Adding storage to my Enphase system

[GXMnow]

My DC panels are a total of 2,000 watts going to a Victron 150-35 charge controller that charges my "48 volt" battery bank. It is sort of "Off Grid" in that the Victron is only a battery charger.

My "Main" solar array is 4,800 watts of panels on Enphase microinverters. That system is AC coupled into a Schneider XW-Pro battery inverter. The XW-Pro is tied to the same battery bank as the Victron charge controller. I thin have a PLC that monitors power flow and sends commands to the XW-Pro inverter to adjust the charge rate when the sun is up and the export rate to my main panel when the sun goes down.

The Victron charge controller never directly exports to grid. Before 4 pm, I don't bother trying as the grid rate is cheaper. I just let it take the extra it needs from the grid and the Victron will keep charging up the batteries. Right now, it is 12:15 and the Enphase system is making 3,200 watts. The house is using about 1,200 watts without the A/C running, so 2,000 watts is charging into the battery. And the Victron system is making 1,500 watts, also charging into the battery. When the A/C kicks on, it pulls 3,800 watts more. The XW-Pro charge rate drops to just 400 watts. The remaining 2,800 watts from the Enphase system is running the normal house loads and only 1/3 of what the A/C needs. The other 2/3 for the A/C comes from the grid. The Victron has no idea this is going on and is still charging the battery at 1,500 watts.

But after 4 pm, the grid rate is nearly double. The PLC then commands the XW-Pro to cover all the loads, including the A/C. If there is enough sun still, it could be inverting from battery and the Victron charge controller input, and it pushes that current back to my main panel to cover running my A/C system and any other loads in the house with the grid exporting a tiny bit.

Here is my battery summary graph from the XW-Pro for July 5th

In the morning, I had it charging slow from grid. It went into tracking control at 6 am. The battery voltage charges up nicely, even when the A/C kicked on, the charge current just dropped, but does not discharge. After 4 pm, each time the A/C kicks on, the battery discharges fast. It does not help that my son is back, and his gaming PC is pulling 600 watts all night. It's cranking out heat too. Each time the A/C stops while the sun is up, the voltage recovers a bit as the Victron just keeps pumping into the batteries. After 9 pm, I also set it to grid charge a bit to make sure I can cover the 4 pm to 9 pm the following day.

 

[400bird]

My question is how late into the evening does your battery run your loads when it is 100+ out?
Or maybe you stop covering loads once the peak rates are done.

 

[GXMnow]

My question is how late into the evening does your battery run your loads when it is 100+ out?
Or maybe you stop covering loads once the peak rates are done.

I am still always making it through the 4 pm to 9 pm high rate time. But if the battery is running low, I use a bit of grid charging at the cheap rate.

Since I was here most of Saturday and Sunday, I did a little extra grid charging, and I got the battery back up to my full charge by 3 pm. But I used 7.3 KWHs from the grid to do it. But since I did top it out, I then exported over 10 KWHs back out again. It was not as crazy hot today, so I think I may actually net out close to zero. I work a long day tomorrow, so I want get the SCE data until late to see how it worked out. It's funny, the JK-BMS on my older battery bank actually reported 99% SoC. I have not seen it that high in months. It has been hitting my normal full 95% voltage, but for whatever reason, the JK was not resetting, and the SoC just kept creeping lower and lower and was telling me 0% SoC when the voltage was still over 3.7 per cell. That is the 50% nominal voltage on NMC cells.

It is 10:30 pm on Sunday night as I am typing this and my voltage is still up at 53.7 volts or 3.84 volts per cell. On the NMC SoC chart, that works out to near 70% SoC. It cooled off enough, I have the windows open, so my only extra load is my son's gaming PC. He is not in a match right now, but it is still pulling about 300 watts. It's 10 hours until there is enough sunlight on the panels to power the house and begin recharging. The system is pulling 1,500 watts from the battery. That is 15 KWHs until sunrise. That is 43% of my capacity. 70 - 43 = 27% SoC remaining. Ouch. I won't let it go that low overnight. That would be 3.6 volts per cell x 14 = 50.4 volts. It could make it. But then it won't fully charge tomorrow.

