diy solar

diy solar

Enphase AC to hybrid system

Oats49250 said:
This was the day before when I had little solar cumming in. I was trying to run my batteries down to try charging them the next day.


View attachment 65363
Click to expand...

I am still trying to think through what might be going on here. Something is certainly wrong, but I can't see exactly what with the information we have. Looking at this screen grab from the Schneider, it looks like it only just touched zero current, and was down hill into discharging from there. At the very end, I see the battery current hit about -155 amps! With the battery bank reporting 54 volts, that is 8,370 watts, well over the 6,800 watt continuous rating of the XW-Pro 6848. Taking that one dip out of the equation, the other places the solar is not helping, I still see the battery current hitting -100 amps. Even that is over 5,000 watts.

Over this 5 hour span, the battery state of charge went from over 70% to under 40%. So let's say you used up 30% of your battery capacity. I am going to round off a lot, to make the math easy, just to get a rough idea. Your battery is made up of 121 Nissan Leaf battery modules. They seem to be about 400 watt hours each x 121 = about 48 KWHs of battery. Using 30% of that is about 14 KWH in just 5 hours = roughly 2,800 watts of average load on the battery bank. But this is not taking into account that the solar was supplying some power as well. Is the time set properly on your Schneider gear? How were you getting sun on the solar panels past 7 pm? I see the ramp down from 5 pm to 7 pm, and it does sort of look like the ramp down slope I see here at dusk, but then after 7 pm, we still see the current cycling up and down 60 amps! That can't be the solar kicking on and off. Something else is going on here. Without being able to zoom in the data, it looks like something drawing 60 amps it turning on for 10 minutes, and then turning off again. A well pump? An air conditioner? Electric water heater? Whatever it is, this is a 3,000 watt load that is cycling. This is not from the solar inverters kicking off.
 
GXMnow said:
I am still trying to think through what might be going on here. Something is certainly wrong, but I can't see exactly what with the information we have. Looking at this screen grab from the Schneider, it looks like it only just touched zero current, and was down hill into discharging from there. At the very end, I see the battery current hit about -155 amps! With the battery bank reporting 54 volts, that is 8,370 watts, well over the 6,800 watt continuous rating of the XW-Pro 6848. Taking that one dip out of the equation, the other places the solar is not helping, I still see the battery current hitting -100 amps. Even that is over 5,000 watts.

Over this 5 hour span, the battery state of charge went from over 70% to under 40%. So let's say you used up 30% of your battery capacity. I am going to round off a lot, to make the math easy, just to get a rough idea. Your battery is made up of 121 Nissan Leaf battery modules. They seem to be about 400 watt hours each x 121 = about 48 KWHs of battery. Using 30% of that is about 14 KWH in just 5 hours = roughly 2,800 watts of average load on the battery bank. But this is not taking into account that the solar was supplying some power as well. Is the time set properly on your Schneider gear? How were you getting sun on the solar panels past 7 pm? I see the ramp down from 5 pm to 7 pm, and it does sort of look like the ramp down slope I see here at dusk, but then after 7 pm, we still see the current cycling up and down 60 amps! That can't be the solar kicking on and off. Something else is going on here. Without being able to zoom in the data, it looks like something drawing 60 amps it turning on for 10 minutes, and then turning off again. A well pump? An air conditioner? Electric water heater? Whatever it is, this is a 3,000 watt load that is cycling. This is not from the solar inverters kicking off.
Click to expand...
My AC is 3000 watts. I have tried some more experimenting today and got very little solar, before I started my panels were putting about 140 watts and I turned on 32 of them, no ac and my loads were still around 1500 watts and I only had about half of the panels working.
Hedges said:
For my SMA string inverters, default frequency range is 59.4 to 60.5 Hz. That is the standard in-spec frequency range, so your microinverters shouldn't go offline below 60.5 Hz.
Start at that frequency (or higher), would like power output to ramp down with frequency. Mine are 100% at 61 Hz ramped down to 0% at 62 Hz.

If your microinverters do frequency-watts, there should be a similar range of frequency where they ramp down output.
There could also be frequency shift rate, Hz/second that the tolerate, and that the Schneider produces. My battery inverter takes several seconds to swing frequency, easy to watch with hand-held meter. It has to source/sink power while waiting or PV inverters to respond.

When I used a PV inverter that didn't respond correctly, frequency continued to rise until it was knocked offline, then frequency ramped back down.
Click to expand...
The Schneider goes from 60 to 61 and then climes to 69 as the panels come on line.
 
Hedges said:
It should shift and just keep shifting until power hits desired amount. Mine takes a few seconds to shift and GT PV inverters respond immediately. Could be the battery inverter shifted at a rate faster than PV inverter thought was acceptable, so it drops off. Are you sure your microinverters are ramping down power? If you have an AC clamp meter and the shift is slow enough you should be able to watch that.

