Trying to figure out self supply on a XWPRO and if it is possible with a pair of inverters.

[Joe_]

I think that's great news!
Can you try inserting that image again? It doesn't show up for me.

My apologies I only saw your post just now. I'm not sure how to fix not being notified exactly. I did just enable push notifications, I hope that fixes the problem.


I fixed my post above so you could see what Schneider tech support said.

I have some questions for you on the virtues of going with a Schneider mppt or staying with my EG4 100s to charge my batteries.
I was following a recent post where n2aws was explaining to someone who was using a non sol-ark mppt to charge his batteries in addition to the sol-ark mppts and it wasn't working well for him due to his configuration of the sol-ark AIO.

The issue for him seemed to be the Sol-ark was unwilling to export the power to the grid that was available from the array from this non Sol-Ark mppt controlled array with his setting because the sol-ark was unaware that the power from this additional array was available to export due to the non Sol-ark mppt which was controlling the array not being able to communicate directly to the Sol-Arc mppt .
Here is the post : https://diysolarforum.com/threads/sol-ark-question-re-additional-ac-coupled-array.74059/post-952051
in case I have misunderstood what the conversation was about.

(I hope that link works better than those pictures I tried to post in that email did )

I think I understand the point N2aws is making, or at least I think I do.

Will I have a similar issue if I don't use a Schneider mppt when implementing supply with the WatNode and context gateway, but a non Schneider MPPT? What do you think? Anyone else that reads this who understands these Shneider units please feel free to set me straight. I'm having some difficulty understanding how these parts of the system interact.

 

[zanydroid]

Yes that is what the convo was about. Glad somebody got value out of the sideways heading thread.

My best guess is that the XW will need to be configured to export within a SoC or voltage band to force the non-Schneider MPPT to generate power. I’m not sure how efficient it can be made, WRT not leaving production on the floor. It can probably be made to work OK in the sense of not wasting energy but likely not as optimal WRT managing SoC and cycling battery as a native MPPT.

 

[Joe_]

If I am understanding, I will have to discharge my batteries to some point, I have no idea what point, lets say for the sake of the conversation 80% just for a number. That to stimulate production from the off brand MPPT? Am I missing something else also? I think that is what you are saying? If so will that have enough effect on battery life to matter.
Is that the real issue or is it likely to be other things that are problematic ? and what is WRT?

 

[400bird]

If I am understanding, I will have to discharge my batteries to some point, I have no idea what point, lets say for the sake of the conversation 80% just for a number. That to stimulate production from the off brand MPPT?

The charge controller needs somewhere to put the energy. Either into the battery, a DC load, or your inverter needs to turn it into AC. Once the battery is full, there's no where else for your charge controller to put energy.
This comes down to settings. This is what n2aws, myself and many others were getting at in that post.

Yes that is what the convo was about. Glad somebody got value out of the sideways heading thread.

I read that thread and thought about the poor SolArk tech support people and how frustrating it must be to try and explain how it works when the customer just doesn't understand what you're telling them.

My best guess is that the XW will need to be configured to export within a SoC or voltage band to force the non-Schneider MPPT to generate power.

But, this is a non-export (to the grid) deployment.

My apologies I only saw your post just now. I'm not sure how to fix not being notified exactly. I did just enable push notifications, I hope that fixes the problem.

Email notifications are turned off by default. The forum owner was facing $thousands monthly in charges that basically trace back to members marking forum emails as spam.

I fixed my post above so you could see what Schneider tech support said.

Looks like good news!

I have some questions for you on the virtues of going with a Schneider mppt or staying with my EG4 100s to charge my batteries.
I was following a recent post where n2aws was explaining to someone who was using a non sol-ark mppt to charge his batteries in addition to the sol-ark mppts and it wasn't working well for him due to his configuration of the sol-ark AIO.

I have good news, @n2aws is a Schneider owner. They've got plenty of knowledge on Schneider gear!