With both the Enphase AC and the Victron DC systems, I am producing a solid 35 KWHs a day. My battery capacity is 36 KWHs.

Today, the XW pulled in 7.4 KWHs from the grid. 6.6 KWHs of that was pushed to the battery. The Victron charge controller pulled in 10.63 KWHs that went into the battery. The Enphase system made 26 KWHs. Most of that went directly to running the house load and the A/C, but some did cycle in and out of the battery as well. I honestly use a bit over 50% of the battery capacity. At 3 pm, I saw the battery topping out on the Victron and it switched to Absorb while I was watching it. I didn't want to throw away any power, and the A/C was cycling on. So I set the XW to allow grid sell a little early and it pulled the battery down enough in just 2 minutes to put the Victron back into Bulk MPPT mode. When the A/C cycled off, the Victron again, got the battery back to full right at 4 pm. The A/C cycled on 12 times today. Each cycle was about 15 minutes, sucking up 3,500 watts or 825 watt hours per cycle. 12 x 0.825 = almost 10 KWHs just running the A/C. And this was less than the last few crazy hot days. Looking at it like this, the DC panels basically ran the A/C today. The Enphase ran everything else. Now if I can just get my son to sleep his PC when he is not gaming. In sleep it drops to about 35 watts, Idle it's 300 watts, in game playing, I see it hit over 600 watts.

The Midea 12,000 BTU "U" A/C unit upstairs is only averaging about 200 watts. I have it in low eco mode. The inverter compressor slows way down to reduce power consumption and it never pulls any start surge. It just ramps up slowly. The highest I have seen it pull is still less than 600 watts. So it cools a large room for less power than my son's PC which is turning his room into a sauna.

 

[GXMnow]

The weather calmed down a little. The A/C still ran a lot, but with my son and I both working, we didn't use a ton of other power. The solar still under produced by a little, with the battery ending the day just 0.8 volts before the start of the day.

It would have been net zero grid power except for a few times where the load required more than 16 amps back in the main panel. The XW-Pro is told it only has a 20 amp breaker and it won't exceed the 80% rule of 16 amps of back feed. So I ended up pulling a little rid power a few times, including a little in the 8 pm hour at the higher rate. On the plus side, if I had it supply that power, the battery would have run even lower. This amount of import and battery discharge once in a while is not a problem at all. It only becomes a problem when it is happening every day for over a week. If the battery creeps down a little, day after day, I end up charging a bit from grid. It would be better to just use a bit less battery and use grid power during the low rate time, rather than discharge the battery have have to charge it back from grid. Even doing the charging at the cheap rate, there is a near 10% energy loss converting the power to DC, charging the batteries, then inverting it back to AC to push into my panel again. To do this, I have changed my setting in the XW-Pro to block it from exporting to the main panel starting at 2:30 am. I may push that back to 1 am or even midnight. My main panel loads ar forced to run off of grid power then and don't pull the battery down.

The utilities certainly were not dumb when the set the high rate to be 4 pm to 9 pm. That is exactly the time when my A/C has to run the most. When my solar panels produce enough power, I end up topping out the battery early and exporting some power before 4 pm. But on these days where it is close or a little short, it is worse. The battery still tops out because the house stays cool enough that the A/C does not need to run. The XW ends up exporting at the cheap rate, but then after 5 pm, the A/C load is pulling the battery down. Being able to store that exported power would have been nice. I have moved the set point to allow the battery to run a bit lower, leaving more room for charge, but for this 6 weeks of the year, I just don't have enough battery capacity or solar panel production. Even if I use it at exactly the right time, it's just not enough.

I think I am going to contact my original Enphase installer about adding on 4 more panels. As tempted as I am to DIY it, doing it safe and legal requires pulling my meter to change the main breaker from a 100 amp to a 90 amp so I can up my back feed breaker from a 20 to a 30 amp. I am not comfortable trying to change the main breaker myself. At the very least, it would mean cutting the seal off the meter to pull it.