There may be some parameter settings that work better.
I was looking for numbers like rate of change Hz/second, which I haven't found, but I did find this. Says for off-grid, so PV inverter settings wouldn't be UL1741, but does XW take care of disconnecting them from grid when necessary? Does XW perform anti-islanding? I don't know, but my Sunny Island does. One of my Sunny Boy models I had to set to "offgrid" when running behind Sunny Island, but now I'm using models which work correctly set to "backup".

"For off-grid: As a second option, when it is not possible to set the battery and PV inverter to the same grid code, it is recommended to disable Freq-Watt P(f) Droop and Freq-Watt P (f) PreDist on the battery inverter (Figure 6) in Grid Codes settings. The smart AC-coupled PV inverter's P(f) curve should be set to start active power reduction at nominal frequency 50/60 Hz and finish it (fully reduce to zero) at 54/64 Hz. This is what the battery inverter frequency shift is configured to in case when grid code P(f) is not activated on it."


"2. Select the appropriate region grid code from the Select Region pulldown menu. a. For grid interactive systems, the appropriate grid code is specified by the utility in their interconnection rules. b. For AC-coupled systems, ensure that the AC coupled PV inverters are set to the same grid code. c. Note, off-grid systems may typically use any of the grid codes. For off-grid AC coupled systems, the California Rule no. 21-2018 grid code is recommended for the XW Pro and PV inverters"

"3. Verify that each inverter accepted the correct Grid Code Region by navigating to Devices -> (inverter) -> Grid Codes"

It did take me a couple tries to get parameters accepted by my Sunny Boys (operator error?)

Click to expand...
When I use this I am off grid, I manually shut down the grid and turn on the XW, the XW does everything that you listed that your sunny island. I have to much solar to run through the inverter it is rated for 6000 watts and my solar puts out 12000 to 13000 watts in full sun. So I use it off grid with 1 or 2 strings of 16 each. about 4000 to 6000 watts. I did try all of this with no change. It seems like that should have worked.
 
Hedges said:
It should shift and just keep shifting until power hits desired amount. Mine takes a few seconds to shift and GT PV inverters respond immediately. Could be the battery inverter shifted at a rate faster than PV inverter thought was acceptable, so it drops off. Are you sure your microinverters are ramping down power? If you have an AC clamp meter and the shift is slow enough you should be able to watch that.

There may be some parameter settings that work better.
I was looking for numbers like rate of change Hz/second, which I haven't found, but I did find this. Says for off-grid, so PV inverter settings wouldn't be UL1741, but does XW take care of disconnecting them from grid when necessary? Does XW perform anti-islanding? I don't know, but my Sunny Island does. One of my Sunny Boy models I had to set to "offgrid" when running behind Sunny Island, but now I'm using models which work correctly set to "backup".

"For off-grid: As a second option, when it is not possible to set the battery and PV inverter to the same grid code, it is recommended to disable Freq-Watt P(f) Droop and Freq-Watt P (f) PreDist on the battery inverter (Figure 6) in Grid Codes settings. The smart AC-coupled PV inverter's P(f) curve should be set to start active power reduction at nominal frequency 50/60 Hz and finish it (fully reduce to zero) at 54/64 Hz. This is what the battery inverter frequency shift is configured to in case when grid code P(f) is not activated on it."


"2. Select the appropriate region grid code from the Select Region pulldown menu. a. For grid interactive systems, the appropriate grid code is specified by the utility in their interconnection rules. b. For AC-coupled systems, ensure that the AC coupled PV inverters are set to the same grid code. c. Note, off-grid systems may typically use any of the grid codes. For off-grid AC coupled systems, the California Rule no. 21-2018 grid code is recommended for the XW Pro and PV inverters"

"3. Verify that each inverter accepted the correct Grid Code Region by navigating to Devices -> (inverter) -> Grid Codes"

It did take me a couple tries to get parameters accepted by my Sunny Boys (operator error?)

Click to expand...
1632536039238.png

1632536082007.png
 
This is a pdf of my setting in grid codes. I made some changes that I will try tomorrow.
 

Attachments

  • grid_codes 9-24-21.pdf
    570.2 KB · Views: 3
  • grid_codes updated.pdf
    570.1 KB · Views: 5
Last edited:
That does not make sense.
Battery volts started at about 55.7 volts which is ok, roughly 70% charged, and the current was at zero. Jut before 3 pm the current shoots to -38 amps. That looks like 2,100 watts or so. What was turned on at that point?

Can you do another screen shot, but zoom in the time to cover just 2 pm to 4 pm? You just put your mouse where you want to start on the plot, hold down the mouse button and roll right. When you are where you want the right end, release the mouse button and it should zoom that section to the full plot width.

What were you using to read the frequency? From the InsightLocal web UI, you can just click on Devices and it dhows the power, voltage, and frequency. The sun is down here, so I am running on batteries.