The issue for him seemed to be the Sol-ark was unwilling to export the power to the grid

Which, you can't do, right?

that was available from the array from this non Sol-Ark mppt controlled array with his setting because the sol-ark was unaware that the power from this additional array was available to export due to the non Sol-ark mppt which was controlling the array not being able to communicate directly to the Sol-Arc mppt .
Here is the post : https://diysolarforum.com/threads/sol-ark-question-re-additional-ac-coupled-array.74059/post-952051
in case I have misunderstood what the conversation was about.

Close enough, sure. I think you're picking up what we're putting down.

Will I have a similar issue if I don't use a Schneider mppt when implementing supply with the WatNode and context gateway, but a non Schneider MPPT? What do you think?

I run an XW and a midnight charge controller. I don't export any power from the charge controller directly to the grid. This requires a Schneider brad charge controller and "enhanced grid support" but you don't want that as you can't export to the grid.

I would still recommend going with the Schneider charge controller, just for the benefit of having all monitoring, settings, and support connected.
But, I don't see any operational benefit for you.

The charge controller will charge through battery. The inverter will use DC energy as needed.
Once the battery is full, you don't have anywhere else to put it, because you can't sell it to the grid.

Running a watt node and giving the XW the ability to cover loads in the main panel should be a separate set up entirely. This is self supply.

I don't see any connection between these two (enhanced grid support and self supply) as far as hardware or settings.

 

[Joe_]

The charge controller needs somewhere to put the energy. Either into the battery, a DC load, or your inverter needs to turn it into AC. Once the battery is full, there's no where else for your charge controller to put energy.
This comes down to settings. This is what n2aws, myself and many others were getting at in that post.


I read that thread and thought about the poor SolArk tech support people and how frustrating it must be to try and explain how it works when the customer just doesn't understand what you're telling them.

But, this is a non-export (to the grid) deployment.


Email notifications are turned off by default. The forum owner was facing $thousands monthly in charges that basically trace back to members marking forum emails as spam.


Looks like good news!

I have good news, @n2aws is a Schneider owner. They've got plenty of knowledge on Schneider gear!

Which, you can't do, right?

Close enough, sure. I think you're picking up what we're putting down.


I run an XW and a midnight charge controller. I don't export any power from the charge controller directly to the grid. This requires a Schneider brad charge controller and "enhanced grid support" but you don't want that as you can't export to the grid.

I would still recommend going with the Schneider charge controller, just for the benefit of having all monitoring, settings, and support connected.
But, I don't see any operational benefit for you.

The charge controller will charge through battery. The inverter will use DC energy as needed.
Once the battery is full, you don't have anywhere else to put it, because you can't sell it to the grid.

Running a watt node and giving the XW the ability to cover loads in the main panel should be a separate set up entirely. This is self supply.

I don't see any connection between these two (enhanced grid support and self supply) as far as hardware or settings.

Does that mean that if my settings are correct, when my batteries reach my specified SOC and there is a demand from the WattNode, the inverter will then begin to shove power out AC1 in accordance to the WatNode's demand to cover a percentage of what would normally be coming in from my grid connection? Do I have that correct?
As long as the batteries then don't droop below the specified SOC and there is still demand from the Watnode then the power will continue tapering or increasing using my SOC value and demand from the wattNode to regulate the back feed to the main panel through AC1. That sound correct?

 

[Joe_]

Does that mean that if my settings are correct, when my batteries reach my specified SOC and there is a demand from the WattNode, the inverter will then begin to shove power out AC1 in accordance to the WatNode's demand to cover a percentage of what would normally be coming in from my grid connection? Do I have that correct?
As long as the batteries then don't droop below the specified SOC and there is still demand from the Watnode then the power will continue tapering or increasing using my SOC value and demand from the wattNode to regulate the back feed to the main panel through AC1. Is that close?

 

[GXMnow]

Does that mean that if my settings are correct, when my batteries reach my specified SOC and there is a demand from the WattNode, the inverter will then begin to shove power out AC1 in accordance to the WatNode's demand to cover a percentage of what would normally be coming in from my grid connection? Do I have that correct?
As long as the batteries then don't droop below the specified SOC and there is still demand from the Watnode then the power will continue tapering or increasing using my SOC value and demand from the wattNode to regulate the back feed to the main panel through AC1. That sound correct?