If (when) I add these new panels, I would probably take the upper row of 4 from my upper roof, and move them to go to the right of my lower array of 8, making that an array of 12 matching panels. Then put the new 4 panels on the upper row on the high roof. The get the most sun up there, and are also hard to see. So it they look a little different, it won't really be visible, and the new more efficient, higher power panels will be in place to get the most sun. Those 4 new ones can also get higher power iQ7A inverters. But that will also mean pulling a second pair of #10 awg wire from the iQ combiner to the junction box in my attic. With 16 iQ7's, I maxed out a single input on the iQ combiner. With this change, it will be 12 iQ7's on one feed, and then 4 iQ7s and 4 iQ7As (or 8A's) on the second feed from the upper roof. That would be 12 amps from the lower roof and 9.8 amps from the upper roof. That's 21.8 amps total, perfectly fine on a 30 amp back feed breaker. That is a 1,392 watt increase in maximum output from the Enphase inverters. That is increasing my peak AC solar power by 36% if I max out the four iQ7A inverters with 400 watt panels. The DC power of panels on the Enphase system would go from my current 4,800 watts to 6,400 watts. That is probably enough to force me to NEM 3.0 if I do it all through proper channels. I have to really think about that one.

 

[Siamac]

Let us know how much they will charge you for the 4 extra panels.

Why not add more DIY panels?

 

[Hedges]

I think I am going to contact my original Enphase installer about adding on 4 more panels. As tempted as I am to DIY it, doing it safe and legal requires pulling my meter to change the main breaker from a 100 amp to a 90 amp so I can up my back feed breaker from a 20 to a 30 amp. I am not comfortable trying to change the main breaker myself. At the very least, it would mean cutting the seal off the meter to pull it.

what bad thing would happen if you backfed 24A through a 30A breaker, into 100A busbar fed by 100A breaker?

(Locations of neutral wires, and amount of 120V imbalance, will determine neutral current in busbar.)

If you fed Enphase through a magnetic-hydraulic breaker, it could be used near 100% rated. e.g. 25A magnetic-hydraulic in series with the new 30A breaker. (fudging a 125% driving speed with 120% speed limit, not enough to get you pulled over?)

 

[GXMnow]

Why not add more DIY panels?

what bad thing would happen if you backfed 24A through a 30A breaker, into 100A busbar fed by 100A breaker?

Adding more DC DIY panels would be easy, and I have thought about making a solar awning across the width of my garage. I have 18 feet wide. I can fit a pretty good sized row of panels. But most of the year, a lot of the extra energy from them would be curtailed away as the battery is full. At that point, I would really want to update my code to understand how to back feed from DC in. I know it can do it, and I have flow charted what the code needs to do.

Adding AC panels, once the battery is full, the extra just goes to the grid and I get credit, very easy on my part. And updating the AC system also giving me that up to 24 amp back feed to the main panel helps cover the load of the A?C running with other loads.

Can I DIY some AC panels? Yes, but... Even with my installer access, I don't seem to have access to add on to my Enphase system layout. I think I could add the new inverter serial numbers, but they would not show up in the layout. Not too big of a deal. I have even thought about just getting one of those 4 channel micros and adding it as a new drop. It would not be metered on the Enphase App, but not too big of a deal. One small issue I am running into it is seems ALL of the new panels on the market are a bit wider than my existing SilFab SLA300M panels. So my idea of replacing the upper 4 on my upper roof is not looking good. Add 40 mm per panel x 4 panels, and the sides will go off the ends of the rails. Oops.

