Adding storage to my Enphase system

[GXMnow]

Does anyone on here have a Schneider Conext Gateway? And if you do, it is connected WiFi to your network? I spent about an hour with the gateway, and I have it updated to the latest 1.08 software. I think I have the WiFi configured correct, I picked my network, entered the WiFi password, and it lit up a nice green check mark. I have it set for DHCP, but I can't see it in my router log. I have tried several things, even having it full reboot without the ethernet cable, and it still is not asking the router for an IP address. As soon as I plug the cable back in, it comes up and works fine, but I was hoping not to have to run an ethernet cable out to the far end of the garage. So I have not even added the XW-Pro to my plant yet in the gateway, and have not even seen the setting in the inverter yet. When I am a bit more awake, I will just take my laptop in the garage so I can get the Gateway connected to the XW-Pro and at least configure it for my battery bank before I let it charge at all. I need to also go pickup a couple more conduit fittings and #8 wire to tie it in to my breaker panels.

With just the battery bank, it lit up in standby mode. I hit the power button and it turned on and was making the 120/240 split phase power (it measured 118/236, close enough) and pulled just over 1 amp from the battery in run mode. When I went to shut it down, a tap on the power button did nothing, so I then held it in and after a few seconds, it did shut off. But as I went to turn off the battery disconnect, it powered back up again. It did not want to stay off or go into standby. I am sure that has to be a setting. I finally held the button in and when the BMS showed zero current and the face of the XW was dark, I shut off the battery bank disconnect. A few seconds later, the XW still tried to power back up, but the caps only lit it up for about 2 seconds as the fans cycled and it went dark.

After some breakfast, I'll be back at it today. I guess you could say there is one good thing about this Covid-19 lock out. I am working from home and as long as I keep my reports getting in on time and I monitor my phone and e-mail and deal with customers, I am here to work on this. But I would so much rather be back at work. We all really need to wear our masks and do what we can to slow the spread.

 

[GXMnow]

I got the Conext Gateway on my WiFi network.

I had to read the setup in the online manual again. I feel stupid, I skimmed one line, since I was talking to the box on an ethernet cable, I figured it was all good as I saw all of the config info, BUT....

The ethernet cable must not be connected "as you select the WiFi to connect to and enter the password"
What?, how?, I have setup many little internet boxes, and this is a new one. You must be connected to the gateway using it's WiFi access point. Then pick your network, and it then get's the DHCP IP address. And it then showed me the network status. Now disconnect from the internal access point, and now I can see it on my LAN at the reported IP, YAY!

Going to plug it into the Xanbus on the inverter and start setting some basic parameters.

 

[GXMnow]

That's annoying

I went into my router and reserved the address for the MAC address, and the device vanished from the network. Never had that happen before. It is one of the features I really like in this ASUS router. It is still DHCP as far as a device is concerned, but whenever the same mac address asks for an address, it will give it the address you select. It is like setting a fixed IP, but it is all managed in the router. The Conext Gateway does not like it. And when I went through the whole mess again, the Gateway changed it's MAC address. So the router sees it as a new device, and gave it a new IP address, WHAT!

So I finally have the gateway on my network via WiFi, and I can see the XW-Pro in my device list on the Xanbus. Now to look through the settings and get this thing working with my battery bank.

 

[GXMnow]

With my battery specs entered into the inverter, and upgrading the inverter to 1.03 firmware, and then having to re-enter all my battery specs...

I temp connected the grid side with a 20 amp breaker. It only took a few seconds to "qualify" the power as ok. I had it set for 3000 watts max charge power, so of course, it started sucking 3000 watts and stuffing it into the battery bank. That is about 56 amps, which is totally fine for the batteries, but I did not want to pull that much now. The solar is only making 2700 watts in this heat. Since the solar is still in the main panel, it thinks it is grid charging only and going by the max settings. While it is live, I was able to go into the charger setting, dropped from 40% to 20% and now it is charging at 1700 watts, pulling just over 7 amps at 240 volts from the line and bulk charging the batteries.

