Adding Schneider XW Pro

[400bird]

Thanks for the backup, that's what I thought load shave did.

It will be at least a few days before I get back to this. Days is generous, prob a couple weeks in reality.
I tore the battery apart today to do the final assembly and install the second pair of batteries.

 

[GXMnow]

I am still trying to move forward on my control system. And I am seeing the same issue again. When I try to have the PLC do the steps in a sequence, it is hanging up again. I will add a little more delay and see what happens.

I like the Swiss army knife analogy. Maybe we are trying to use the nail file and the cork screw at the same time.

DUH!!!
I figured out what I was dong wrong. I think I am going to get this totally working soon now.

I was having it run my basic language functions on every scan. It was trying to open the TCP connection a few hundred times before it moved on to the next step in the sequence. I needed to select single shot mode, they call it "Differential Up". When the input goes high, it runs the function once. This is looking a lot better now. I now have Grid Watts, Battery Watts, and Load Watts, all being updated about once every 6 seconds.

 

[GXMnow]

Are you just using time to tell it to go into charge mode?

I am thinking about having it monitor the load side watts. When that exceeds -400 watts, then I can start to charge at 5%,.
5% of 140 amps is 7 amps, 7 amps at 52.2 volts is 367.5 watts. That is the lowest charge rate possible. From that point I will monitor the grid watts. On mine right now, that will just be the grid input side of the XW. If that goes to more than 300 watts going out to the grid, then I will add 1 % to the charge power. If it falls to less than 100 watts to the grid, I will subtract 1% from charge power. The loop will happen about every 15 seconds.

At 4 pm, charge will be blocked. My PLC also has real time, So I will have it set the charge power back to 5% at 4pm, and then sit and wait.

Step 2, once the charge side is working, will be to have it adjust the grid export current. Until I have a grid side power meter in my breaker panel, I have a current switch that will tell the PLC when the air conditioner compressor is running. From 4 pm to 9 pm I will have it export just 300 watts when the A/C is off, and then go to exporting 10 amps, 2400 watts, the the A/C is on. After 9 pm, I will have it go to zero export when the A/C is off, like it is now. When the A/C kicks on, I will export maybe 5 amps. I will have to see how much battery I have left after the 4 to 9 time frame. Once I have the DC solar added, and cycle the batteries deeper, then I can zero the A/C further into the night. And I may also have the PLC read a grid power meter to actually adjust the export power, like Schneider does with the WattNode. But I can do it with a much cheaper meter. If I do have enough solar and battery, I could crank out the full power for the A/C compressor. Once it is up to speed, it is only 14 amps at 240.

 

[400bird]

Got it (hadn’t understood the PV was grid tied).

If I had DC coupled solar, the Pi and external energy monitor program would not be needed for battery charging control.
The XW does everything really well when DC coupled.

Are you just using time to tell it to go into charge mode?

Yes, I should go back and address that, but I have not yet.

I am thinking about having it monitor the load side watts. When that exceeds -400 watts, then I can start to charge at 5%,.
5% of 140 amps is 7 amps, 7 amps at 52.2 volts is 367.5 watts. That is the lowest charge rate possible. From that point I will monitor the grid watts. On mine right now, that will just be the grid input side of the XW. If that goes to more than 300 watts going out to the grid, then I will add 1 % to the charge power. If it falls to less than 100 watts to the grid, I will subtract 1% from charge power. The loop will happen about every 15 seconds.

I'd recommend leaving max charge amperage exactly where you want it, as the MAX.

Use EPC charge control, it is in watts (saves you the math to get amps and % charge) and is designed to receive the command every 1-2 seconds and has significantly finer control. You can set charge watts as low as something like 20-60 watts. The control isn't exact, it might be plus or minus a few percent.

Port 502, address 40210 EPC max charge power

At 4 pm, charge will be blocked. My PLC also has real time, So I will have it set the charge power back to 5% at 4pm, and then sit and wait.

Step 2, once the charge side is working, will be to have it adjust the grid export current. Until I have a grid side power meter in my breaker panel, I have a current switch that will tell the PLC when the air conditioner compressor is running. From 4 pm to 9 pm I will have it export just 300 watts when the A/C is off, and then go to exporting 10 amps, 2400 watts, the the A/C is on. After 9 pm, I will have it go to zero export when the A/C is off, like it is now. When the A/C kicks on, I will export maybe 5 amps. I will have to see how much battery I have left after the 4 to 9 time frame. Once I have the DC solar added, and cycle the batteries deeper, then I can zero the A/C further into the night. And I may also have the PLC read a grid power meter to actually adjust the export power, like Schneider does with the WattNode. But I can do it with a much cheaper meter. If I do have enough solar and battery, I could crank out the full power for the A/C compressor. Once it is up to speed, it is only 14 amps at 240.