XW-Device.PNG

My load side is outputting 792 watts at 246.24 volts, and frequency was 59.99, it bounces from that up to 60.01
AC1 input power is zero, but I am actually selling about 1,030 watts back out to my main panel. That is why the battery DC Power is showing -1,737 watts. The numbers move around a bit based on loads. I know it can't be putting out more than it is taking from the batteries, but it is pretty close. It does measure over 95% efficient at this moderate load.

I don't have any solar coming in at this time, so I can't show that, but basically, the AC load side power will go into negative. I will try to kick it off grid again this weekend and see if I can capture what it does. But it certainly is acting different than yours.

The XW Grid Codes section does not effect how it acts while off grid. That is how it responds to grid frequency shift while running on grid so the sell to grid output current meets the local grid code. Looking through my setting, I do not see anything for setting how it handles frequency shift when it is creating the local grid for AC coupling.

Here is a Battery Summary screen grab from my last power failure when the solar was making power.GridFailWithSolar.PNG
At 8:36 am, I clicked on force charger state (Bulk) to make it start charging before I left for work. I do not have this automated at this time while I am on grid. You can see the battery current jumped right up to 23.4 amps. But that only lasted for 2 minutes as the grid failed at 8:38 am. The battery current drops to -16 amps discharging. The solar was putting out less that 100 watts total, and half of the panels saw the glitch as a bad grid, so, no power from solar yet. Around 8:47 I was up to 11 panels producing power, but the total was still under 150 watts, so not keeping up with my load. The other 5 panels were stuck in a "grid instability" mode, and would not restart. I sent the command to the Enphase Envoy to change the grid code and they all reset and started coming online one by one. As the sun rose, the current went from discharging to charging, exceeding 12 amp of charge by 9 am, while the solar is also supplying all of the power to run my loads in the backup breaker panel. From 9 am to 10 am, the solar keeps ramping up, but at 9:29 9:39 9:49 and finally 9:59 you can see the charge current dip to about 10 amps. I had left the maximum charge rate still set to just 17% as I normally slow charge while on grid, and I forgot to change it as I was running late for work. I think those dips in the current are where the XW saw the current getting close to the charge current limit, so it started shifting the frequency. 5 of my 16 panels saw it as an unstable grid, but they didn't lock out, they reduced power, and then the frequency went back down, so they went back online. 4 times, exactly 10 minutes apart. It is hard to tell, and the log does not have an entry, but it looks like the grid came back on at either 10:03 am, or 10:12 am as those are the two places I see a small inflection in the traces. It reverts to charging from grid right at 23.4 amps again. But since the solar was online, it was actually getting the charging power from the solar, and any additional solar power was selling back out to the AC1 "grid" input of the XW which is in my main panel helping power the loads that are not on the backup panel. With my charge rate set to just 17%, 23.4 amps (over 1,200 watts), I still end up selling back as high as 2,000 watts out to the grid when in full sun. This rate pushes about 9 KWHs into my battery bank before 4 pm. Since I doubled my battery bank, I now push it to 20% charge rate, and push over 10 KWHs.

Mine is still on the older 1.08 firmware, if anything, yours should actually be working better. Something is either wrong in the setup, wiring, or with the device itself.
 
GXMnow said:
That does not make sense.
Battery volts started at about 55.7 volts which is ok, roughly 70% charged, and the current was at zero. Jut before 3 pm the current shoots to -38 amps. That looks like 2,100 watts or so. What was turned on at that point?

Can you do another screen shot, but zoom in the time to cover just 2 pm to 4 pm? You just put your mouse where you want to start on the plot, hold down the mouse button and roll right. When you are where you want the right end, release the mouse button and it should zoom that section to the full plot width.

What were you using to read the frequency? From the InsightLocal web UI, you can just click on Devices and it dhows the power, voltage, and frequency. The sun is down here, so I am running on batteries.

View attachment 66228

My load side is outputting 792 watts at 246.24 volts, and frequency was 59.99, it bounces from that up to 60.01
AC1 input power is zero, but I am actually selling about 1,030 watts back out to my main panel. That is why the battery DC Power is showing -1,737 watts. The numbers move around a bit based on loads. I know it can't be putting out more than it is taking from the batteries, but it is pretty close. It does measure over 95% efficient at this moderate load.

I don't have any solar coming in at this time, so I can't show that, but basically, the AC load side power will go into negative. I will try to kick it off grid again this weekend and see if I can capture what it does. But it certainly is acting different than yours.

The XW Grid Codes section does not effect how it acts while off grid. That is how it responds to grid frequency shift while running on grid so the sell to grid output current meets the local grid code. Looking through my setting, I do not see anything for setting how it handles frequency shift when it is creating the local grid for AC coupling.

