diy solar

diy solar

XW Pro Grid Support limitations?

So even with the wattnode, power is still leaking back to grid? I would rather pay for a several KW a day (if enough Solar) than deal with grid-tie agreement.

Well, I wasn't fully clear. With the wattnode, you can set a limit of how much can leak back, OR a minimum amount you always want to pull from the grid. So, IF you are ok with always pulling "at least X watts" from the grid, you can likely eliminate the leakage.

As an example, I have an air fryer, and a microwave on the same leg (different circuits, same leg)
If I happen to run both at the same time (unusual, but not impossible) I can have somewhere near 15-20 amps of imbalance from my normal loads. If I wanted to, I could set the minimum utility draw to be "-1800" (15a X 120v) and this would make it so that if the power draw from the grid was under 1800w, the inverter would curtail and produce "0w", above 1800w, it'd start producing and selling back to the main panel, but not to the grid.

My feature request above would allow the system to stop monitoring it as "the sum of both legs", and instead monitor that either leg is leaking. if so, curtail to the point where neither leg is pushing back to the grid. This would allow you to set something more like -300, and get much closer to "net zero" without leaking to the grid one one leg because you're pushing it too close.
 
I'm going to try an experiment.
(Will take a few weeks to get the HW in place.)
Dispensing w/ all the 120v loads on the inverters, converting all to 240v.
Interested to see if there is any effect on the leakage back to grid.
 
I consistently have a fair bit of imbalance on my system. Right now it is showing 3.7 amps going out L2 while L1 is drawing in 3.3 amps. The So Cal Edison power meter shows less than 20 watts. And my hourly usage graphs from SCE also agree that it is averaging the 20 watts I am seeing locally. In my case, I have it exporting that 20 watts as I do have a net energy agreement, but it is just a value change to make it importing 20 watts instead. I doubt anyone is going to ever notice the tiny import export swings and they only bother to report the average total for the hour. Starting and stopping a motor could cause some current reversal from time to time, but the long term average is going to stay importing from the grid.
 
My loads are also pretty imbalanced. The grid load normally sits around 80-100 watts from the grid. As in, I'm constantly buying a little power.

Right now, the grid load is 1.7 amps on L1 and -1.6 on L2.
 
I'm going to try an experiment.
(Will take a few weeks to get the HW in place.)
Dispensing w/ all the 120v loads on the inverters, converting all to 240v.
Interested to see if there is any effect on the leakage back to grid.

I suspect your results will be just like mine. 240v loads will be pretty evenly balanced, and thus.. the only time you would "leak" would be for a second or two after a large load stops, and the inverter scales down the output. That takes a moment.

The "real" leakage we're discussing comes from imbalanced loads caused by 120v loads. It's generally this 120v imbalance that will occasionally see 1 leg importing, and 1 leg exporting a handful of amps.
 
Received the equipment necessary to conduct the experiment.
Dispensing w/ all the 120v loads on the inverters, converting all to 240v.
Looking to see if this effects the "leakage" back to the grid.

I found this voltage converter on a "inventory liquidation" sale, cheap, free shipping.


Because it was free shipping, I did not look closely @ the weight,
seemed to be "52" something.
It was kg....114lbs.
I opened it up, lots of copper.
A 240 transformer, and a 208 for data center use.
Have moved the 120v loads to the converter, will see the results over time.
vertiv_liebert_voltage_converter copy.jpg
 
Yeah I placed an AutoTransformer on my SW and it made a pretty big difference in AC Support Mode as to how much it pulled form Battery/Solar and how much it pulled from Grid for the loads. Issue does indeed seem to be load balancing.
 
I saw some small leak yesterday while messing with some setting. How would the utility notice this? Wouldn't this get used up on the main panel as load?

1702680523086.png
 
I saw some small leak yesterday while messing with some setting.
What settings?
How would the utility notice this?
Depends on your utility. I believe their smart meters can set to alert if the energy flows the wrong way at all or past some trigger point they set.
Wouldn't this get used up on the main panel as load?