I set the "Bulk Termination Voltage" to the maximum 54 volts it would allow. That is well below the actual bulk charge of these cells. But it then has an "Absorb Voltage Set Point" which I set at 57.6 volts which is just 4.11 volts per cell. I may back this down a little more. The pack is still just room temp (nearly 100F already in the garage) and 28 amps is nothing to these cells. The status still says it is in Bulk Charge even though the voltage is at 56.1, 2 volts above the 54 bulk termination setting. They need to work on their UI to make that less confusing. I had to set it to "Custom" battery type, as even in "Lithium" it did not give me any setting for the voltage limits and did not even report what it was set to. If you have a "Default" it really needs to let you know what it is. My guess would be it is expecting a 16S LiFePo4 pack, but I really don't know.

I have it set to go to charge block at 3PM, and then Grid Sell from 4 pm to 9 pm, but at just 7 amps. That would be exporting 8.4 KwH out of the inverter, but my home will be consuming more like 11 KwH at that time since my A/C will be on almost constant, so I will actually still be buying a little power. With no WattNode, it can't self adjust the sell current.

One thing I have not been able to tell yet, and it is not clear in the online info... If I have it "selling" 7 amps out the AC1 grid connection to feed my main panel, Will it supply say 10 amps on the "essential loads" and still be pushing 7 amps out the grid input? If that is the case, it means I would actually be pulling 4,000 watts or 74 amps from the battery bank. Right now there is nothing on the load output side. On the status screen, it is passing through the AC1 input to the output. It shows about 7 amps coming in and no current going out.

Obviously having my grid tie solar still on the main panel is not giving a true picture of how it works. I need to still get a few fittings and the #6 and #8 wire to tie it in correctly, and I don't want to take down the solar or A/C while it is 102F out.

 

[GXMnow]

Sorry, no new pics, I am not proud of the temp wiring, I have not had a chance to get the other fittings yet to do it right. I have been spending most of my time going through the config.

I found the first thing I truly HATE about the XW-Pro. There is a light 60Hz buzz that I can ignore, but there is also a high frequency that I CAN'T!
I can't hear it in the house, but when I am in the garage, while it was charging at 20 amps, the high frequency noise is obnoxious. I have not checked what frequency it is at, but it has to be up close to 18,000 Hz or so. Most of my work now is as a sound engineer working on very high end professional systems, and this noise is annoying is really annoying to me. In it's basic idle state, inverting, but not putting out current, I can still hear it, but it is quite a bit lower. At 4PM it should start exporting power to my home, I am curious if that makes more noise. If I have to work in the garage for a decent amount of time, I can just disable it if it starts bothering me too much. I'll ask Schneider if it is normal, and see if maybe they can up the switching rate in software to get it over 21,000 hz where I know I can't hear it. My girlfriend hears it too, but does not think it is bad. My son has a damaged eardrum, so he is only hearing it in one ear and though it was his ear ringing, as he has mild tinitus. So it is likely most people would not complain about this.

On the inverter settings, it has a high voltage cutout setting. This is separate from the charging settings. I just set it to 58.8 volts which is 100% full 4.2 per cell which my pack should never hit, but in an hour of charging, it store an over volt warning 3 times, even though the battery voltage graph was just a steady climb and was still under 57.4 volts at the time. When I looked u the cause of the fault, it said to set the cutout voltage to 2 volts higher than the battery should ever go. Since I upped this to 60 volts it has not set that error again. The first thing I thought was my BMS went open while it was charging, as this will cause the error, but the charging side of the XW-Pro just stayed charging the whole time with no error, and the BMS is reporting no error. So I guess the threshold on the inverter side is just too picky, so that 2 volt margin is needed. I know I am running way low at 900 watts of charging, but the current is not as steady as I would have expected. I can see it fluctuate about 10%-15% on my BMS as well as in the Conext Gateway web monitoring app. I am only looking at it on my local network, it is not up on "Insight 2" yet. I am not going to register the system until the essential loads panel is wired up.

Since this will be the first test of power export, I have dialed it down to just 5 amps. About 16 more minutes until we go to peak rate and see what it does.

 

[GXMnow]

SUCCESS!!

At 4PM the inverter started to push power back to the main panel, and it ramped up smoothly. It was very close to the set 5 amps. I then upped it to 7 and finally to 8 amps where it is now. The sun is still shining pretty good and my solar is still making about 2400 watts, and the XW is pushing another 1700 watts. So there is almost 4100 watts going back into the main panel, this is just a hair more than my 16 amp export limit, but since my A/C is running in 107F heat, I am not exporting at all. My house is using all 4100 watts at the moment. The solar is starting to ramp down pretty quick now as the sun angle is getting low in the sky.