Here is a Battery Summary screen grab from my last power failure when the solar was making power.View attachment 66235
At 8:36 am, I clicked on force charger state (Bulk) to make it start charging before I left for work. I do not have this automated at this time while I am on grid. You can see the battery current jumped right up to 23.4 amps. But that only lasted for 2 minutes as the grid failed at 8:38 am. The battery current drops to -16 amps discharging. The solar was putting out less that 100 watts total, and half of the panels saw the glitch as a bad grid, so, no power from solar yet. Around 8:47 I was up to 11 panels producing power, but the total was still under 150 watts, so not keeping up with my load. The other 5 panels were stuck in a "grid instability" mode, and would not restart. I sent the command to the Enphase Envoy to change the grid code and they all reset and started coming online one by one. As the sun rose, the current went from discharging to charging, exceeding 12 amp of charge by 9 am, while the solar is also supplying all of the power to run my loads in the backup breaker panel. From 9 am to 10 am, the solar keeps ramping up, but at 9:29 9:39 9:49 and finally 9:59 you can see the charge current dip to about 10 amps. I had left the maximum charge rate still set to just 17% as I normally slow charge while on grid, and I forgot to change it as I was running late for work. I think those dips in the current are where the XW saw the current getting close to the charge current limit, so it started shifting the frequency. 5 of my 16 panels saw it as an unstable grid, but they didn't lock out, they reduced power, and then the frequency went back down, so they went back online. 4 times, exactly 10 minutes apart. It is hard to tell, and the log does not have an entry, but it looks like the grid came back on at either 10:03 am, or 10:12 am as those are the two places I see a small inflection in the traces. It reverts to charging from grid right at 23.4 amps again. But since the solar was online, it was actually getting the charging power from the solar, and any additional solar power was selling back out to the AC1 "grid" input of the XW which is in my main panel helping power the loads that are not on the backup panel. With my charge rate set to just 17%, 23.4 amps (over 1,200 watts), I still end up selling back as high as 2,000 watts out to the grid when in full sun. This rate pushes about 9 KWHs into my battery bank before 4 pm. Since I doubled my battery bank, I now push it to 20% charge rate, and push over 10 KWHs.

Mine is still on the older 1.08 firmware, if anything, yours should actually be working better. Something is either wrong in the setup, wiring, or with the device itself.
GXMnow said:
That does not make sense.
Battery volts started at about 55.7 volts which is ok, roughly 70% charged, and the current was at zero. Jut before 3 pm the current shoots to -38 amps. That looks like 2,100 watts or so. What was turned on at that point?

Can you do another screen shot, but zoom in the time to cover just 2 pm to 4 pm? You just put your mouse where you want to start on the plot, hold down the mouse button and roll right. When you are where you want the right end, release the mouse button and it should zoom that section to the full plot width.

What were you using to read the frequency? From the InsightLocal web UI, you can just click on Devices and it dhows the power, voltage, and frequency. The sun is down here, so I am running on batteries.

View attachment 66228

My load side is outputting 792 watts at 246.24 volts, and frequency was 59.99, it bounces from that up to 60.01
AC1 input power is zero, but I am actually selling about 1,030 watts back out to my main panel. That is why the battery DC Power is showing -1,737 watts. The numbers move around a bit based on loads. I know it can't be putting out more than it is taking from the batteries, but it is pretty close. It does measure over 95% efficient at this moderate load.

I don't have any solar coming in at this time, so I can't show that, but basically, the AC load side power will go into negative. I will try to kick it off grid again this weekend and see if I can capture what it does. But it certainly is acting different than yours.

The XW Grid Codes section does not effect how it acts while off grid. That is how it responds to grid frequency shift while running on grid so the sell to grid output current meets the local grid code. Looking through my setting, I do not see anything for setting how it handles frequency shift when it is creating the local grid for AC coupling.

Here is a Battery Summary screen grab from my last power failure when the solar was making power.View attachment 66235
At 8:36 am, I clicked on force charger state (Bulk) to make it start charging before I left for work. I do not have this automated at this time while I am on grid. You can see the battery current jumped right up to 23.4 amps. But that only lasted for 2 minutes as the grid failed at 8:38 am. The battery current drops to -16 amps discharging. The solar was putting out less that 100 watts total, and half of the panels saw the glitch as a bad grid, so, no power from solar yet. Around 8:47 I was up to 11 panels producing power, but the total was still under 150 watts, so not keeping up with my load. The other 5 panels were stuck in a "grid instability" mode, and would not restart. I sent the command to the Enphase Envoy to change the grid code and they all reset and started coming online one by one. As the sun rose, the current went from discharging to charging, exceeding 12 amp of charge by 9 am, while the solar is also supplying all of the power to run my loads in the backup breaker panel. From 9 am to 10 am, the solar keeps ramping up, but at 9:29 9:39 9:49 and finally 9:59 you can see the charge current dip to about 10 amps. I had left the maximum charge rate still set to just 17% as I normally slow charge while on grid, and I forgot to change it as I was running late for work. I think those dips in the current are where the XW saw the current getting close to the charge current limit, so it started shifting the frequency. 5 of my 16 panels saw it as an unstable grid, but they didn't lock out, they reduced power, and then the frequency went back down, so they went back online. 4 times, exactly 10 minutes apart. It is hard to tell, and the log does not have an entry, but it looks like the grid came back on at either 10:03 am, or 10:12 am as those are the two places I see a small inflection in the traces. It reverts to charging from grid right at 23.4 amps again. But since the solar was online, it was actually getting the charging power from the solar, and any additional solar power was selling back out to the AC1 "grid" input of the XW which is in my main panel helping power the loads that are not on the backup panel. With my charge rate set to just 17%, 23.4 amps (over 1,200 watts), I still end up selling back as high as 2,000 watts out to the grid when in full sun. This rate pushes about 9 KWHs into my battery bank before 4 pm. Since I doubled my battery bank, I now push it to 20% charge rate, and push over 10 KWHs.