View attachment 183194
Depends on your set up and main panel loads, but yes, probably.
 
What settings?
Not sure exactly what triggered it. Was messing with modbus
Depends on your utility. I believe their smart meters can set to alert if the energy flows the wrong way at all or past some trigger point they set.

Depends on your set up and main panel loads, but yes, probably.
Good to know. I always have something running so hopefully these leaks are very small.
 
I saw some small leak yesterday while messing with some setting. How would the utility notice this? Wouldn't this get used up on the main panel as load?

View attachment 183194
This is exactly what the "Zero sell". or "zero export" feature does when you add a wattnode and enable it in the insight device. It "exports" back to the main panel, until the current is around whatever you specified. In my case, I leave it at about -400 to make sure I'm always drawing something from the grid.
 
@ohsolar From your graph, it looks like the XW sent just 0.62 KWHs to your main panel on that one day. That is very small indeed. What loads are in your main panel? That could be 620 watts for one hour, That could show as an export. Or it could have been 62 watts for 10 hours. That would be a lot less likely to show at the grid. Or if you hit the wrong setting in the XW, it could even have been just 10 minutes, but pushing a healthy 3,720 watts. The grid would certainly see that.

Go to your battery summary page on Insight Local. Pick that date Dec. 14th and look at your battery discharge current graph. How far negative does it go, and for how long? Here is mine for the same Dec 14th day.

XW-BatSum12-14-23.JPG
The green trace is battery current. My largest inverter power happened just before 9 pm. The current went to -44 amps for 2 minutes. I think that was the microwave oven and the toaster at the same time. Or maybe my girlfriend used her hair dryer. In any case, the XW had to invert 54 volts x 44 amps = 2,376 watts to keep up with the loads in my house at that moment. You can see how that yanked the battery voltage down too. In my case, I do have my own export limiting setup. I have it set so I am exporting 20 watts to grid most of the day, and up to 80 watts when the Enphase solar is making more than I need. The XW charged from Enphase power from 8:22 am to 3:32 pm. But I used a little grid powe after 2 pm because it only goes down to 7 amps of charge current when clouds/night moves in. At 3:30 I let it stop charging and start doing grid support and seel to grid. So it goes right from 7 amps of charging to 5 amps of discharging into the inverter.

The "Energy Comparison" page might also show you what hour it happened in. Since I have the AC coupled Enphase inverters, the Grid Sell stats also include all of the power that goes through the XW to get to the grid. That is how it shows me discharging more power than I charge. The XW does not see the power from my DC solar controller.

Here i my Energy Comparison. It only sows a single number for the whole hour, so you don't get the detail of the battery summary, but it can give you a lot more data.
XW-EnergyComp.JPG
Here you can see that I was only putting energy into the battery from in the 8 am hour to the 3 pm hour with the orange bars. The yellow bars are the energy coming back out of the batteries. That floats between 700 an 800 watts this time of year. The blue bars shows that power l have been inverting is almost always exporting a little to grid. But it is only 69 watts for the first few hours. When the PV solar takes over, the export goes up, because the XW is seeing the power going through it from output back to the grid. At this point, the XW is charging, not inverting.Gren is power coming in from the grid. There is not much on that graph. Just a iny bit at 8 am and 3 pm. This happens when te PV current falls below the house demand and the minimum 7 amp charge current. Rather than stop / start charging, this is "Super off Peak" time, so I just let it pull that grid current during the transitions.

Here is how this all looked from Southern California Edison.
Unlike @n2aws I have mine set to always export a little since I do have a net metering deal which is pretty good.
But I still keep it very low, I want to keep power in the battery for the whole night until the next sunrise, or a grid failure.