At 1700 watts out, it is pulling 31 amps from the battery bank at 57 volts = 1767 watts in, give or take some rounding. So the rough numbers agree with the rated 96% efficiency. I am very happy with that. I tried to get a screen grab on my phone, and the current reported by the BMS is fluctuating from 29 to 38 amps. Not quickly, but it changes a little every second or so. I think the XW is hunting and the sample times between the XW and the BMS make it look worse. The pic I got shows 35.4 amps at 56.27 volts after exporting 700 watt hours. That would suggest 1991 watts, which would make the efficiency look a bit worse.

The BMS was at 98% state of charge when it switched over, After it pumps out for an hour or so, I will see how well the state of charge loss matches the XW Export log. It is down to 94% when I snapped the pic. The batteries are staying cool, they actually feel cooler than the garage floor. The BMS though is pretty warm at 35 amps. Even with 100F ambient air, I can keep my hand on it, it is certainly not dangerous, but it is rated for 200 amps. This is only 35 amps. The BMS App says the FET's are up to 45C, while the temp sensor on the battery pack is still at 36C. Air temp is near 38C. I also have the Schneider battery temp sensor on top of the battery with a small wrench on top holding the metal tab against the cell case. That sensor is reporting 95.72F or 35.4C so all right in the ballpark. For a quick check, I measured the voltage drop through the BMS. It was just 39 milli volts. Doing a quick Ohm's law calculation, 0.039 volts / 35 amps = 0.0011 ohms. The spec sheet says 1.5 milli ohm, so it is better than the spec sheet. That heat coming off the BMS is only 1.36 watts, so it is not bad at all, but with how hot the air is around it, it can't get ri of the heat. I will add a PC CPU cooler to the BMS board to help pull some heat out. It is does get to 75C it will cut th current output to protect itself, it is not even close.

Here are a few pics. Sorry for the pics of the lap top screen. That PC is not on the home network right now so I couldn't easily take a screen grab.

Charge current started at 40% of the XW's 140 amp max, which I think is a bit much. I then dropped it to 20% then 10%. You can see each set down as the battery voltage was climbing. Once it hit the charge level, it turned off charge. I reset the inverter and it went back to bulk charge for a bit. Then you can see it go to discharge as it started to export. First exporting 5 amps, then 7, and finally 8 amps at 240 volts, or about 1900 watts.
Here it shows the power going from the batteries to the grid. It only shows KW, so 1.9 looks very small.
The power meter still shows I am still buying 1.08 KW from the grid, not actually exporting
The BMS shows the power going out of the battery bank at 56.27 volts and 35.4 amps. This is a little more than the average I am seeing on my Fluke meter. You can also see the cell balance voltage is all within 0.006 volts, and the FET temp at 45C, battery bank at 36C.
This is the solar production today. I grabbed this about 40 minutes after the pics of the Schneider operation. As the sun is falling, it is now down to 1.9 KW at 5PM. The Schneider has been exporting for an hour. It says I have exported 2.1 KwH's and the BMS is down to 87% capacity showing 313 amp hours left. This started at 98% so it has used 11% to put out 2.1 KwH's. Doing some quick math, that short test suggests a battery capacity of 19 KwH's, a little more than the LG spec sheet. The Amp Hour gauge went from 358 down to 313, so used 45 amp hours to make 2.1 KwH. Just divide and we get an average voltage of 46 volts, which is a bit low. Hmmm. There is an efficiency loss here. The real pack voltage went from 57.6 down to 55.2 under load so lets take the middle of that as 56.4 volts x 45 amp hour = 2.538 KwH was taken from the battery. The numbers keep moving, but this should be close. This looks like a worst case 83% efficient, but then it means my pack is 22 KwH's so it is not that low, I think we have some rounding errors

 

[GXMnow]

The XW-Pro stopped exporting power after 4.9 KwH's went out, but it was only 7 PM, I had the export time set to 9 PM. Hmm. I checked and the battery voltage was at exactly 53 volts. Look in the advanced grid support, and sure enough, grid support voltage 53 volts. As soon as it dipped beow, it stopped. At that point, I checked the BMS and it says I have used 27% of the capacity. I know it is far from perfect with some rounding etc., but if I take 4.9 and divide by 0.27 (27%) it still comes back that my battery bank is 18.148 KwH's total. Almost the same at the first 11% figure. So it is holding. Doing it by Amp hour, It estimates I have 257.4 left, started at 358, so used 100.6 at about 54 volts = 5.4 KwH. If that is ONLY 27%, it puts my full capacity at closer to 20 KwH's. In any case, the battery is out performing it's specs.