In all honesty, I totally agree, If I just put in the 30 amp back feed breaker, nothing bad will happen. There is absolutely no time when my home will pull more than 100 amps. So there is no concern of exceeding the bus bar rating. But if I ever do have a fire etc. and an inspector finds a 100 amp main and a 30 amp back feed, it exceeds the 120% rule and my insurance could tell me to take a hike. Yes, my DC system is all also unpermitted, but it can be easily wired to be "Off Grid" in a hurry. As long as it is not the source of a fire, I am not worried about it as it sits. But my main panel is a concern for me. It is original from 1969 as well. The bus bars look fine and it runs cool. With my sub panel now, the only big load in the main panel is the air con, but when I do get an EV, that will also be a big load in that panel. I will try my best to NOT be charging a car when the A/C is running, but it could happen. But I will most likely also limit EV charge power to about 30 amps. I will also have a 240 volt outlet for the EV charger on the backup sub panel. I hope to never need it, but that would allow me to charge a car at 20 amps during a grid outage. I also will never "Export" more than 16 amps out to the grid. I need the extra current to cover the A/C load in the main panel. With the Enphase system back feeding through the XW-Pro, that one breaker could be taxed with 28 amps from the XW plus 16 amps from the Enphase inverters. That's a possible 48 amps of inverters. It is only software limiting it to 16 amps now.

Safe is fine, and I am sure a 30 amp breaker and software limiting would be completely safe. But to be legal, I also need to change that main breaker to a 90 amp unit, and GE does make it for my panel. It's just a matter of having to pull the meter to swap it. I have thought about just biting the bullet and upgrading to a new 200 amp panel. But it is not a trivial job. With the meter being in my existing main panel, and it all being flush set into a stucco wall, it becomes a bit of a construction project.

I talked about it when I did the original solar install, and that was the big reason I only put in the 16 iQ7 inverters. It limits right to 16 amps. He estimated about $4,000 more to upgrade the breaker panel as part of the solar install. Knowing what I know now about the changing NEM rates and such, I should have done it then. Along with it though, would have been 4 more 300 watt panels and iQ7 inverters, taking on yet another $3,000 or so.

Just doing a little hunting on the net, I found a place selling surplus modules. They have 365 watt Solaria modules with iQ7+ Enphase microinverters already on the back of each panel. I might be able to fit 5 of those across the width of my garage. It would make an awning just over 5 feet out from the face of the garage. I am sure some nosey neighbor would hit me up for a building permit.

 

[Hedges]

Put a GT PV inverter after the relay for A/C compressor?

 

[GXMnow]

Put a GT PV inverter after the relay for A/C compressor?

That is an option I have seen for people who don't have an energy export agreement. But in my case, I don't see it being very cost effective.

The A/C cycles are typically just 15 minutes, and the GT inverter waits 5 minutes before it starts producing any power. I lose 1/3 of the energy it could make while the A/C is running. And I lose all the extra production while the A/C is not running. If I could route that extra energy into the batteries or export closer to my monthly limit, it helps pay for the gear. Limiting it to only the AC running slashes the value a lot. Today, my A/C didn't make it's first cycle until 1 pm as solar noon was already passing. Half the production would just be thrown away.

With iQ7+ inverters on the new panels, I could have my PLC connect them one at a time as the power production is below my limit. Even today, due to the heat, my maximum from the Enphase system only hit 3,200 watts or 13.3 amps. I can safely export 16 amps x 240 volts = 3,840 watts. So I could have had 2 or even 3 more panels running without exceeding my back feed breaker. And obviously, I can use even more and never hit the limit actually exporting. I may lie to the XW-Pro and tell it the breaker is a 22 amp to squeaze a little more when running the A/C. It's annoying knowing I have a 6,800 watt inverter, but it has to cap under 4,000 watts because the breaker is too small at the main panel. Of course, I could also be running another 2,800 watts of load in my backup sub panel. Even more if the Enphase panels are pushing current. It's 5:50 pm and the Enphase panels are still making 1,300 watts. 900 watts are being used in the backup sub panel when the A/C is not running. It just cycled on again, and the furnace blow going to high makes th sub panel load hit 1,500 watts. So right now the XW is inverting 4,300 watts as it tries to zero the grid power.

 

[GXMnow]

I just posted in the thread "Panel Cleaning!!" about rinsing off my panels this morning.

Panel Cleaning!!

Anyone use somthing like this to clean panels on their roof? Curious is the rubber on the squeegee could harm them some how? Thought about spraying them with water from ground level and then dipping the scrub pad in some soapy water and scrubbing them. Then rinse them off and squeegee them...

diysolarforum.com

You can check it out there and add your thoughts. My power increase over yesterday at this time was nearly 22%. They were DIRTY!

