnight handle an AC coupled, grid ties, net metering system. Thank you.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)

    No offense, but you did:
    RCinFLA wrote: »
    The GT inverter will likely disconnect long before the XW ramps the freq down to 5 Hz.

    This is not a slur on you (RCinFLA is one of the smartest people on this forum), but rather to point out that when terminology gets mixed the meaning gets confused.
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    Thank you very much - this was instructive. I understand a lot more now.

    I don't think that the Xantrex GT inverters know how to limit power. If they did, I would expect that there would be a Modbus register to set the power limit or at least read it back. SMA/Sunny identifies registers 30835 as "Active Power Limitation Mode" and 30839 "Active Power Setpoint %" but they are "Read Only" (SUNNY WEBBOX MODBUS Interface Technical Description). The SMA/Sunny system power output seems to be only managed via the "droop mode" (AC coupling) line control.

    Have I missed anything? I'd like to learn as much as I can about finer grained control of these systems, as well as how the other manufacturers like Outback and Midnight handle an AC coupled, grid ties, net metering system. Thank you.

    You are correct that the XW inverter has no control over a GT inverter's output power other then causing it to turn off by slewing frequency outside of regular grid spec limits.

    My personal opinion is I think the Xantrex method is fine with the caveat that your maximum PV generated power must not cause your batteries to exceed their maximum charge current based on your batteries AH size. (PV power also cannot exceed rated power of the XW inverter)

    I am sure Xantrex has no interest in doing anything to promote the sale of Sunny Boy or any other manf's equipment. Xantrex would rather sell you their new high input voltage charge controller instead of you buying a non-Xantrex GT inverter.

    If the Sunny Boy freq control of GT inverter output power is patented then Xantrex would have to pay them a royalty for including its control mechanism in XW inverter.
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    RCinFLA wrote: »
    If the Sunny Boy freq control of GT inverter output power is patented then Xantrex would have to pay them a royalty for including its control mechanism in XW inverter.

    I don't believe it is. Currently the Kaco inverters also support something similar, although it's not a linear derating, but instead a 2 step process: at 50Hz it produces 100% and at 50.3Hz it produces 50%, above 51.5Hz it disconnects.

    The 2009 German code for renewables has made linear derating based on frequency a requirement for inverters in the >100kW range, so it may not be too long before we see the same thing in other residential scale GT inverters.

    There's quite a cute feature of the Sunny Island which gives an indication of how much thought went into the system:
    Some walls clocks using the grid frequency for their timing control, so if the grid frequency is fast, the clock would run fast. So the Sunny Island keeps track of how much time it spends in over-frequency mode and then when the frequency adjustment is not necessary anymore, it reduces frequency below line voltage to make up for it.
    E.g.
    If it spent an hour at 62Hz, then it would also spend another hour at 58Hz so that over time your clocks stay accurate!
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    stephendv wrote: »
    Hehe, nope. The comms line is purely for monitoring the systems (and it's optional), there's no communication between the SI and SB over it. You have to manually program the SB to be in "off-grid" mode, and once that's done then it knows that it should vary it's output based on the frequency.

    You can have a look at the SI installation manual which explains the process.
    Hehe yourself ;). The SI installation manual shows the comm line irrespective of whether or not there is a WebBox in the system. I have installed a couple of these systems and spoken to SMA tech support about them, and they say this connection is necessary with no mention of it being only for monitoring. Maybe they just want to sell me their daughter boards?

    I confess to not knowing exactly what is passed over the line, but I have a SI installation manual at home. Can you point me to the section of the manual that explains exactly what the comm line does and why it is not necessary in a non-monitored system? I would appreciate it.
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    From a practical stand point the XW/GT and SB/SI systems are identical.

    To run the XW like the SB, the GT inverters and connected on the Output of the XW, just like how the SB/SI work.

    The only difference is the XW can also sell back to the grid and auto drop off with the built in disconnect.

    Maybe I am misunderstanding you, but the SI/SB system sells back to the grid. The SB feeds the grid through the passthrough of the SI when the grid is up and the loads on the microgrid are less than the output of the SB. The SI has a built in disconnect that separates the microgrid from the main grid when the main grid goes down, too; is that what you mean by "auto drop off"?
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    ggunn wrote: »
    Hehe yourself ;). The SI installation manual shows the comm line irrespective of whether or not there is a WebBox in the system. I have installed a couple of these systems and spoken to SMA tech support about them, and they say this connection is necessary with no mention of it being only for monitoring. Maybe they just want to sell me their daughter boards?