 

[400bird]

Nice results! Great first day. Once your fully wired in with full sized breakers, you should run those cells full to empty to capacity test them.

It's funny how similar our systems could be. AC coupled, XW Pro, I was going to get a complete used bolt pack and use a portion of it. Initially two 14s packs.

I'm excited for your results. Hurry up and turn down the thermostat so you can get to work. Lol
I want answers now! Haha

 

[GXMnow]

Too bad I don't have any A/C in the garage. It's down to 86F outside now with the sun down, so I will do a bit of clean up work, but I never made it out to get the parts I need. I made up a good shopping list for tomorrow. I am a bit nervous about leaving 2,000 watts flowing through temp wiring, so I have been here checking it every 15 minutes. I'll shut it down when I go out.

Those Chevy Bolt packs are amazing. So far I have only pushed up to 60 amps, but the packs never got warm in the least. The voltage did not even dip with a 30 amp change in current. It only dropped with discharge time. If you do end up buying an entire Bolt pack, would you have any interest in selling me the two 8S blocks? I could graft them into a third 14S string and add 50% more capacity here. If I do get them safely apart, I can even send back the extra 6 cells. It is a bit of a pain that the cells are packed in pairs, but wired in parallel groups of 3. The easiest way I found to make the 14S was just cutting the middle buss bar in the 8S and then putting each half in series with a 10S block. So I totally understand if you don't want to give up the 8S ones. There are only two.

 

[400bird]

I'm just having fun, it's scorching hot out!
AC in the garage is a dream, I wish I could pull that off.


Those two 8s packs are the most important. They are key to making a 14s pack, as you know. I'm going to see if I can actually separate the two 4s halves. If I can, it gets easier. I could pull down a 10s into a 4s and 6s.

 

[GXMnow]

Cutting the buss bar between the 3P group was not bad at all, I did it with a Dremel tool with a small rasp bit and a vacuum cleaner sucking up the chips. But being that the 3P cell groups do not line up with the 2 cell holder frames, makes it a pain, to physically split them up.

If you always work from the negative end (on both of my 10S bricks), you should be able so split it at every 3rd cell holder, in 2S3P chunks. But when you get to the positive end it has an oddball cell holder. The 8S brick I got was the other way around. The positive end started the normal cell groups, and the oddball one was at the negative end. If you tried to combine them again, you would have to turn it around to keep the polarity correct. I used the original mounting plate on the back end, but I only had 1 of the plastic end plates for the other end, so I just removed it and have a 1/8 inch thick aluminum rack panel plate on the end. It did bow a little when I bolted it up tight, so I added a washer under each bolt, between the plate and the aluminum cell holder plate. Then I also added another 1/8 inch thick 1 3/4 inch wide bar running vertical across all 3 packs. There is certainly some pressure against the cell face now, but no longer bowing the rack panel. Once they are in the final rack cabinet, I may add another brace to ensure the panels don't bow outward again. From what I understand, these pouch style cells like to have a bit of compression load on them. If I had 2 more of the plastic separator plates, I would have left those between the rack panel and the cell. In the full Bolt pack, they go between the two 10S packs running the width of the car, so they also have a cell pushing on the other side. That area may need to be filled to make sure the plastic does not bow out.

I did not look at the whole battery layout close enough. When both my 10S packs had the same polarity, I figured they all were the same, but then the 8S was flipped. Not sure if both are flipped, or just one? Are some of the 10S flipped? I'll have to look at the battery tear down video again. Just follow the buss bars around the car.

If they ever do come up for sale again at Battery Hookup, I may just get 2 more of any size they get, so I can make another 14S stack and have a few extra cells to mess around with. If I get 2 10S bricks, that is a total of 60 cells. The 14S3P uses 42 cells, then with the extra 18, I could make a 14S e-bike battery at 60 amp hours, and still have 4 more cells for a car jump starter.