 

[400bird]

I am still trying to move forward on my control system. And I am seeing the same issue again. When I try to have the PLC do the steps in a sequence, it is hanging up again. I will add a little more delay and see what happens.

I like the Swiss army knife analogy. Maybe we are trying to use the nail file and the cork screw at the same time.

DUH!!!
I figured out what I was dong wrong. I think I am going to get this totally working soon now.

I was having it run my basic language functions on every scan. It was trying to open the TCP connection a few hundred times before it moved on to the next step in the sequence. I needed to select single shot mode, they call it "Differential Up". When the input goes high, it runs the function once. This is looking a lot better now. I now have Grid Watts, Battery Watts, and Load Watts, all being updated about once every 6 seconds.

Congrats!
Nothing like banging your head against a problem for weeks only to figure out it was something you had overlooked or not quite understood.

 

[GXMnow]

I know this is probably like watching paint dry for some people, but I am having fun watching it work. The error was so dumb, but now it makes so much sense. Every time I was adding a custom function, it would ask, "Do you want to run on every scan?" If I said no, it would not add the function. So I said yes, and when I would click on the function manually, it would work. But when I make it run in a loop, it would just go stupid and send tons of crap on the network. The little PLC is actually very fast.

Now that it is working, I had to microwave myself an early lunch. That took power from the solar coming in, so the charge current ramped down from 21% to 6%. Sure, it takes a minute to change 6%, but who cares? When the microwave shut off, it just ramped back up, changing 1% every 10 second, back up until grid export fell below 500 watts again. It is now up to 28% (2,000 watts) charge rate as the sun is coming up more.

I will look into the EPC max charge power, but I think this is going to work just fine. Each step is just 80 watts. And going down to 5% seems low enough. And only changing the setting once every 10 seconds does not need fast control. The refrigerator just kicked on again. Charge current dropped from 28% to 23%. Five 1% steps x 80 watts = 400 wats to run the compressor. If I had to be truly zero export, then I would have to do faster and finer control, but for my situation, I am vey happy. The real test will be tomorrow morning, when it actually goes into bulk charge by itself.

With how nice this is working, I may just bail on the DC panels and add 6 more Enphase micros. No worry about Arc Fault or RSD, it's all just there. The iQ7+ inverters which will take the larger 400 watt 72 cell panels are just $160 each now. x 6 = $960 That is less than the MPPT charge controller and the RSD/Arc fault box. And does not even include the 3 Tigo boxes. I'm torn, the DC charging does have some advantages, bt the main disadvantage of the AC coupling is now gone.

 

[400bird]

I know this is probably like watching paint dry for some people, but I am having fun watching it work.

I have watched that paint dry for way more time that I want to admit.

The error was so dumb, but now it makes so much sense. Every time I was adding a custom function, it would ask, "Do you want to run on every scan?" If I said no, it would not add the function. So I said yes, and when I would click on the function manually, it would work. But when I make it run in a loop, it would just go stupid and send tons of crap on the network. The little PLC is actually very fast.

Glad you got it figured out.

With how nice this is working, I may just bail on the DC panels and add 6 more Enphase micros. No worry about Arc Fault or RSD, it's all just there. The iQ7+ inverters which will take the larger 400 watt 72 cell panels are just $160 each now. x 6 = $960 That is less than the MPPT charge controller and the RSD/Arc fault box. And does not even include the 3 Tigo boxes. I'm torn, the DC charging does have some advantages, bt the main disadvantage of the AC coupling is now gone.

For me, one big pull of the DC coupled panels is black start capability. In a power outage, if you run down the battery overnight, in the morning you will have no way to get the system started.

I too have been considering adding more AC coupled panels. It would allow the car to charge a bit faster, then I could convert that to battery charging once the car is full.

Maybe have it set up so I can swap in a cheapo DC charge controller on a few of the panels, for grid down situations.

 

[400bird]

I've mentioned Ohmconnect before, but they sent me tax paperwork.

Yeah, I set enough juice back in their program over the past year that they paid me a total of $900!

The events happen for an hour a couple times a week over the summer and maybe once a month in the winter.

So, they cut down my investment into this by a not insignificant portion. I don't want to admit to how much I spent on the battery, inverter, wiring, control hardware, battery box, battery heater, etc.
They didn't quite cover 10% of my investment, but it's a dent.

I hate tax time, I had add up all the expenses to write off the battery costs.

 

[400bird]

I removed the 2 packs and tore apart the battery box last weekend expecting to get some time during the week to work on the final build.