    I confess to not knowing exactly what is passed over the line, but I have a SI installation manual at home. Can you point me to the section of the manual that explains exactly what the comm line does and why it is not necessary in a non-monitored system? I would appreciate it.

    I'll see your "hehe" and raise you a "shazam" :D
    http://files.sma.de/dl/5612/SI5048-TB-TEN110340.pdf
    It's going to be a tall order to find where they explicitly list the things that aren't necessary, but you can infer from the paragraph about the comms module that it's not a required component:

    Chapter 6.5:
    "You can connect SMA Solar Technology communication devices (e.g., Sunny Boy Control, Sunny WebBox) or a PC with the appropriate software to a communication interface. A detailed cabling diagram can be found in the communication device manual, the software or on the Internet at www.SMA.de/en.
    You can incorporate an RS485 communication interface into the Sunny Island."


    Nowhere in that chapter (or anywhere in the manual) do they describe connecting a sunnyboy to a sunny island via the comms modules.

    Chapter 17 Describes how to setup a sunnyboy for off-grid use, including:
    "Depending on the device type, you can set the PV inverter to "stand-alone grid operation" using Sunny Boy Control, Sunny WebBox, Sunny Data Control or Sunny Explorer."

    If I had to guess, I think this is why the SMA chaps are saying you need the comms module in the sunny boy. But once you've programmed it you can remove the module and keep it for the next install.
  • Dr. Strangelove
    Dr. Strangelove Solar Expert Posts: 49
    Re: AC Coupled mini-grid (micro-grid)

    Can someone help clarity a few more things for me:

    The XW hybrid inverter has AC connections (see page 2-19 here)
    • AC1 connection to the external grid
    • AC2 connection to a back-up genset
    • ACOut connection to the local mini-grid (sources & loads)

    The Sunny Island has AC connections (see page 15 here):
    • AC1 connection to external grid AND back-up genset
    • AC2 connection to internal mini-grid

    So the XW system appears to incorporate the transfer relay between the grid and the generator into its cabinet. However, an external transfer relay is necessary for the Sunny Island to switch between the external grid and the backup generator. So how does the Sunny Island know when to switch back to the grid once it is disconnected?

    Second, the Sunny Island seems to be a single phase; to make a 240V/60Hz split phase power system, you need two Sunny Islands. Is this correct?

    Third, the XW system requires a transfer switch between larger GT inverters and the external and internal AC grids. I don't understand why this is necessary: if the XW can control charger power from the external AC grid, why can't it control it when drawing from the internal grid? Isn't the ACOut port basically the common pole between the AC1 & AC2 port contacts?

    Fourth, the Xantrex documentation worries about a GT inverter back feeding the AC2 genset when off grid and speaks vaguely about some unspecified relay to disconnect the genset. If the XW is using an AGS (Automatic Generator Start), can't it simply shut off the genset - problem solved?

    Finally, the XW6048 can produce 25 Amps per phase (or is it 50?). It appears that the XW hybrid inverters were designed for off grid use, and grid tied/mini grid AC coupled applications were an afterthought. In an AC coupled system, there might be more than 50 Amps or more of generated capacity on the mini grid. Can the XW6048 pass through this power in addition to its sell back? What is the maximum power that it can push back to the grid?

    Thanks-

    Andrew
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: AC Coupled mini-grid (micro-grid)

    Your XW link is to a European 230 vac 50 Hz single phase/non-split phase model.

    Not a U.S. model terminal diagram. U.S. version has L1,N,L2 for each of the ports split phase connections.

    Pass through relay will handle 60 amp. You should not hook GT inverter with output of more then power rating of XW unit. It may work with greater power pushing pass thru to grid but when grid goes down there would be possibility of overload since inverter will not be able to push greater then inverter rating to batteries. If there is nowhere for the PV power to go the AC voltage would likely rise causing both the GT and XW to shut down.

    Split phase 240vac XW is like a 240/120 vac generator, like center tapped transformer. The transformer wire IR loss limits a half side load to maximum of about 41 amps. Total limit for XW6048 is 6kVA max combined load with 41 amps max on one side.