Yes, I think you are getting a good understanding here.

Look at the XW-Pro as a completely independent device.
The "Grid support" and "Grid Sell" functions can be set to operate based on battery voltage (or state of charge, if you have a BMS or Battery Monitor that is supported). In my case, I am just using voltage control. As long as my batteries are above 50.5 volts, I let it do grid support all day. That will provide power to my backup loads panel any time that panel asks for more than 1 amp. But then I also have "Grid Sell" turn on after 3:30 PM and off again at about 9:30 AM. My external controller works very similar to what the Watt-Node and "Zero Sell" will do for you. While Grid Sell is on, and the battery voltage is high enough, it will "Export" current out the AC1 input port to cover the loads back in the main panel. This does not care what charged the battery, it just works. I have mine set to always export a tiny 20 watts, but you can set it to always import some as well, since you do not want any export.

Now look at the DC MPPT charge controller as an independent device.
The MPPT charge controller will always try to pull as much power as it can from the PV solar panel array. Even my cheapo BougeRV charge controller just keeps trying to pump energy into the batteries. Most of them only work on battery voltage, and that is just fine. In my case, I have it set to push as much current as it can harvest until the battery voltage hits the absorb limit. I set that to 57.6 volts for my 14S Li NMC battery system. For a 16S LFP setup, I would probably back that down to about 55.2 up to about 56.32 volts or so. When the battery voltage get's to the Absorb voltage, the MPPT charge controller actually stops doing MPPT. It switches to a constant voltage output. The idea here is it will try to hold the battery at your set voltage. Depending on the complexity of the controller, it may have an absorb or bulk timer setting, and then it will switch to a float mode. In my controller, I have the float setting set the same, but with LFP cells, you will want it to drop to no more than 3.45 volts per cell, or 55.2 volts if you have it bulk higher. At this point, the Charge controller is still running in constant voltage mode. What this means is it will try to keep the battery voltage from dropping. Let's say the XW-Pro inverter is pushing out 1,200 watts to run your loads. When the sun is high your solar panels might be pushing 4,000 watts into the battery bank. Neither one knows about the other, but it does not matter. When the battery is not full, it will be charging at the different, 2,800 watts. If you turn on a huge load, and some clouds move in, you may now be pulling 5,000 watts, but the MPPT is only able to push 1,000 watts. In this case, the battery is actually discharging, but at only 4,000 watts.

If the charge controller makes enough energy to overcome what the XW is using, and the battery does manage to get to the absorb limit, the current should end up dropping to just match what the XW is using. As the battery discharges, the voltage dips, and the solar charge controller does what it can to keep the voltage at full (or float). Now I did find one serious flaw with my cheap BougeRV charge controller. It has a fair bit of internal resistance between where it is measuring the battery voltage, and the actual battery terminals. This causes the current to begin tapering off over 0.5 volt early. And the internal calibration is also about 0.2 volts off. Every review I have seen finds it shows 0.2 volts higher than the batt terminal voltage. So in the summer, when my system was topping up the batteries, the current would fall and it would take a full hour for the actual battery voltage to get from 56.9 volts up to 57.4 volts with the charge controller showing Absorb mode. A better charge controller, or one that can use remote voltage sensing at the battery terminals will do much better at this end of charge time. In reality, if it tops up the battery, it does not matter. But on days I fall short, I left some energy on the table that I should have been able to harvest. But then on real bad days, like this winter time, it again does not matter as it never get's the voltage high enough and it stays in MPPT until the sun fall and I am still only at 56 or even 54 to 55 volts.

So what is the problem of not having communication between the inverter and charge controller?
If you can't sell to the grid, there is no problem at all.
The whole point of "Advanced Grid Support" is to monitor how much energy is coming in from the DC charge controller, and set the system to sell the excess energy out to the grid.
Why is this a good thing?
The advantage is only there if you get credit for the exported energy. Without this function, the system has to throw away any extra energy after the batteries are full. With the communication, once the battery voltage reaches full, the XW will know how much power the MPPT is making, and it can grid sell the extra that the house does not need.