Well, work sucks and it was a busy week.

This weekend, I cut and the cement board for 5 sides. Some fire resistant caulking in the corner and 2 coats of paint and I'm where I thought I would be Tuesday or so...

That's the 4/0 cable wrapped in blue painters tape. With how rigid it is, there's no way I'd pull it out just for paint. Everything else came out.

 

[GXMnow]

Have you had your XW-Pro charging above 50 amp yet?

Any charge rate below 50 amps, the current stays solid as the voltage climbs up, but any time the charge current is above 50 amps, it seems to be constant wattage. I don't think it is any kind of a problem. When I first noticed it, I thought it might have been the charger getting warm and ramping down. The odd part is, I think it is limiting to this charge wattage. My PLC data is not being logged, so I can't be sure, but when I got home from work, I saw the PLC commanding 50% charge rate or 70 amps, and the battery was only taking 53 amps an I see the downward slope in the battery summary graph.

I am going to look through the charger settings and see if there was another field that limits maximum charge power. It is not often that my solar is cranking out this much power AND there is no one home and light load on the system. It was really trying to limit export by cranking up the charge rate.

 

[GXMnow]

I think I figured it out.

The EPC charge power limit was set to 3,000 watts. Doing the math, it was limiting to about 2,950 watts. I won't know for sure until we get that kind of solar power coming in again, but I upped it to 4,000 watts to see if it changes the behavior.

While I was at it, I also am allowing it to discharge another 0.5 volts on the bottom, and added 0.28 volts on top. This makes my maximum cell volts now 4.07 volts, and my minimum cell voltage 3.64 volts. With the 4 strings of Bolt cells, that added about 1,270 watt hours to each cycle. Having it charge up the little extra, it pushed 13.2 KWH into the battery bank today. I also s the PLC to drop the charge power to 20% when the battery is near full, instead of the previous 10%. At 20%, it pushed that 610 watt hours into the battery in just 26 minutes.

 

[400bird]

Have you had your XW-Pro charging above 50 amp yet?

My immediate answer was, yes of course and I've never seen that. But wait, have I? Then I reread your post and also arrived home.

Any charge rate below 50 amps, the current stays solid as the voltage climbs up, but any time the charge current is above 50 amps, it seems to be constant wattage. I don't think it is any kind of a problem. When I first noticed it, I thought it might have been the charger getting warm and ramping down. The odd part is, I think it is limiting to this charge wattage. My PLC data is not being logged, so I can't be sure, but when I got home from work, I saw the PLC commanding 50% charge rate or 70 amps, and the battery was only taking 53 amps an I see the downward slope in the battery summary graph.

I am going to look through the charger settings and see if there was another field that limits maximum charge power. It is not often that my solar is cranking out this much power AND there is no one home and light load on the system. It was really trying to limit export by cranking up the charge rate.

So, once I got home I popped up Grafana and something jumped out right away.

It's subtle and there's a lot going on, but I ended up pushing some current to the grid in the middle of the day.
Yellow is solar
Blue is grid energy and negative readings are my system pushing power to the grid.


Zoomed in, there is a definite cap to battery charging at about 3.4 kw


But here is battery current and voltage. Current starts at 63.1 and drops to 61.6 as battery voltage increases from 53.5 to 55.5 (the scaling on voltage vs amperage makes it look like a slightly different story)


I was assuming it was that I was accidentally limiting charge current to 45% (140 amps * 45% = 63 amps)
But, no it was at 55%.
I upped it to 65% just in case.

I decided to check the battery graph in Schneider Insight.

Holy crap, 55 amps?!
Now this is interesting!
I knew the XW and Batrium didn't exactly agree on current, I didn't realize the difference was quite so drastic. 15% is more than I like. I'm going to get some 3rd party current numbers tomorrow, if I can. I've got a couple amp clamps I can throw on there to figure out who's reading is incorrect.

I am tempted to play with this now, but I'd have to adjust too many settings. That's a plan for tomorrow for sure.



While poking around looking for an alternate reason for the charge current limit, I found this. I can't remember if EPC Discharge Power was present before. I am using EPC Charge Power.
I guess I never noticed it before, it is present in the 2020 modbus map. Address 40152.
Without knowing what Schneider had it mind, I would suspect these two are designed for the type of control we are doing. For me, it's handy because the units are already watts, no conversion needed.