But my post here is about something else. My original SilFab panels on th Enphase microinverters have bee 100% rock solid. They just plain work. Historically, the panels on my upper roof have produced more power than the group on th lower roof. And that makes sense as the lower roof has some shadows at different times of day where the upper roof really never has shadows. But lately, when I look at the array energy per panel chart, I have been noticing that one panel on my lower roof is consistently been the top producer. Even out doing all of the upper roof panels by a fair amount. Yesterday with all the dirt still on the panels, that panel made 1.57 KWHs and the next best panel only did 1.52 KWHs and the best on the upper roof was at 1.50 KWHs. It's only 3% more energy than the other best on the lower roof, but 5% better than the non shaded upper roof panel, and over 6$ better than the average of all of the panels. And while I was up on the roof cleaning them off today, I noticed something. That panel looks "different". It's still a SilFab Black on Black panel with the same dimensions and cell type, but the cells inside had a different shade, with a slight blue tinge compared to the more grey/brownish of the other 7 around it. I am certainly not going to complain. From the street, it looks identical. What I think happened, is they ran out of the 300 watt version, and that panel is probably either a higher grade, or maybe even a 315 watt panel which they did sell at the time, but at a much higher price.

The panels are starting to heat up and the power improvement over yesterday is down to 16%, but that's still huge.

 

[GXMnow]

I'll keep it short here since I did a full report in the Panel Cleaning thread.

The sun was up for 2 hours before I did the cleaning.
But the total for both systems still produced 4.4 KWHs more than yesterday, with both systems gaining over 13% from the day before. Yeah, they were crazy dirty. Tomorrow should do a little better with the morning sun being on clean panels as well.

 

[faccnator]

Man that’s so much more than I would have expected. Does it rain a lot where you are?

 

[GXMnow]

Man that’s so much more than I would have expected. Does it rain a lot where you are?

No, we get very little rain. That's the problem. We also have had several brush fires around the area, and 2 are still burning. Between the ash flying in the air and the dirt kicked up when the use bulldozers to cut fire break lines, and then high winds throw it all in the air, it was like a layer of grey/brown snow on everything. The truly scary part about it is that I had just rinsed off the panels about two months ago. They had more crap on them than the previous 6 months of buildup. Talk about horrible air quality... We are breathing this air full of smoke, ash, an dirt.

 

[Recovering EE]

Time for a system operation update.

To start off, here is the chart for so far this billing month from So Cal Edison.

View attachment 226328
As the heat moved in (and my son got home from college) my excess energy export dropped quite a bit, and 2 days I had to net import a bit of power. But so far, for this crazy hot month, I am still averaging a net export of over 3 KWHs per day. The Enphase production has been very good, It has been very consistent, topping 25 KWHs every day except June 23rd it only did 24.4 KWHs. Here is just July 2nd as they all look pretty much the same.
View attachment 226330
The reason for the reduced export is my consumption. My Central A/C is a fairly big draw, and I also have the 12,000 BTU window unit running upstairs. But to be honest, it is reducing my electric bill as it vastly reduces how much the central air has to run. To give you an idea of the power draw, here is what the XW-Pro battery had to do on July 2nd.
View attachment 226331
Each one of those big power spikes after 1 pm is when the A/C has to kick on. I did make a few small changes to the XW settings. After 1:30 AM, I now block selling to my main panel and the grid. From 1 am until the sun comes up, we don't use much from the main panel, so I end up just using a small amount of grid power running a few little loads around the house. Also from 8 am to 4 pm, I won't stop the XW from charging if the load in my house exceeds solar, it just drops it to 7 amps of charge current. So when the A/C runs before 4 pm, it is taking some grid power. The end result is the battery ALMOST ended the day at the same voltage as the start. Here is the SCE trace for the same day.
View attachment 226333
From 2 am to 7 am you can see the small grid usage, peaking just over 150 watts at 6 am. Once the sun is up, I start to export, but most of the power is going into charging the battery. In the 1 pm and 2 om hours, the A/C had to run, so that pulled some grid power. But in the 3 pm, 4 om, and even 5 pm hours, the battery was full as far as the XW-Pro was concerned, so it exported a bit of the extra Enphase system power. From 6 pm on, my PLC programming is using battery power to zero the grid to about 60 watts as usual. The A/C cycled on 5 more times, but you could never tell from this. The batteries and XW inverter easily covered the load. The end result, is I still managed a net 1.29 KWHs of export, and most of it ended up at the higher peak rate. Without the DC solar panels, this just would not be possible. The Victron charge controller has been working big time.