In my system, I could spend a couple hundred hours working on new code for my PLC to make it sell off the extra power from my DC panels, but it's just not worth it. The days where my DC system goes into float and gives up energy is small, and the total energy lost each day it has happened is really small. When I had some very sunny but cool days, the worst I saw was giving up maybe 4 KWHs. And for that to happen, the XW stopped charging from the Enphase system, and the extra Enphase power was already exporting as much as 3,000 watts. So what I did was set the XW to stop charging at an even lower battery voltage at 56.7 volts. So the XW goes into stand by mode, the Enphase panels power the house and export the excess, and the DC panels push another 3 KWHs into the battery. I could take that even lower in the XW, but when I am not getting enough sun, I need it to charge more and not export. This has turned out to be a good compromise.

 

[n2aws]

@400bird thanks for tagging me, I hadn't seen this thread, and you *know* I love to talk Schneider! lol

So, reading through this thread, I feel there are a few separate but disctinct questions being asked, and at least 1 (very common) misunderstanding about a setting in the Schnieder ecosystem.

Lets start with the misunderstanding, because clearing that up will be a critical point a little bit later in the post.

Enhanced grid support: The way this works is, it *always* prioritizes charging the batteries above all other grid related activities (selling back, peak load shaving, etc). The misunderstanding is that (as seen in this thread), people tend to think it'll charge your batteries during the day, and you can use your batteries overnight. This is inaccurate unless you use some kind of outside control. With Enhanced grid support, your batteries will always be at 100% SOC. When the sun goes down, it will not use the batteries to support loads. The only time it will draw from the batteries, is if the grid were to go down. Then it'll use batteries to support only the critical loads connected to "AC Out".

The primary purpose for Enhanced grid support mode, is to always keep your batteries topped off.. but after they are at 100%.. selling the excess solar energy back to the grid. (Otherwise, the MPPTs would curtail output since there is no place for the energy to go..)

So, with that clarification in mind.. "Enhanced grid support" is one of two primary reasons why you'd want to stick with Schnieder branded MPPTs. I'll get to the second reason in a moment. But, if you don't have a NEM agreement and can't sell energy back (and don't ever intend to), then Schnieder branded MPPTS aren't required because "Enhanced grid support" isn't a useful configuration for your needs.

So, the other reason (which may or may not matter to you) for using Schnieder MPPTS would be for the monitoring aspect. All of your production, and usage information in the insight dashboard. If you can collect that data another way or just don't necessarily care about the data.. there aren't many other reasons for staying within the Schnieder ecosystem for everything. That said, I do like the single dashboard, so I have Schnieder MPPTs LOL


ok, so on to the questions:
- If using a non-Schnieder MPPT, you can still do zero sell. The caveat there is, you'd set your grid support SOC to something less than 100%. As an example, I set mine to 40%. This allows the inverter to "sell back" to the main panel and offset grid import all day and night, until my batteries hit 40% SOC (actually 39%) and then it stops using the batteries to support the load. I vary this between 40-60% and haven't decided yet what the best value for me is. I just like to have enough to support my loads overnight if the power were to go out, because I'm in a rural area prone to having occasional short outages.
- Caveat to this: IF your batteries don't get back to 100% SOC occasionally.. you'll want to manually set your grid support SOC limit to 100% occasionally to let them get a full charge and top balance the cells. Otherwise, over time the cells will naturually become imbalanced and you'll notice a reduced capacity until you get them back in balance. I typically do this on the 1st of each month now using nodered so that it's automated.
- When your batteries are at 100% SOC, the non-branded MPPTs will curtail power to only serve loads.. BUT, so will your schnieder MPPTs since there is nowehere else for the power to go. Without enhanced grid support.. the MPPTs will behave exactly the same making decisions on when to enter bulk/absorb/float based solely on battery voltage. (Even the Schnieder MPPTs are based on voltage, not SOC)
As for the sol-ark thread, this is the issue that poster is running into, and for whatever reason, he's just not grasping the fact that "if you discharge your batteries.. as the voltage drops the MPPT will wake up and do what it does", he's essentially wanting whatever sol-arks equivelant of "Enhanced grid support" is, and not understanding that it's impossible without the inverter and MPPT communicating.. So, he'll *have* to set his inverter to use some of the battery if he wants the non-branded MPPT to stop curtailing power.