SCE-usage-12-14-23.JPG
This is basically the data from the grid power meter at my house. You can easily see the 20 watts of export the whole time I am running on battery. And when the sun was up, I le it export to just over 100 watts. Again at 3 pm, here you can see I bought the super off peak power, but only 140 watts averaged over the hour. Total for the day ended up exporting 770 watt hours, or 0.77 KWHs. Very little. As the sun was going down, the Enphase system was not covering the loads AND the 7 amps of battery charge current. In the past, I would have it turn off charging right away, but I figured it was better to get some energy into the battery on these short solar days. When the days are longer, and the sun falls after 4 pm, then it will stop charging right away, so I don't buy any of the on peak power from 4 pm to 9 pm.

If your electric company will supply you with hourly charts like that, you can see exactly where you are using or exporting too much power. This is different from the XW-Pro grid energy graph because of the power used in my main panel. Most of the time, that was only 50 watts. All my normal running loads are in my backup panel, so they are covered by the inverter power output. I didn't have it graph that, but you can pick that on the Energy Comparison page.
 
@ohsolar From your graph, it looks like the XW sent just 0.62 KWHs to your main panel on that one day. That is very small indeed. What loads are in your main panel? That could be 620 watts for one hour, That could show as an export. Or it could have been 62 watts for 10 hours. That would be a lot less likely to show at the grid. Or if you hit the wrong setting in the XW, it could even have been just 10 minutes, but pushing a healthy 3,720 watts. The grid would certainly see that.

Go to your battery summary page on Insight Local. Pick that date Dec. 14th and look at your battery discharge current graph. How far negative does it go, and for how long? Here is mine for the same Dec 14th day.
hi @GXMnow,
It was for about 50 mins starting around 4 p.m. At 8 p.m. my charge cycle starts. As I remember it, the dashboard flow was showing it charging and discharging at the same time. I think it matches the chart below. So it was .62 for the whole day where on other days it is like .03 / .04. I just happen to monitor it during this time and kinda had a small panic :)

1702734280705.png
View attachment 183236
The green trace is battery current. My largest inverter power happened just before 9 pm. The current went to -44 amps for 2 minutes. I think that was the microwave oven and the toaster at the same time. Or maybe my girlfriend used her hair dryer. In any case, the XW had to invert 54 volts x 44 amps = 2,376 watts to keep up with the loads in my house at that moment. You can see how that yanked the battery voltage down too. In my case, I do have my own export limiting setup. I have it set so I am exporting 20 watts to grid most of the day, and up to 80 watts when the Enphase solar is making more than I need. The XW charged from Enphase power from 8:22 am to 3:32 pm. But I used a little grid powe after 2 pm because it only goes down to 7 amps of charge current when clouds/night moves in. At 3:30 I let it stop charging and start doing grid support and seel to grid. So it goes right from 7 amps of charging to 5 amps of discharging into the inverter.
I see. I have my AC unit hooked up(240v) but not in use yet. From what I read on this thread, I would only see this when there is an in-balance between the legs which would mostly happen with 120v loads. Wondering what the difference in AC loads would be between the legs.
The "Energy Comparison" page might also show you what hour it happened in. Since I have the AC coupled Enphase inverters, the Grid Sell stats also include all of the power that goes through the XW to get to the grid. That is how it shows me discharging more power than I charge. The XW does not see the power from my DC solar controller.

Here i my Energy Comparison. It only sows a single number for the whole hour, so you don't get the detail of the battery summary, but it can give you a lot more data.
yeah about .6 for the hr
1702734831124.png
View attachment 183237
Here you can see that I was only putting energy into the battery from in the 8 am hour to the 3 pm hour with the orange bars. The yellow bars are the energy coming back out of the batteries. That floats between 700 an 800 watts this time of year. The blue bars shows that power l have been inverting is almost always exporting a little to grid. But it is only 69 watts for the first few hours. When the PV solar takes over, the export goes up, because the XW is seeing the power going through it from output back to the grid. At this point, the XW is charging, not inverting.Gren is power coming in from the grid. There is not much on that graph. Just a iny bit at 8 am and 3 pm. This happens when te PV current falls below the house demand and the minimum 7 amp charge current. Rather than stop / start charging, this is "Super off Peak" time, so I just let it pull that grid current during the transitions.