View attachment 226335
The latest day is on the left. This billing month began on June 17th. The total production was the lowest on that day as you can see it went into absorb and float. After that, with the A/C sucking down some power, there was more room left in the battery. From the 22nd on, it never went to float again. It pushed over 10 KWHs, and even over 11 KWHs a few times each day since. Without this extra energy going into my battery bank, I would have had to pull over 6 KWHs from the grid every one of those days. The bar for today is low because I took the screen shot before 11 am. Needless to say, the added DC panels are paying for themselves very quickly. Just one summer, shaving off over 6 KWHs each day (more like 10 really) for 3 months. at 27 cents, that's over $140 in savings (could be $240 at 10 KWHs a day) in under 3 months just from the power of the DC panels alone. Here are the details from the DC production yesterday.
View attachment 226337 View attachment 226336
It actually produced 10.82 KWHs, leaked at just 1,452 watts due to heat and dusty panels.

I am not too worried about it, but I do have to think about "Micro Cycles".
Each time the A/C system fires up, the battery goes from charging to discharging, and then when it cycles back off, it goes back from discharging to charging again. You can easily see how the voltage dips and then rises again. It is only doing it 1 or twice in a full hour, so it's not all the time. And if we look at these cells in an EV, each time you accelerate and then brake, it will be doing this at much higher current levels. I doubt it is hurting these cells at all, but I never had this before the DC system was added. The charge and discharge rates would fluctuate, but it either stayed charging or stayed discharging. Now the energy flow is reversing with each A/C cycle. Yesterday, between the A/C load, and everything else in my house, the battery power hit as high as 4,868 watts. For the battery to never see the current reverse, I would need enough DC coupled solar power to exceed that power level. But as the sun is falling, it will still happen.

This is only really happening when the batteries are between 60% and 90% charged. Worst case would be going from about 2,500 wats of discharge to 2,500 watts of charge. That is a C rate of less than 0.07 in or out. as it swing. I doubt this will be any issue at all, but seeing the battery bank voltage dip and recover 10 times just looks wrong.

You're clearly inland. The front half of my production curves are jagged and spikey.

The marine layer nearer the coast is a biotch. Regularly impacts both skirts of the curve. The front especially.

 

[GXMnow]

You're clearly inland. The front half of my production curves are jagged and spikey

Correct.
I am in a valley close to the edge of the Mojave Desert. We have far more clear sunny skies than a lot of areas of So Cal. We do get clouds from time to time, and when we do, they hang in all day.

 

[GXMnow]

Another billing month has come to an end. Between this heat wave, the dirt on the panels, and my son using a crap load of energy, my solar didn't quite keep up. We didn't miss by much, but my energy credit balance is going down a tick.


In 31 days, we ended up buying just 43 KWHs from So Cal Edison. That $5.47 shown there is just the taxes an stuff. The energy will take about $12 off of my energy credit. I didn't get the actual bill yet, but it will come soon. With the panel cleaning, and the weather calming down a bit, I should do a little better, but my son is still using a lot more power than I expected. Last year, on the same billing period, I actually gained an energy credit. Of course, the bad heat wave came later in August last year.

One of the big issues is my son is doing 1 or 2 loads of laundry, EVERY DAY. How does he go through so much? Between the gas dryer and water heater, my gas bill is also up 35% since he is back here. I had a little talk and we have to figure out how to make it a bigger load once every 2 days.

Oh well. I still have a decent credit, but I am not sure I will be able to keep all year without adding a few more panels.

 

[GXMnow]

I was out of town for work for 4 days, so no one was watching my system since Monday night.