In schneider land, it's a simple fix. Set your grid support SOC to something lower than 100%, and USE the battery that you paid for! lol


- Zero export/zero sell with a single inverter works really well. This is how I use it.
- Sadly, I only have 1 inverter, so I have no useful insight on if you can do zero export with multiple units.

- The setting @GXMnow mentioned to always pull a little energy from the grid works really well. One slight clarification though. The 30second "trip time" setting doesn't actually take 30 seconds to adjust to changing loads. In the real world, my XW Pro reacts within a few hundred milliseconds. I rarely see it take a second to react. The "trip time" settings is actually more of an "oh shit" setting. meaning that if it exceeds whatever value you have set for the "Export limit" for 30 seconds or more, it'll trip the relay and disconnect you from the grid to prevent compliance issues with your utility. (The other time setting there is how long it'll wait before reconnecting to the grid after it's tripped). In practical experience, I've never actually seen the trip time be triggered.

- My loads in the main panel are imbalanced. I'm likely never going to take the time to try and balance them.. so, I occasionally see as much as an 8 amp difference between L1 and L2 (though 3a is far more common). On top of that, I'm not sure what my utility would do for "selling" energy on 1 leg, and "buying" energy on the other to ensure that I have zero export. so, I split the difference between 3amp normal and 8amp max imbalance and choose a very conservative "-1200" export limit. This ensures that I'm always drawing at least 1200 watts from the grid, before the inverter will start inverting power from the batteries. This works fine for me, as I don't have enough solar to cover 100% of my loads anyway, AND my local utility imposes a minimum usage fee, so even if I could offset 100% of my usage.. I'd still be paying the utility for power I didn't use.. so, I may as well use it

I'm sure there were more questions asked than I'm missing right now, but I also fee that this post is already way too long and wordy, so I'll stop here. Hope it helps!

 

[MisterSandals]

- If using a non-Schnieder MPPT, you can still do zero sell. The caveat there is, you'd set your grid support SOC to something less than 100%. As an example, I set mine to 40%.

I thought that either a Schneider MPPT or battery monitor were required to provide the SoC. And that without this SoC, it is advised to turn SoC controls off and use the voltage based controls.

I am banging on this with an SW4048 but thought this concept was the same for SoC "everything".



Source:

https://download.schneider-electric.com/files?p_enDocType=User+guide&p_File_Name=990-91227F-01.pdf&p_Doc_Ref=XW-Pro-NA_990-91227

 

[n2aws]

I thought that either a Schneider MPPT or battery monitor were required to provide the SoC. And that without this SoC, it is advised to turn SoC controls off and use the voltage based controls.

Apologies, and good catch. I was assuming a lithium battery with a BMS that talks to the insight. Anywhere I said "SOC" you can also use voltage based settings. The insight platform supports either/or.

But yes, to get SOC, you either need a battery monitor, or a supported BMS-based battery.

*Edit* unfortunately, a Schnieder MPPT alone will not supply SOC. it has to be a schnieder battmon, or a supported BMS

 

[GXMnow]

- The setting @GXMnow mentioned to always pull a little energy from the grid works really well. One slight clarification though. The 30second "trip time" setting doesn't actually take 30 seconds to adjust to changing loads. In the real world, my XW Pro reacts within a few hundred milliseconds. I rarely see it take a second to react. The "trip time" settings is actually more of an "oh shit" setting. meaning that if it exceeds whatever value you have set for the "Export limit" for 30 seconds or more, it'll trip the relay and disconnect you from the grid to prevent compliance issues with your utility. (The other time setting there is how long it'll wait before reconnecting to the grid after it's tripped). In practical experience, I've never actually seen the trip time be triggered.

Thank you for this clarification. I thought the trip time seemed too long for the actual control loop. My PLC based control updates every 5 seconds. Since it has this delay on both increasing and decreasing the sell current, the average still zeros out for even the 1 minute run of the microwave.