Here is how this all looked from Southern California Edison.
Unlike @n2aws I have mine set to always export a little since I do have a net metering deal which is pretty good.
But I still keep it very low, I want to keep power in the battery for the whole night until the next sunrise, or a grid failure.

View attachment 183249
This is basically the data from the grid power meter at my house. You can easily see the 20 watts of export the whole time I am running on battery. And when the sun was up, I le it export to just over 100 watts. Again at 3 pm, here you can see I bought the super off peak power, but only 140 watts averaged over the hour. Total for the day ended up exporting 770 watt hours, or 0.77 KWHs. Very little. As the sun was going down, the Enphase system was not covering the loads AND the 7 amps of battery charge current. In the past, I would have it turn off charging right away, but I figured it was better to get some energy into the battery on these short solar days. When the days are longer, and the sun falls after 4 pm, then it will stop charging right away, so I don't buy any of the on peak power from 4 pm to 9 pm.

If your electric company will supply you with hourly charts like that, you can see exactly where you are using or exporting too much power. This is different from the XW-Pro grid energy graph because of the power used in my main panel. Most of the time, that was only 50 watts. All my normal running loads are in my backup panel, so they are covered by the inverter power output. I didn't have it graph that, but you can pick that on the Energy Comparison page.
this looks ok. From my provider
1702735107182.png

Thanks for all the info. One more thing to keep a watch on :)
 
This is exactly what the "Zero sell". or "zero export" feature does when you add a wattnode and enable it in the insight device. It "exports" back to the main panel, until the current is around whatever you specified. In my case, I leave it at about -400 to make sure I'm always drawing something from the grid.
Interesting.. Is the 400w being pulled by the wattnode? I don't have solar.
 
Interesting.. Is the 400w being pulled by the wattnode? I don't have solar.
No, the wattnode just has some CT's that go around the mains coming in from the utility meter. it monitors how much current is flowing and in what direction. The insight can read that data, and then command the inverter to "sell back to the grid" (the inverter doesn't know if it's the grid or not, it just sells back the amount of current it's told to by the insight). So, you set "-400" or whatever in the insight, and the insight will have the inverter "sell back to the main panel" until it either maxes out, or until there is only 400w coming in from the grid (combined, the legs may not be balanced.. so you might have 50w on 1 leg, and 350w on the other coming in))

Essentially, if you don't have a NEM agreement, it allows you to have a critical loads panel with the absolute essentials, but also offset the electrical consumption used by the non-critical loads while the utility is up, maximizing your self-consumption.
 
TY. I need to
No, the wattnode just has some CT's that go around the mains coming in from the utility meter. it monitors how much current is flowing and in what direction. The insight can read that data, and then command the inverter to "sell back to the grid" (the inverter doesn't know if it's the grid or not, it just sells back the amount of current it's told to by the insight). So, you set "-400" or whatever in the insight, and the insight will have the inverter "sell back to the main panel" until it either maxes out, or until there is only 400w coming in from the grid (combined, the legs may not be balanced.. so you might have 50w on 1 leg, and 350w on the other coming in))

Essentially, if you don't have a NEM agreement, it allows you to have a critical loads panel with the absolute essentials, but also offset the electrical consumption used by the non-critical loads while the utility is up, maximizing your self-consumption.
Think I got it..TY
 
Something odd happened between 4 pm and 5 pm. It looks like it was flipping between inverting and charging. Do you know what setting made it do that? It did export over 500 watt hours, but it imported almost as much, so the hour total is not too bad. The inverter mode went as high as 2,600 watts, but it also charged at 1,600 watts. It's very odd to see it swing up and back like that in such a short time. Can yu zoom in the graph to just show that onw hour? Here is a 2 hour slice of mine today as the sun was going down into the trees, and we used the microwave etc.