Mine is still on the older 1.08 firmware, if anything, yours should actually be working better. Something is either wrong in the setup, wiring, or with the device itself.
Click to expand...

I use the insight local to see what the voltage and frequency are doing. It sound like you are having the same problem I am having, is that I can only get about half of my inverters to come online. My loads are so high that most of my power is used up and little is left for charging.
Click to expand...

1632659365186.png
 
I don't like using to much of my solar up, since I need to get extra credits for the fall. So I usually test in the after noon when I am done working and have little power coming in. I did test it yesterday late afternoon and was able to get most of my panels to work, but due to the sun setting and my loads I didn't get above my chagrining threshold. If I have my absorption voltage set to 57.4 how far below that do I have to go before my batteries should start to charge? Or will it start at what ever voltage I my recharge voltage set to?

1632659676244.png
 
I am starting to think you have something wired wrong. How is the XW-Pro wired into the system? What connects to AC1, AC2, and the Load Output terminals? Where is your main panel, backup loads, and Enphase solar connected?

For the XW to go into charge mode by itself, the battery voltage has to drop below the "Recharge Voltage" setting under the charger settings menu. But you can force it to charge by going to controls, select "Bulk" under Force Charger State and then click apply. But for that to work, the power has to be coming in on the AC1 or AC2 terminals. When running in AC coupled off grid, the Solar power should be on the output side of the XW and if the solar is producing more than the loads demand, the extra will just flow to the batteries. It is not a charge mode, it is just stabilizing the system by taking the extra power needed to keep the voltage at 240/120. If it creeps up, it pulls more current and feeds the batteries, if the voltage dips, it will pull from the battery to keep the voltage up.
 
GXMnow said:
I am starting to think you have something wired wrong. How is the XW-Pro wired into the system? What connects to AC1, AC2, and the Load Output terminals? Where is your main panel, backup loads, and Enphase solar connected?

For the XW to go into charge mode by itself, the battery voltage has to drop below the "Recharge Voltage" setting under the charger settings menu. But you can force it to charge by going to controls, select "Bulk" under Force Charger State and then click apply. But for that to work, the power has to be coming in on the AC1 or AC2 terminals. When running in AC coupled off grid, the Solar power should be on the output side of the XW and if the solar is producing more than the loads demand, the extra will just flow to the batteries. It is not a charge mode, it is just stabilizing the system by taking the extra power needed to keep the voltage at 240/120. If it creeps up, it pulls more current and feeds the batteries, if the voltage dips, it will pull from the battery to keep the voltage up.
Click to expand...
AC1 is my incoming power AC2 is my generator and AC loads is connected to my sub panel in my barn which is fed by a sub panel on the outside of my barn which combines the solar sub panel from enphase and my power coming from my house.
 
Last edited:
So I ran a test today and it work for awhile, I was trying to adjust my setting in the grid codes of the XW. It still was limiting the number of inverters that were working but it was allowing the ones that were working to come up to full power.
As the voltage started to clime it started to drop in current, I thought maybe a cloud went over, but it was a cloud free day.
The system charges fine on the AC1 and 2 input, but on the load input the XW sits at 60.1 for long periods of time, I think interfering with the enphase inverters. Makeing changes in the grid codes seemed to make a difference, I have never seen more than 5 amp before.

1632708224871.png
 
This looks a lot better. It hit 20 amps of charge current on the first run. And then 40 amps on the second run.

Why did the gateway lose data, twice? What was going on with the voltage, current, and frequency at the output of the XW? The first run does show the power ramp up and then start ramping down. The second run started in discharge, went to charging, but then fell to 5 amps of charge, and then to 10 amps of discharge. Did some loads cycle on? Did the system shut down on it's own, or did you shut it off?
 
GXMnow said:
GXMnow said:
I bought my Schneider XW-Pro from "Real Goods" which is a California arm of Alt E Store. I caught a 10% off sale and free shipping.