    To your question about selling from DC line and backfeed GT at same time... It could theoretically but not sure how software would react. Selling is managed by measuring power flow between the three ports (AC1in, Inverter, and ACout) and regulating battery voltage. Any attempt to raise battery voltage above XW regulation setting will cause power to be sucked off battery line and pushed to AC output from inverter. When you turn sell off the SW just prevents any power from flowing out AC1in port. It will push to ACout loads if just grid support is enabled.

    I am not very familar with Sunny Island but again you are looking at European 230vac manual. The U.S. version, i believe, is 120 vac so you need two for 240 vac split phase setup.
  • Dr. Strangelove
    Dr. Strangelove Solar Expert Posts: 49
    Re: AC Coupled mini-grid (micro-grid)
    RCinFLA wrote: »
    Your XW link is to a European 230 vac 50 Hz single phase/non-split phase model.

    Not a U.S. model terminal diagram. U.S. version has L1,N,L2 for each of the ports split phase connections.

    Yeah, I'm an idiot. It's impossible (or more difficult than I care to handle) to link to Schneider's online documentation so I searched for the first one I could find online. Here is the US split phase 240VAC manual (see page 2-17).
    Pass through relay will handle 60 amp. You should not hook GT inverter with output of more then power rating of XW unit. It may work with greater power pushing pass thru to grid but when grid goes down there would be possibility of overload since inverter will not be able to push greater then inverter rating to batteries. If there is nowhere for the PV power to go the AC voltage would likely rise causing both the GT and XW to shut down.

    Split phase 240vac XW is like a 240/120 vac generator, like center tapped transformer. The transformer wire IR loss limits a half side load to maximum of about 41 amps. Total limit for XW6048 is 6kVA max combined load with 41 amps max on one side.

    While this all sounds reasonable, and I do not doubt that you are correct, could I trouble you for a source that I can study?
    To your question about selling from DC line and backfeed GT at same time... It could theoretically but not sure how software would react. Selling is managed by measuring power flow between the three ports (AC1in, Inverter, and ACout) and regulating battery voltage. Any attempt to raise battery voltage above XW regulation setting will cause power to be sucked off battery line and pushed to AC output from inverter. When you turn sell off the SW just prevents any power from flowing out AC1in port. It will push to ACout loads if just grid support is enabled.

    I see your point. That seems like a good reason for the transfer relay on the GT inverter output. It seems that a mechanical relay would be very slow. Has anyone implemented a UL listed transfer relay using solid state relays? Yes a double pole, double throw SSR will require power in both directions - for which the AUX power could be used.

    Any thoughts on the XW AGS obviating the need for an additional relay?
    I am not very familar with Sunny Island but again you are looking at European 230vac manual. The U.S. version, i believe, is 120 vac so you need two for 240 vac split phase setup.

    Again, I am 0 for 2. Here is the US 120V Sunny Island version, and page 17 shows 2 Sunny Islands supporting 240VAC split phase. Are there any SMA/Sunny experts on this board? I have found the SMA boards somewhat ideological: "Are there any reasons that don't justify the price of our inverters?"

    Andrew
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    Second, the Sunny Island seems to be a single phase; to make a 240V/60Hz split phase power system, you need two Sunny Islands. Is this correct?

    Correct. IMO a killer reason to opt for the Xantrex if you need split phase functionality.
  • RCinFLA
    RCinFLA Solar Expert Posts: 1,484 ✭✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    While this all sounds reasonable, and I do not doubt that you are correct, could I trouble you for a source that I can study?
    Andrew

    Don't have an info source or actual situational experience, just basic understanding of bi-directional inverter and fact that more PV incoming power then can be consumed means something has got to give.

    Maybe the inverter will just overload trying to push all the power to battery instead of vac rise, maybe Vdc will overvoltage first depending on battery size since a short overload on inverter is allowed, result is same though, protection will kick in and shut it down which, in turn, will shut down GT inverter.
  • ggunn
    ggunn Solar Expert Posts: 1,973 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    stephendv wrote: »
    I'll see your "hehe" and raise you a "shazam" :D
    http://files.sma.de/dl/5612/SI5048-TB-TEN110340.pdf
    It's going to be a tall order to find where they explicitly list the things that aren't necessary, but you can infer from the paragraph about the comms module that it's not a required component:

    Chapter 6.5:
    "You can connect SMA Solar Technology communication devices (e.g., Sunny Boy Control, Sunny WebBox) or a PC with the appropriate software to a communication interface. A detailed cabling diagram can be found in the communication device manual, the software or on the Internet at www.SMA.de/en.
    You can incorporate an RS485 communication interface into the Sunny Island."