I run about 4 amps of imbalance most of the time I am at light loads. Right now I have 1.7 amps coming in on L1 and 3.6 amps going out on L2. The utility power meter only reports the sum. In this case, exporting 1.9 amps. Throw in some power factor, and it's only 49 watts going out.

The sun is managing to break through the clouds a little, and it hit 300 watts more then my backup loads panel a few times. So it keeps going in and out of charge mode. I may have to shut down my PLC in this crappy weather.

 

[Joe_]

I appreciate all the good information and will have some more questions once I reread all the posts a few times and think about their content. A lot of information and insight there.


For information my batteries are (2) EG4 power pro’s. They can each supply 200 amps. And have a capacity of 14.3 kWh nominal. They do support a Schneider RS485 protocol I am hopeful that they will properly communicate SOC to the Insight Home. I think I am OK there?

Reading through the posts the first time pops a complex question to mind.


Originally I bought a Victron BMV-712 Smart shunt to manually monitor SOC. At the time I had planned on using a single inverter and only supporting my backup panel. No self supply to the main panel was planned because I was unaware of that feature.

The Smart Shunt does not communicate with the insight home as far as I have been able to determine, and I haven’t found an interface for it to easily talk to the Insight Home through.
It didn’t seem necessary.


I wasn’t going to install it in an effort to keep the DC cabling simple. Maybe install it later if I found a use for it. But it sounds like it might be useful and worth the extra expense and time to put it in place now?

Choosing to Instal it brings a question about if I have a large enough shunt. More below.

As I understand it when in self supply under the supervision of the WattNode Modbus meter any large induction motor starting surges will be supplied mainly by the grid. If the grid were down they would come from the inverter and the batteries but I lack enough PV or batteries to do that at present.

One of my loads, the one I really hope to offset it’s power usage with self supply the is my 5 T ac. It’s compressor has a running load amps of 26.4 (RLA) and a locked running amps of 134 amps (LRA). This is likely a somewhat conservative figure considering that manufacturers don’t really design equipment to consistently run in desert climates due to how few units actually end up in that environment.
I am aware that I will need a soft start should I try to supply this load without assistance from the grid to start it. But I don’t have a soft start yet and I also don’t have enough Pv or batteries to do so. In this iteration I will not be attempting to start that 5 t off grid. It would be helpful to know for certain that the grid will supply actually supplies the starting surge.

The Smart Shunt says it comes with a 500 amp shunt. There are larger sizes available.
But the shunt can only safely surge to 500 amps I believe? I am fairly certain I ran across spec that said that it was not rated at 500 amps continuous duty though I am unable to find that information at the moment I am writing that. I think it’s continuous rating is about 80%.

Without a soft start which I will eventually add the nominal 134 amp LRA starting surge would be 670 amps at the batteries. With the mandatory soft start installed it will be much less of course. But I am wondering if need the soft start in order to use self supply now, and also if I need to install a 1000 amp shunt initially due to the size of this load and the batteries I am using and having two inverters drawing on the batteries?


Any thoughts?

 

[n2aws]

For information my batteries are (2) EG4 power pro’s. They can each supply 200 amps. And have a capacity of 14.3 kWh nominal. They do support a Schneider RS485 protocol I am hopeful that they will properly communicate SOC to the Insight Home. I think I am OK there?

I don't have the power pros, but the EG4 lifepower4 batteries I have speak to the insight perfectly fine. I suspect the power pros would too. I'm willing to test it if you wanna send me one! lol

As for the rest of the post.. I'll have to defer to @GXMnow on the startup surge. I've never actually thought to test/measure where the surge comes from when on-grid. So, I'll have to take him at his word. Though it completely makes sense. If the A/C is powered from the main panel rather than the critical loads panel, it'd take the wattnode a moment to notice the increased demand and tell the XW to ramp up its output. At the surface, it completely makes sense. But, I've never actually tested it. (Now I'm wondering how a sudden surge would be handled on the critical loads side too.. in both on-grid and off-grid scenarios. I suspect it'd be powered by the batteries either way, but not 100%)

 

[GXMnow]

I'll have to defer to @GXMnow on the startup surge.