 

[GXMnow]

So it seems the Schneider software has a logic error.

After reading a few other posts and such, it just did not seem right, but now after experiencing it myself, I have to agree something is wrong and they need to do a software fix.

According to the videos online, I have the settings correct. They may not be perfectly optimized, but it should work.

I have the charger set to charge just 17% (of the 140 amp max) so it is running at a bit over 23 amps. When my A/C is not running, I am still exporting a bit of power, probably up to 500 watts right now, but when the A/C comes on, I have to buy at about a 2KW rate. Since I only have it connected on the input, it has no idea about the solar production, so the current is just manually set right now.

At 3:50 PM, I start "Charge Block" so it will stop, that part works fine.

At 4 PM it will come out of "Sell Block" and if the battery is above 55 volts, it will start exporting power to the AC1 input which goes to my main panel. I have it set to push just 7.5 amps at 240 volts. Just 1,800 watts. This is set to run for up to 5 hours, with "Sell Block" starting again at 9PM. So far so good.

At 55 volts, my battery bank still has 3/4 of it's power left. But as soon as the voltage falls back below 55, it stops inverting and pushing power to my panel. So I set this down to 49 volts which is still 3.5 volts per cell, so totally fine. As expected, it fired back up and continued pushing power. At the 9 pm mark, it had pushed 8.7 KwH from the battery into my house. All seems good.

I did not want to use grid power overnight yet, as my panels should make enough, so I did not have it come out of "Charge Block" until 6 am. Still no real sun yet, but figured it can start and run at a lower current to put back the 8.7 KwH's. When I checked the system at about 7 am, it still was not charging. Hmmm. Go in and check the settings. Ah, the "Recharge volts" was set at 50 or so, and the battery was still at about 50.5 volts. I raised this up and it still would not charge. As I understand it, this should be a level where it forces a charge to start. So still nothing. Hmmm. The change that finally made it start charging was raising the "Grid Support Volts" to above what the battery was at. Even though it was in "Sell Block" and not in "Load Shave" this setting still makes it not able to start a charge. I put it back up to 55 volts, and it started charging, but the battery is about to go over 55, so let's see if it keeps charging. Yup, it is up to 55.48 volts now, and still charging. I'll give it a bit more and see what it does. I sure don't want to have to manually kick start it each morning though.

I sent a very long detailed explanation of what is going on to Schneider. I hope their engineers understand what it's doing.

When it is in "Sell Block" and not in Charge Block" it should be able to start charging, even if the battery is at 50% charge.

 

[GXMnow]

Here are two screen grabs off the Schneider just now.

When I first got the charger running, it was at 10%, then I upped it 17% to hopefully get the pack fully charged as I hit the peak rate time charger disable. I am pretty surprised how linear the voltage climb is at this slow 23 amp charge rate. There was a small curve right after I upped the current, but it went flat soon after.

My BMS is only reporting that is is up to about 80% at this point. The charger claims to have pushed over 8 KwH's in already. The battery started at 27% on the BMS, so if we go by those numbers, 80 - 27 = 53%. So if 53% is 8 KwH, then it looks like just under 16 KwH's total. A bit less than the discharging estimate. Obviously not a super accurate, and the voltage is still pretty low at 54.8 (3.91 per cell). So the pack is going to take a lot more to get to 58 volts.

 

[GXMnow]

Up to 4.037 volts per cell, 92% charge on the BMS, with 1.5 hours left until charge block begins and it starts feeding the power back out. It has put 8.9 KwH's into the battery today. The charge voltage graph is still very close to linear. Climbing just under 1 volt per hour. It just topped 57 volts. Since the charge is constant current, it is actually increasing in wattage, now pushing 1,300 watts into the battery. Looking at the AC power in to charge power out, it is 94.5% efficient. Losing 77 watts to heat in the XW. It would actually be more efficient if I charged at a higher current. 23 amps is a trickle to this thing.

 

[GXMnow]

The way Schneider has the charger settings does not agree with how I would expect it to work.