My son is a total energy hog. Between his PC and leaving EVERY light on all day, doing 2 loads of laundry every day, running, and not opening windows at night, and more... WOW. The battery went from full (57.9 volts, about 95% SoC) on Tuesday at 4:08 pm all the way down to 50.7 volts by midnight. The A/C kept cycling past midnight to even one cycle after 2 am. Opening the windows does wonders when the outside temp falls. It ran all the way down to just 50 volts even before the sun came up at 8 am. That is 3.57 volts per cell at rest. On NMC cells, that is down to 25% SoC. So the house sucked down 70% of my 36 KWH battery bank, on a day the Solar panels also made 24 KWHs from the Enphase and another 10.4 KWHs from the Victron system. I do know he also worked on his car, so I am sure the shop air compressor ran a bit, and the big ShopVac is also plugged in by his car. But that is a lot of energy being used. I see one power spike at 7:30 pm the day I left where the XW was pulling over 108 amps from the batter bank, exceeding 5,800 watts. It looks like the A/C was on, and the compressor in the garage started. And of course it was at the highest rate time of day.

Starting from so low, the system didn't recover. On day 2, the battery voltage only peaked to 55.5 volts (about 82% SoC) before the sun was going down. Again, the Enphase system made 25.4 KWHs and the Victron shows 10.4 KWHs again. By midnight, the battery bank was down to 50 volts again, and kept dropping the next morning. At 6:29 am, it hit low battery cut which I have set at 49 volts (just 15% SoC). I was surprised to see the XW actually went into a charge cycle and started using grid power, but by 7:30, the solar panels were making more than the house was using, so th grid power went to zero and it started charging faster, but again, starting so low, and the A/C coming on even before noon, it had no chance of catching up. It only managed to get back up to 54.5 volts (75 % SoC) at 3:40 pm. So 7% lower than the day before.

So that brings me to today, I got home about 6:30 PM. It's now about 7:30. I do not see a log entry why it happened, but the XW triggered another charge cycle 2:51 am this early morning. The battery voltage had dropped quite low, but it was still at 49.1 so it should not have triggered an error. It charged from grid power for a little over 3 hours at just 7 amps. Got th voltage back up to just 50 volts even by 6 am where my PLC takes over and stopped it from charging while it watches for PV coming in. So the battery ran down again, to 49.3 before the sun came up and started charging again. And today as well, the A/C was cycling on even before 11 am. So once again, the battery is dropping low.

On the good side, it still rarely pulled any power at the high rate. It pulled from battery as needed, and the battery always made it well past 9 pm. So the bulk of the extra grid power used was still all at the lowest rate, but that is now 27 cents per KWH so it still adds up fast. Peak weekday rate is now up to 50 cents after my low usage discount, it 60 cents if I use past me baseline allowance.

To keep up with my son here on these 100+ hot days, I need like another 2,000 watts of panels to break even. And even then, it looks like my battery capacity might not be enough as he uses a ton of power when the sun is down.

I was a bit nervous that maybe I was down a string of batteries. But I checked, and all 4 strings are pulling current, the JK BMS is showing only half the flow to the XW, and the total Amp Hours cycled still meshes with my total battery bank being a bit more than 36 KWHs of capacity. The XW is reporting that it was discharging up to 24 KWHs out of the battery on these crazy days. But the XW only charged in about 10 KWHs a day, because another 11 KWHs was coming in from the Victron system. So it fell short by 3 KWHs. 24 / 36 = 67% capacity used out of the battery. That is close to the 70% I calculated above from the voltage change alone. And the real numbers are probably closer because the XW power per day is midnight to midnight, where I calculated from 4 PM full charge to 8:30 am lowest SoC reading.

Until this weather calms down a bit, I am going to block sell starting at 1 am, so any load in my main panel will just get from grid and reduce the overnight discharge on the battery. If I have to recharge from grid, it costs a little more due to efficiency losses. I am just better off using from grid rather than charging from it to use during off grid time. It only makes full sense to use grid charged battery power during peak rate. If I am going to run out of battery, Charging from grid is still wasting a bit.