In all of my testing, the start surge of anything in the main panel always comes from the grid. My control takes 5 seconds before it starts to ramp up the battery current. The LRA current is long gone by then. My AC is about 3T with a 105 LRA current. Run current is only 14 amps. My single XW-Pro is able to keep it running with the grid current actually exporting 20 watts still. But when it starts, the surge and then 14 amps is on the grid until the XW ramps up.

I do not ever expect to put my A/C compressor on the backup loads panel. We don't have very many or long power outages, so it's not that big of a concern here. But we do have a few decent loads in the backup panel. I can't say for sure as I do not have a second amp probe, but it does lok like a portion of the initial current comes from the grid, but the XW is much faster to increase it's output current to drive loads in the backup panel from the load output. And this does make sense. If it is being told to export 1 amp to the grid input, the instant it sees a 14 amp increase on the load output, it will jump to cover it and get back to exporting to the grid again. I see this when my girlfriend turns on her hair dryer. The draw from the grid is gone in less than 1 second. But that initial 105 amp LRA current surge from the A/C would still show on the AC input from the grid. When running "Grid Tied" the XW is only a current source. It needs to know how much current to add, and the line voltage is controlled by the grid.

When you go off grid, it is a very different beast. The XW is then running in the Voltage Source mode. Cycle by cycle it is monitoring the output voltage and if it dips, it very quickly ramps up the current to keep the voltage at 240 across the legs. I did not run my A/C, but I id run a "3 HP" shop air compressor and a Hobart Handler 150 amp MIG welder on the inverter output while off grid as a test. It easily ran them at the same time with no voltage sag. The LED lights only showed a tiny flicker. I was not monitoring the battery current, but I am sure it exceed the 140 amp constant rating. But my battery bank is fused at 225 amps, and it never blew. Fuses do take some time to trip.

Quite a few people have done some serious testing and the XW pushes out 12,000 watts without complaint. But that is still just 50 amps at 240 volts. The short term surge can certainly go even higher. But I doubt it would like the 105 amp surge from my A/C compressor. LRA is actually "Locked Rotor Amps" which means with the rotor not turning at all. That condition lasts only microseconds. As soon as the rotor begins to turn at all, the current drops quickly. The MicroAir Easy Start claims to be able to drop that peak current to less than 1/3. The trade off is it take twice as long to get the rotor up to full run speed. But we are still talking less than 2 seconds. I looked into getting one, but they were about $600 USD, but now they dropped under $400, so I might do it. That reduced start current is probably good for the life of the compressor as well. I don't like hearing the wires buzz in the conduit when mine starts at full current. SO I feel the Easy Start is not a bad idea, even if I never run the A/C off the inverter only.

 

[Joe_]

I’m out of questions.
I’m fairly certain I understand what I am doing now.
Thank all of you for your patience and excellent answers.?

 

[n2aws]

I’m out of questions.
I’m fairly certain I understand what I am doing now.

Famous last words! We'll see you in a day or two. lol

All joking aside, best of luck. One bit of advice, change one setting at a time after you initially get the system doing "mostly what you want". The Schneider firmware is *not* DIY friendly IMO. and all-to-often, changing one setting without fully understanding the ramifications.. will cause 2 or 3 other things to just "behave differently". Sometimes it's documented, sometimes it's not. so make a single change, and observe it for a period of time. then make your next change.

 

[400bird]

I’m out of questions.

Famous last words! We'll see you in a day or two. lol

Preach it brother!

Joe,
I'll admit, that I thought I had things down and was doing good with my Schneider system. Then I did an update but it "broke" some features that I needed. So, I'd roll back. I fought it on and off for a few months. Updating, trying a few things, giving up and rolling back to the old software.

The entire time, the issue was simply that I needed to reset the clock/time in my Schneider gear. Somehow the new update didn't transfer forward the correct time.

 

[n2aws]

Preach it brother!

Joe,
I'll admit, that I thought I had things down and was doing good with my Schneider system. Then I did an update but it "broke" some features that I needed. So, I'd roll back. I fought it on and off for a few months. Updating, trying a few things, giving up and rolling back to the old software.