    Nowhere in that chapter (or anywhere in the manual) do they describe connecting a sunnyboy to a sunny island via the comms modules.

    Chapter 17 Describes how to setup a sunnyboy for off-grid use, including:
    "Depending on the device type, you can set the PV inverter to "stand-alone grid operation" using Sunny Boy Control, Sunny WebBox, Sunny Data Control or Sunny Explorer."

    If I had to guess, I think this is why the SMA chaps are saying you need the comms module in the sunny boy. But once you've programmed it you can remove the module and keep it for the next install.

    Interesting, thank you. When SMA opens today I will call their tech support line and ask them point blank; when I talked to them when I was installing my first SI/SB system, they knew there was no WebBox in the system and they still told me I had to have the piggyback modules and comm line to do the install. This was separate from the SB parameter change; they knew I had the interface cable, laptop, and Sunny Data Control software to do that. In the installations I have done since then, I have just included the comm connection as a matter of course.

    Film at eleven.
  • maskani
    maskani Registered Users Posts: 7
    Re: AC Coupled mini-grid (micro-grid)

    Hello all, I'm reading all of your post here, they are so useful and I thanks to all. I think Dr. Strenglov said true about sunny webbox, sunny webbox is necessary for programming inverters but after that it can be removed.

    So, I have a question, is it possible to replace Sunny Island with a Ups and V/f Controller? I want to setup a not expensive system, with a 350w microinverter, and 200 or 300 watt UPS, maybe I can design a V/f controller, too. SMA is so good, but it's so expensive. we can't find customer with this prices. we must build a system with lower power and not expensive. if we can build this system, we can sell more and more pv system, and expanding usage of renewable energy in the world.
  • stephendv
    stephendv Solar Expert Posts: 1,571 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    maskani wrote: »
    So, I have a question, is it possible to replace Sunny Island with a Ups and V/f Controller? I want to setup a not expensive system, with a 350w microinverter, and 200 or 300 watt UPS, maybe I can design a V/f controller, too. SMA is so good, but it's so expensive. we can't find customer with this prices. we must build a system with lower power and not expensive. if we can build this system, we can sell more and more pv system, and expanding usage of renewable energy in the world.

    For a budget system, and really for most systems, there isn't a compelling reason to use AC coupling at all. Much simpler and more efficient to go the classic DC route: panels to charge controller to battery bank to inverter.
  • NorthGuy
    NorthGuy Solar Expert Posts: 1,913 ✭✭
    Re: AC Coupled mini-grid (micro-grid)
    stephendv wrote: »
    Much simpler and more efficient to go the classic DC route: panels to charge controller to battery bank to inverter.

    May not be that inefficient. Sunny Boys claim to be 98% efficient, compare to MPPT+Inverter at 85%, so you get much better utilizaton of solar when used directly. When going to batteries, you get (Sunny Boy + Sunny Island) efficuiency at 90%, compared to MPPT at 95%. Depending on the usage pattern, SMA system may turn out to be more efficient.
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Re: AC Coupled mini-grid (micro-grid)
    NorthGuy wrote: »
    May not be that inefficient. Sunny Boys claim to be 98% efficient, compare to MPPT+Inverter at 85%, so you get much better utilizaton of solar when used directly. When going to batteries, you get (Sunny Boy + Sunny Island) efficuiency at 90%, compared to MPPT at 95%. Depending on the usage pattern, SMA system may turn out to be more efficient.

    Standard GTI's are 95+% efficient when converting PV power directly to AC.
    If you use them for charging batteries the efficiency drops because you are converting DC to AC and back to DC again.
    An MPPT charge controller is as efficient at supplying DC for charging batteries as the standard GTI is at supplying AC for grid feed.
    The exception to this is if very high Voltages are needed to overcome long wiring distances, where the typically lower Voltage DC arrays can't do it. Although with the advent of the MidNite Classic 250 the DC arrays can run nearly at the same level as a GTI and with an XW 600 MPPT 80 they can run at 2X the Voltage of a GTI output.

    It's never a "one size fits all" answer. :D