ChargeEnd12-16-23.JPG
At cheap rate time, I try to keep mine charging. The times it dipped to 7 amps of charge current, I was probably using a little grid power, rather than stopping the charge. Even as the XW charge current is going flat, the DC system is still feeding charge current. The Enphase system, feeding to the XW is also powering the house, so the XW stops charging a while before the DC system stops. In fact, After 3:30 pm, the XW will switch to inverting to help power the house because the Enphase system is no longer making enough to run the house. But the DC system was still feeding a few amps to the battery system. So in effect, the DC panels were going straight to AC. It is now 4:42 pm as I type this, and the DC panels are still pushing 14 watts 0.2 amps to the battery, but the XW is inverting, pulling 13.5 amps, about 800 watts from the battery to run the house. So yes, it is discharging, but a tiny bit slower than if I didn't have the DC panels. The Enphase inverters have all shut down for the night. Last production from them was over 20 minutes ago.
 
Something odd happened between 4 pm and 5 pm. It looks like it was flipping between inverting and charging. Do you know what setting made it do that? It did export over 500 watt hours, but it imported almost as much, so the hour total is not too bad. The inverter mode went as high as 2,600 watts, but it also charged at 1,600 watts. It's very odd to see it swing up and back like that in such a short time. Can yu zoom in the graph to just show that onw hour?
Here is the one hour zoom. I am not sure what occurred but once I put it back into standby and back, it worked as normal. I have a charge block from 6 a.m to 8 p.m. so not sure why it started charging.
1702823803816.png

Here is a 2 hour slice of mine today as the sun was going down into the trees, and we used the microwave etc.

View attachment 183426
At cheap rate time, I try to keep mine charging. The times it dipped to 7 amps of charge current, I was probably using a little grid power, rather than stopping the charge. Even as the XW charge current is going flat, the DC system is still feeding charge current. The Enphase system, feeding to the XW is also powering the house, so the XW stops charging a while before the DC system stops. In fact, After 3:30 pm, the XW will switch to inverting to help power the house because the Enphase system is no longer making enough to run the house. But the DC system was still feeding a few amps to the battery system. So in effect, the DC panels were going straight to AC. It is now 4:42 pm as I type this, and the DC panels are still pushing 14 watts 0.2 amps to the battery, but the XW is inverting, pulling 13.5 amps, about 800 watts from the battery to run the house. So yes, it is discharging, but a tiny bit slower than if I didn't have the DC panels. The Enphase inverters have all shut down for the night. Last production from them was over 20 minutes ago.
Neat! Solar is next but my reserves have to get charged first :)
 
That is very odd. It looks like the system went into some kind of oscillation. My guess is the Watt-Node was over correcting, but too late. It saw grid export, so it reduced the inverter power. But then it saw too much grid power coming in, so it increased the inverter power. It went to far, and was exporting again, so it reduced the inverter power yet again, and too far, again. Rinse and repeat.

When I programmed my PLC system, I intentionally made it only correct 80% of the error. It may take it a few updates to hit the right current, but it keeps it from over shooting. I am not using the built in Zero Sell functions.

Neat! Solar is next but my reserves have to get charged first :)
Since you have the XW-Pro, I have to recommend adding DC coupled solar first. Then the sun just charges your battery ant time it shines on the panels. No messing around.

I started with the Enphase microinverters first, then added the XW for storage, and it was a headache. The only reason it works as it does is the PLC logic controller that I programmed to command the XW based on the power flow in my system. When the controller sees the AC coupled solar making more power than I am using, it sends the command to tell the XW to charge. Then it measures the grid export and adjusts the charge current to keep my grid export at the value I set. The Schneider programming can't do that.

With enough DC solar coming in, the XW stays in the invert mode all the time. IT runs the loads in the backup panel on it's AC Out all the time. You can limit it by using the grid support mode, and with the Watt-Node, it can also grid sell just enough to run the main panel loads as well. If I ever build a new system, I will use mostly DC coupled solar to the battery.
 
Back
Top