The entire time, the issue was simply that I needed to reset the clock/time in my Schneider gear. Somehow the new update didn't transfer forward the correct time.

If I kept track of the number of times I thought "I've finally got it figured out" only to have Schneider humble me.. I could write a book.

The most recent time: I was trying to mock something up for a user on the forum. It required draining my batteries to nearly 0% SOC. Well, at 5%, the batteries alarmed, triggering a bunch of (what I thought were related but weren't) things in the insight and inverter. long story short, I ended up factory resetting *everything*. insight, XWs, charge controllers, etc. After reconfiguring everything, it took me 3 days to figure out why my "peak load shave" and "zero sell" weren't working anymore.. That day (just a few weeks ago), I learned something new.

1) Typos happen. I meant to type -400 for my export limit (zero sell). and typed -4000.
2) I learned that "Peak load shave" even when set to zero amps.. won't actually kick in until the export limit is met, *even though* peak load shave isn't exporting! That was an insanely frustrating couple of days, all caused by typing an extra "0" into a setting. I spent insane amounts of time trying to figure out why it would only "load shave" to my critical loads panel only occasionally. I realized at some point the only time it was load shaving, is when my central AC was running and the house loads were high, compared to my steady state.. thats when I double checked my zero export setting. I'll be honest, if I hadn't observed it worked when my AC was running, I'd likely still be scratching my head on it.

 

[Joe_]

Oh, I didn’t mean it that way. Lol

What I meant and should have said was I am greatly relieved now and can get a decent nights sleep because I can stop worrying if the equipment I bought will work together or not. I have woken up in the middle of the night anxious about that recently on several occasions . Kind of takes the fun out of it.


But I am really excited that it turned out good so far.
There will be plenty of time for it to go sideways in the future I’m certain.? Hooking it all up and configuring it correctly is a project for another day.

I really appreciate all the help from you and the others. It’s going to take a long time for this to come together. At least I have the right parts now. That was exactly what I was hoping for but couldn’t be sure of without everyones help.
Life is good!

 

[n2aws]

I run about 4 amps of imbalance most of the time I am at light loads. Right now I have 1.7 amps coming in on L1 and 3.6 amps going out on L2. The utility power meter only reports the sum. In this case, exporting 1.9 amps. Throw in some power factor, and it's only 49 watts going out.

I was thinking about this a little bit last night.

In your case, since you have a NEM agreement, you have a bidirectional meter.
So, this works fine. However, for many (maybe even most?) non-NEM users, the meter is single directional. It can't really tell if current is flowing in or out. Just that it's flowing. so in your example, if I have 1.7a in, and 3.6a out, my meter would see that all as consumption. meaning, I'd be paying for 5.3a of "consumption". Essentially, I'd be paying them to take my surplus on the imbalanced leg. As a result, we have to be a bit more "strict" with our imbalance.

A while back, I talked to one of the engineers at Schnieder. I mentioned a "flaw" in the wattnode logic, in that with an imbalance, one leg could be selling back because it does the math and views the flow as a "whole". I told them they should have an option that allows the insight to make decisions based on the status of each individual leg. IE, have a setting/GUI button that essentially stops looking at the total, and looks at the status of each individual leg, and if either one is selling back.. ramp the inverter down.

His response was "yeah, but.. how do we make that an option in the GUI that makes sense to users". And I thought about it for a bit, honestly. Then I was like "well, why does the end user need to know *specifically* what it does? I mean, lets be honest.. many of the other features are unclear.. so why not just have a toggle box that just says "NEM agreement" with a little help bubble that says "enable this button if you have a Net metering agreement with your utility provider, and are allowed to sell excess production." Let that be the only config option/indicator for the underlying change.

Under the hood, that toggle option could make zero export a bit smarter. ie, if you set it to always draw 200w from the grid.. with that button disabled, it'd behave as it does now. if it's enabled, it'd become an OR statement. "draw 200w from the grid OR ramp down if one leg is exporting, even if that means draw 500w from the grid for a short time"