If I was going to do it again.. I may have gone with the Outback Skybox. It costs a bit more, and is a little less power, but everything you need is all in one box. I also had to buy a UL rated conduit box and the gateway just to get the XW-Pro connected. The Skybox even has a DC coupled MPPT controller if you want to add a few DC panels, but it also works just AC coupled. A Schneider MPPT controller is close to $1,000.

As a backup inverter, the Schneider is hard to beat. This thing is very solid. It is built to take it and has a 10 year warranty once you register it. It is pretty efficient and puts out solid clean power. My only regret is the time of use power shifting issue. From what I understand, the Outback Skybox handles that without a problem. The Sol Ark did not recommend running my 4,800 watts of solar panels in an AC coupled only setup. I am not sure if that is a serious limitation, or just an efficiency issue or what.

I have not done any kind of long term test yet, but from the descriptions, all of the majors that are rated for AC coupling will do the frequency shift power curtailment on solar inverters that support it. I have not had mine in a backup off grid mode long enough with decent sun to see it go into that mode yet. Basically, if the solar power coming in exceeds what the loads need, it will use the extra power to charge the batteries. If this power is greater than the maximum charge rate, or if the batteries are getting full, the inverter will shift the line frequency up. Most newer solar inverters should be able to be set in a grid profile that will cause the inverter to curtail it's output as a ratio of the line frequency. Enphase calls it Frequency/Watt control. I have that turned on by using the California "Rule 21" grid profile. As I am in So Cal, I need to be in this mode anyways to be legal.
Click to expand...

I bought my Schneider XW-Pro from "Real Goods" which is a California arm of Alt E Store. I caught a 10% off sale and free shipping.

If I was going to do it again.. I may have gone with the Outback Skybox. It costs a bit more, and is a little less power, but everything you need is all in one box. I also had to buy a UL rated conduit box and the gateway just to get the XW-Pro connected. The Skybox even has a DC coupled MPPT controller if you want to add a few DC panels, but it also works just AC coupled. A Schneider MPPT controller is close to $1,000.

As a backup inverter, the Schneider is hard to beat. This thing is very solid. It is built to take it and has a 10 year warranty once you register it. It is pretty efficient and puts out solid clean power. My only regret is the time of use power shifting issue. From what I understand, the Outback Skybox handles that without a problem. The Sol Ark did not recommend running my 4,800 watts of solar panels in an AC coupled only setup. I am not sure if that is a serious limitation, or just an efficiency issue or what.

I have not done any kind of long term test yet, but from the descriptions, all of the majors that are rated for AC coupling will do the frequency shift power curtailment on solar inverters that support it. I have not had mine in a backup off grid mode long enough with decent sun to see it go into that mode yet. Basically, if the solar power coming in exceeds what the loads need, it will use the extra power to charge the batteries. If this power is greater than the maximum charge rate, or if the batteries are getting full, the inverter will shift the line frequency up. Most newer solar inverters should be able to be set in a grid profile that will cause the inverter to curtail it's output as a ratio of the line frequency. Enphase calls it Frequency/Watt control. I have that turned on by using the California "Rule 21" grid profile. As I am in So Cal, I need to be in this mode anyways to be legal.
Click to expand...
@GXMnow
Been trying to 'digest' your posts and make the most out of that precious source of information. So I thank you for your posts.
Could you share your wisdom on this challenge?
I have 15 400W panels with IQ7 microinverters recently commissioned. (Yay!)
My next project is to use 18 x LiFePo4 BYD 24V (~3.5kWh/each) recycled batteries for backup (power cuts) and hopefully off-grid use during high-peak prices (sub panel loads)
How you recommend me to tie all together using a Outback Skybox? I figure this all-in-one inverter would be easy to install and operate.
One thing I couldn't figure out, though, is how they connect to my microinverters.
Or should I just set up a few more panels just for this purpose and keep it simple?
 
15 x 400 watts = 6,000 watts. That is acceptable to AC couple to a Skybox as long as the iQ7's are set to the Rule 21 grid code with Frequency Watt control. Since those are 400 watt panels, are they 72 cell? IF they are, are those iQ7 A or + inverters? The straight iQ7 is only rated for 60 cell panels. 72 cell would hit over voltage limits. Those 18 batteries at 3.5 KWH each works out to 63 KWH. That is a lot of backup power.
By my quickie math, that totals well over 1,000 amp hours at 48+ volts. I am at 720 amp hours with my system, about 36 KWHs. You should double check with BYD if it is okay to series them in pairs to make 48 volts. That can be a problem with the BMS in many packaged 12 and 24 volt batteries.

I have not actually setup a Skybox myself. We do have another member on the forum who is using one with AC coupling. Hopefully he will see this thread and step in if I have anything wrong.

The wiring should be very similar to what I have with my Schneider XW-Pro inverter.