For now I just set the charge rate down to 17%.
The "Recharge Volts" is at 52, but I messed with that this morning when I was trying to make it charge.
2 Stage, should be Bulk and Absorb. I have not seen it go to absorb yet. It ran into the time "Charge Block" yesterday. It may hit today, it is getting close. I raised the "Absorption Voltage Set Point" to 58 volts to push a bit more in the pack, and it is at 57.5 right now. It still shows "Bulk" on the charger status screen. But the "Bulk Termination Voltage" is at 54, and it will not let me set that any higher. The "Bulk/Boost Voltage Set Point" is at 57.6, I will soon see if anything happens at that voltage. I am in a "No Float" mode, so the "Float Voltage being 56.7 probably does not do anything. I have the Temp Comp at 0, I can't find anything that says I should temp compensate the voltages on NMC cells. If anyone know about temp compensation, let me know if I should change this for these cells. It just now hit 57.6 volts, and has not changed the current yet. 9.6 KwH's into the battery bank so far today. 45 minutes until "Charge Block", then 10 more minutes until it starts selling power.

3:26 pm, Just hit 58 volts. Still in Bulk.
It switched to Absorption, and then threw an over voltage fault, still at 58 volts, current went to -0.9 amps. That's odd, Why would Absorption pull current? It should be holding at fixed voltage, and it is not, it has dipped more than 0.1 volt already. The BMS now shows 99% SOC 358 AH in the pack. 4.104 volts per cell with highest to lowest cell within 0.002 volts. And the XW pulling 1 amp from the pack. The XW reports it sent 10.1 KwH into the pack.

 

[GXMnow]

It started exporting right at 4PM as scheduled. Pushing 1.7 KW into my main panel. While the A/C was running, I was still buying 1.18 KW from the grid, and the Enphase solar is still making just under 2KW. A/C just shut off, so I am now exporting 2.35 KW out to the grid, but that won't last long, it is still 95F in the shade outside, so the A/C will be back on shortly.

Yes, it would be nice of the Schneider would go back to charging, or even just rest instead of the power going to the grid. But since I do not have the WattNode box, it does not know now that I am not using the power. Even without the XW running rigth now, the solar alone would be exporting a little. It will take a few more days before I can see the usage data from SCE.

 

[400bird]

Good start. I believe it can output current to cover loads it can see with either the internal current sensors or a watt node type energy meter. Shouldn't be much problem there.

My concern is charging. As you found out, there is room for improvement. Or maybe I should call it a design mishap? Software bug?

My solar will cover loads (so, it could charge the battery) but, I don't think it will quite run the AC in addition to normal loads other than 2-3 hours at peak production. The XW would have to switch back and forth from charge to discharge.

 

[GXMnow]

SCE just posted the hourly usage for yesterday. Here is a quick comparison, but, it is not totally valid.

On Aug 13th, and Aug 18th The A/C running most of the evening, we sucked up 15.96 and 15.55 KwH's on the peak rate 4-9 schedule.

Yesterday, Aug 20th It was even hotter, the A/C never stopped, and my son was working on a car and the 3 hp air compressor was running much of the time as well. We only pulled 9.23 KwH's of peak rate power. The last hour from 8 to 9, was only 0.11 KwH. After my son was done using air tools in the garage. Today's numbers might be better, but my son did weld a pair of exhaust hangers. That won't be a ton of power as the welding current was only live for 10 minutes.

Now the flip side. The 13th and 18th (tried to pick semi average days) Off Peak power was 28.14 KwH's on the 18, as it was hot all night, only dropping into the high 80's, so the A/C was running even in the morning. On the 13th, it was much cooler the night before, the temp got down into the 70's, so the A/C didn't really start cranking until 1 pm. So the off peak power was low at 18.57 KwH's (considereing the heat outside). Yesterday, the off peak power was the highest at 29.19 KwH as I pumped over 8 KwH's into the battery.

 

[GXMnow]

Good start. I believe it can output current to cover loads it can see with either the internal current sensors or a watt node type energy meter. Shouldn't be much problem there.

My concern is charging. As you found out, there is room for improvement. Or maybe I should call it a design mishap? Software bug?

My solar will cover loads (so, it could charge the battery) but, I don't think it will quite run the AC in addition to normal loads other than 2-3 hours at peak production. The XW would have to switch back and forth from charge to discharge.