 

[GXMnow]

I am working from the home office again tomorrow, so if we have good sun again, I will watch it and see if the EPC charge limit changed it on mine. It was definitely going into a wattage limit just under 3,000 watts. I never thought I was going to be charging at 3,000, let alone wanting it to do more. But with 4 stings of these batteries, they are not breaking a sweat. Even the full 6,800 watts of charge power would still be less than 12 amps per 60 amp hour cell. So 0.2C charge rate. That's nothing for these. My brother charges his Chevy Bolt at 15,000 watts. That works back to about 16 amps per cell. At a DC fast charge unit, it can charge at over 35,000 watts. That is over 37 amps per cell. That is finally getting over 0.6C, and the cells are rated to 1.0C.

I blew the deal on the Q-Cells panels. I waited one day too long to decide to order, they sold out over the weekend. They only have 4 different panels left on the free shipping deal, and they are all quite a bit more expensive, and you still need to buy 9 for the free shipping. With the 72 cell sized panels, I can't fit 9 on my roof.

And the sun and cool air was crazy here today. My array produce 26.3 kwhs, out of just 4.8 kw of panels. That's 5.48 sun hours, on March 7th. That is a what PV Watts estimates average for the whole month. Last year I didn't hit this much until the start of the third week. I am pretty sure it is a combination of the cool wind we had, and the recent rain cleaned off the panels. I even clipped out the inverters again.

 

[400bird]

We'll I don't understand what is going on. It's like there is a hard limit at 3.4kw of charging.
Have you figured anything out?

 

[GXMnow]

We'll I don't understand what is going on. It's like there is a hard limit at 3.4kw of charging.
Have you figured anything out?

I am watching my system today, but so far, I have not had enough extra solar to hit the limit. It might get close soon though. The battery voltage is up to 55.6 and my solar is putting out 3.79 KW. But the house is using a chunk right now, holding the charge rate down to 2,400 watts. The PLC is commanding 34% right now. 140 x 0.34 = 47.6 amps. I'll go shut off the fridge and dryer and see what I can get it up to.

Okay, I pulled the plug on the PLC, and manually turned up the charge rate. I am charging at 69.95 amps right now 3,925 watts. And it would not go higher that that when I upped it to 55% max charge rate, so it looks like it is limiting to the EPC charge power limit again, since I upped it to 4,000 watts. Just to be sure, I upped it again to 60%, and still 3,925 watts, give or take. I am just going to plug the PLC back in and see how long it takes to get it back to zero import. I am pulling 1,000 watts from the grid right now.

By the time I got back to my PC, the current was falling, and got to 52 amps in under a minute, back exporting 160 watts from solar. With EPC charge power limit set at 3,000 watts, it capped at 50 amps, and at 4,000 watts, now, it capped at 70 amps.

 

[400bird]

Ugh, I'm frustrated. I played with the settings some and didn't make any progress.
I'm going to post some screenshots later, maybe there's a charge setting that I am missing?

 

[GXMnow]

I am still running the 1.08 firmware in mine, so that may may a few differences. Here is my charge settings page.

My PLC adjusts the Max Charge Rate %.
The EPC max charge power is what is limiting my charge power. It was at 3,000 watts, and I was seeing it limit to 2,950 or so. With it set to 4,000, it now went into limiting at about 70 amps, 3,923 watts by the math. I only had it up there for 4 minutes, so I didn't see it slope like I was seeing with the 50 amp limit over a few minutes. I have BMS control turned off, could that be effecting your current limit?

 

[400bird]

I'm glad you posted again, I would have forgotten.
I do have the BMS connected, but the charge cycle is not set to "External BMS"

When I was experimenting, I had:
Max charge rate at 100%
Both max current setting at 140 amps
EPC max charge power I typed in 4,000, 5,000, and I put the slider at the max. None made any difference.
I also put pack capacity at 2,000 Ah, I forgot to put that back down, I just changed it now.

Here's my settings:


I put all these at the max and even disabled EPC.
EPC charge power will display the command from my power monitor program, but it only updates with a complete page refresh. The sun set 4+ hours ago, so the program is requesting 0 watts.


I may try a full factory reset if I can't figure this out.

 

[GXMnow]

That is odd. Your charger settings should be producing 105 amps.

Oh, I have a thought. I also noticed mine limiting power due to grid support settings. I would see the charge current drop off when I turned on the microwave. I think was due to "Load Shave". If the load caused the XW to pull too much power from the grid input it would reduce the charge power to limit what was coming in from the grid. My Load Shave is set to 3.5 amps. I think that means it will allow 3.5 amps to pull from the grid input before it reduces charging or starts inverting to keep the grid current below this setting. Since my solar was cranking good when I did the test, I hit 70 amps of charge current at less than 1 amp from the grid.

 

[400bird]

I agree, it should have been charging well over what I was seeing.

I have grid load shave disabled currently.