You need to start by figuring out what circuits you want to have on the battery backup. The Skybox can run 5,000 watts from battery, but does not have much overload capacity past that. So this is what you need to work with. If you feel you need more power than that during a blackout, then you will have to look at something more powerful. The Skybox should be able to double stack for 10,000 watts, but last I saw, they were not supporting both stacking and AC coupling at the same time. This may change with new firmware though. But for now, let's figure the limit is a single 5,000 watt setup. This is going to be a new breaker panel for all of your loads that you want to run during a blackout. If I was planning again, I would have used a large panel. I already had a 6 space (up to 12 half width breakers) 100 amp Square D panel, so I used what I had from my old generator backup system. I have a pair of 2 pole full size breakers, one is the main for the feed from the battery inverter, and the other is the breaker where the Enphase solar is connected. That just took up 2/3's of the spaces in my panel. So I used 2 dual breakers (half width tandems) to give me 4 back circuits. They feed (1. my refrigerator and gas stove igniter), (2. my PC and internet devices and the gas furnace), (3. the bathroom lights, garage door opener, and the microwave), and (4. the living room lights Dish network and TV). Each circuit also has a few extra outlets, so I can move over plugs if I need to run something else. It would be nice to have 2 more circuits, but this does work fine. If I was doing this again, I would have used at least a 12 full size space panel so I would have room to grow. With just 5,000 watts, you are a bit limited while on battery at night, so keep that in mind when you decide what to back up. But when the sun is shining, you do have all the power from the solar also going into that backup panel, so you can run much more during the day, and also the Skybox will pass through grid current while it is live, so the power limit is only a real concern when the grid and sun are both down.

AC coupled battery inverters like my Schneider XW-Pro and the Skybox are able to provide a valid grid on their output side for the grid tie solar inverters to sync up and produce power. The input side of the battery inverter is equipped with a transfer relay that disconnects from the grid input when it detects a grid problem. I have seen a few minor glitches where the power didn't actually fail, by my system went to battery mode for a minute or so because the grid went out of spec. But for the most part, it just sits there waiting for a grid failure. Your solar will be feeding the output side of the battery inverter and it's power will first feed any loads that are in your backup breaker panel. If the solar is making more than your loads in the backup panel, then the extra current will feed back into the battery inverter. If the grid is up, that power will go back into your main panel, just like it does now with the solar connected to the main panel. If the battery inverter is charging, it is taking it's charge power from the excess that would be going back to the main panel. If the solar is weak, and it is still trying to charge, it can also take grid power to charge, by we should adjust the settings to minimize that.

If the grid goes down, this is where things get interesting.
Let's say it is at night, and you lose grid power. That's easy, the battery inverter transfer relay isolated from the grid, and then it starts using battery power to produce the 120/240 power to feed the back loads breaker panel. The switchover is typically quick enough that you hardly notice a glitch. So now your essential loads are running off of the batteries. When the sun comes up, the Enphase iQ inverters see a good "grid" because they are on the output of the battery inverter, so they will start to produce electricity. As they ramp up, they solar inverters will start to power your backup loads, and the power needed from the batteries will drop off. If the solar ends up making more than the backup loads need, then the battery inverter becomes a charger and the extra power get's pushed back into the batteries. Here is another thing you need to think about. Things do not happen instantly. If there is a large load running, and you have lots of sun shining, it's all good, but then that large load turns off. The solar is still making all that power. The battery inverter will very quickly grab the power and start pushing it to the batteries. So you need a battery bank that can tolerate all of the power from the solar array. In your case, that is a potential of 6,000 watts from 15 x 400 watt panels. At 48 volts, that means, it could try to stuff 125 amps into the batteries. As long as the batteries can take that, the system should be stable. After a short period of time, the inverter can start using frequency shift to start controlling the charge rate. The battery inverter can command the Rule 21 compliant solar inverters to scale back their output by raising the line frequency. Here in the USA, it can shift from 60 Hz up to say 61.5 or 62 Hz to reduce the solar current to safely charge and top op the batteries. It is possible some of the solar inverters may see it as bad power and shut off, and if it does, the battery inverter will either reduce charging, or even flip back to inverting to make sure the loads get enough current. As soon as the current drops, it will also ramp the frequency back towards normal to get the solar inverters to make power again.

When the grid comes back on, the battery inverter will qualify the power as good, and sync up and reconnect to the grid.

I have now had 3 real power failures here, and my Schneider has worked perfectly. One time, we did not even know the power had gone out. I just happened to notice the Schneider dashboard showed my battery was discharging because it was early, and quite overcast, so the solar was not making much power yet. Another time, my girlfriend had thought the garbage disposal had failed because it wouldn't turn on. I don't have that on the backup power. The solar was running all the backup loads and charging the batteries.
 
luisfcosta said:
@GXMnow
Been trying to 'digest' your posts and make the most out of that precious source of information. So I thank you for your posts.
Could you share your wisdom on this challenge?
Click to expand...
I agree...@GXMnow you are a great person that spends tons of time to write out helpful replies but I never have time to read them.... Maybe a summary, then less details? Thanks for all you do...
 