As I understand the manual, yes, it should cover any loads at the output of the inverter while still "exporting" the set amount to the input AC1 terminals. So when I have it set to export 7.5 amps like it is doing now, and then I add a 5 amp load on the "Essential loads" panel, the inverter should increase it's output to 12.5 amps. I should have some of the panel wired tomorrow. I am hoping to pick up the rest of the wire and conduit fittings tonight, and once the solar is no longer producing, I can shut it off and move the wires over.

The switching into charge mode is certainly odd. I hope it is a setting error, but I doubt it. I think they need to rethink the logic so it can charge from a higher voltage as long as the grid support is in time out.

My solar covers my house with a bit of power to spare, IF the A/C is not running. That is the one issue with living here. My A/C draws about double the average power of the rest of the normal house loads. When the grid is up, I don't expect it to flip up and back, from charge to discharge at all. Just charge when the time block says it should have solar coming in, and discharge when the time block says I am in peak rate time. Now if we can add power control based, It would be cool to ramp up and down to match the house load. It looks like the WattNode will do that. While in charging, right now it can't, it will just use the set charge current.

I have not gotten any direct contact with Schneider yet, but these are the things I have in my draft for them to work on.

Improve the time of use by making sure it can be pushed into charge at a set time, no matter what the voltage is as long as it is not at the Absorption voltage.

Maybe add a time/current chart, so if it can't measure the current from the solar, at least we could put in, 10 amps from 8 am to 10 am, 20 amps to 11 am, 30 amps to 1 pm, 20 amps to 3 pm, then stop at 4 pm.

Add a separate charger settings page when the grid is out. If I am off grid, I want it to charge as hard as it can, use all solar that is not going to my loads, do not curtail output unless it is over the max charge current or the batteries are full. Not the low current I am pulling while the grid is up.

With a WattNode measuring the current at the grid input to the main panel, it has all the info it needs. Right now, it can adjust its inverter output to try and zero out the power going to the grid. If the solar is AC coupled, it should also have no issue adjusting it's output to still hold the grid at zero during the peak rate time, using some battery if needed. It does not matter if the current is also coming from the AC coupled solar. But if the solar is making more than the house is using, it will just go down to zero output during "Grid Support", but if it is in the charging time, it should be able to work in reverse. Just sample once per minute. If the house is using power, stop charging the battery. If the solar is pushing power to the grid, raise the charge current until grid export stops. Simple math, does not need to be fast. Have the setting in time blocks. For my system I would have it in charger limiting mode from 6 am to 4 pm. Export limit mode from 4 pm to 9pm, no charge until 6 am again. Seems very simple.

If Schneider can't do this, I may have to program a small microcontroller to send the commands and adjust the XW-Pro setting based on a timer and the reading from a WattNode externally. I am going to try and find a paper for the commands over the xanbus. When the Gateway tells it a new current setting it responds right away. Maybe I can fake it out and send my own commands on the fly?

 

[GXMnow]

At 1.7 KW going out of the inverter, the battery voltage is dropping 1.66 volts per hour. Starting at 58 volts, it will reach the shutdown at 50 volts in 4.8 hours. My Peak rate time is for 5 hours. Maybe I will drop the current I export just a little bit. The charge and discharge is so linear, I did not expect that. I can see my state of charge off voltage very well by this. These LG Chem Chevy Bolt cells are so well matched that the balancer has not had to come on at all. At the worst case, I saw the highest to lowest cell delta just hit 0.006 volts for about a minute. I have the balancer set to start moving energy when the delta exceeds 0.008 volts. I am 2 hours and 20 minutes into discharging. The pack had "topped out" but at just 58 volts on Bulk, and no Absorb at all. So probably a true 80-85% charge. The BMS is reporting pack at 77% capacity, at 279 amp hours remaining, average cell voltage at 3.840 volts. It started at 99% and 358 AH. so used 22% and 79 AH at an average voltage of 55.65, so it used 4,396 watt hours to output 4.2 KwH's so far. And the BMS did calculate off the current used, as 22% of 360 is 79. The cells seem to be living u the the ratings. The tag from the Chevy Bolt under rates them a bit at 5.94 KwH for the 10S brick. That holds 30 of the cells. By that number, my 14S and 6P (84 cells) setup should be 16.63 KwH's. At the nominal 51.8 volts, that would be just 321 amp hours. They are certainly measuring closer to the cell package rating of 60 per cell, or 360 amp hours for 6P.