@GXMnow
Thank you for your words of wisdom. Yes, they are 72 cells/each and IQ7+ inverters. Batteries are recycled and range 3-3.5kWh each.
So I think my next challenges will be:
- Find stackable inverters to reach 10kW, or a single one (big enough). Frequency shifting, off-grid, AC coupling, ...
- Move essential loads to a loads panel, matching ~ inverter(s) power
- Assemble the 24V battery pack (48V system) and install an intelligent BMS to handle the input current (125A)
Interesting how the setup will work - in theory. I hope I can 'tell' the system to prioritize battery charging (to face a coming power outage, very frequent here in NorCal), or to do night discharges (essential loads)
Quick question: The essential loads panel will be only fed by the inverter and won't be connected to the grid, right? I can potentially leave other major loads running directly from the grid (car charger, heat pump,...) at the same time the inverter+battery are feeding the essential loads?!
Thanks a ton!
 
If your concern is having power for an outage, I have to say the Schneider XW-Pro works great for that. The inverter can push 6,800 watts continuous, with a 12,000 watt surge, and even 8,500 watts for several minutes. It is a much stronger inverter than most for it's size and ratings. My only gripe with it is the software has some limitation for trying to do energy time shifting. I want to time shift 10 KWH's from when my solar is producing to the 4 pm to 9 pm peak rate time, and with only AC coupling, the system just won't do it on it's own. You either need manual intervention, or an external controller to do it. But for a backup power system, it will keep the batteries charged up over your "Recharge Volts" setting, and if there is a grid failure, it switches over and seamlessly keeps your backup loads running. You can stack up to 3 of them if you need crazy power. With 2 units stacked, you do not need any external components as the internal transfer relays are rated to 60 amps. With a third unit, they recommend an external transfer relay.

During the power failure, a single unit can handle 6,000 watts of AC coupled solar and I did see mine do the frequency shift to control charge power, so I know that works. It seamlessly went from charging to discharging as the amount of load and solar changed.

While the grid is up, you have several choices for how the loads are powered. If you have it setup just as backup power, the grid energy will flow through from the AC1 input to the output and it will keep the batteries charged. If you set the recharge volts lower, and turn on grid support, then it can use battery power to help power the loads on the output side. Grid support in the Schneider works by using battery power to keep the grid input power below the limit you set. For example, you could set the grid support to 2 amps. If your backup loads are pulling less than 2 amps, all of the power will come from the grid input. But if your backup loads are pulling 10 amps, then the inverter will will supply 8 amps to the loads so the grid is only supplying 2 amps. If you turn on "Grid Sell" then you can tell it to back feed battery power out the AC1 input to your main panel to help supply loads there, and even sell out to the grid.

When you have AC coupled solar like I do, it does change things a little. When the grid is up, the XW-Pro has no way to control the AC coupled solar. So lets say you have grid support set to 2 amps, and grid sell set to 5 amps. And you have your 6,000 watts of AC coupled solar. If the solar is making 4,800 watts, that would be 20 amps being fed into the backup loads panel. If the backup loads are only using 5 amps, the other 15 amps will flow backwards through the XW-Pro and go back to the main panel and the grid. The XW-Pro will basically just sit there in standby as all of the power demands are being met. As the solar power fall off in the evening, The XW will start to use battery power to maintain the 5 amps of grid sell. If the battery falls below the sell limit, or grid sell is blocked, then it will go back to only supplying the loads in the backup panel.

The Skybox, Radian, and Sol-Ark all should work similar to this, but I only have direct hands on experience with the Schneider XW-Pro.
 
luisfcosta said:
- Find stackable inverters to reach 10kW, or a single one (big enough). Frequency shifting, off-grid, AC coupling, ...
Click to expand...

GXMnow said:
My only gripe with it is the software has some limitation for trying to do energy time shifting. I want to time shift 10 KWH's from when my solar is producing to the 4 pm to 9 pm peak rate time, and with only AC coupling, the system just won't do it on it's own.

The Skybox, Radian, and Sol-Ark all should work similar to this, but I only have direct hands on experience with the Schneider XW-Pro.
Click to expand...

Sunny Island is another (I use with Sunny Boy). Get two of them for 12kW, and they do frequency-shift for AC coupling.
I don't think they have time-shifting function built in.
 
You must log in or register to reply here.

Similar threads

A
Connecting new system to existing Enphase setup
Replies
4
Views
407
wheisenburg
wheisenburg
wheisenburg
IQ8 Grid Profile for AC Coupled?
Replies
2
Views
202
wheisenburg
wheisenburg
O
Enphase IQ7x Micro Inverter FW Update
Replies
4
Views
316
solar8484
S
R
Hybrid grid/solar system
Replies
9
Views
466
RockOn
R
L
Chinese "Pure" sine wave AIO inverter AC quality.
2
Replies
23
Views
935
timselectric
T
Back
Top