diy solar

diy solar

Adding Schneider XW Pro

I’m new to solar, so went with the old “KISS” principle on this first system, using 8 Rolls FLA 445Ah.
XW Pro 6848, 100A-600vdc MPPT charge controller, battery monitor, SCP, gateway, Mini-PDP.
(I basically just checked all the Schneider boxes when ordering)
16 REC-365AA, for 5840 PV.
I did connect to the grid for a little while, only for battery charging. No sell-back.
I shut off the grid breaker a couple weeks ago, as the inverter was using grid to supplement the batteries under load.
(Don’t want to run up the utility $$$ unnecessarily.)
I have left it in “off grid” mode, just letting the loads use battery/PV during the day, battery only @ night.
The loads are: 1HP well pump, (cycles infrequently), workshop/shed wall AC (runs only when there is good Sun), and a small refrigerator in the shed.

Here are the stats from the Schneider gateway.
I’m not exactly sure how to interpret all this data.

PS: I am in awe of you guys that can take an EV battery and make the pieces into a functional battery bank. I would like to learn how to do that, so as to be able to replace the FLA w/ better tech when the FLA reach end of life-cycle.

Screen Shot 2021-06-12 at 10.02.37.png
 
Your solar is clearly producing way more power than you are using. You system is coasting along. I don't know how far into the day or week you are, but I can see that your solar production is way below what 5,840 watts of solar should produce. Can you take a snapshot of the "Battery Summary" page? Pick a day before, so it has a graph for the full 24 hours, like this. I picked June 11, yestrday. XW-Batt_06-06-21.PNG
Since I don't have a DC charge controller in my system, we just see the battery voltage and current. I have a fixed current charge from before 8 am to just after 3 pm, and then the inverter is running my loads from 6 pm, and it wraps into the next day, the previous day ran in to about 2 am this day. Once we see this data for your system, we may be able to offer some settings advice.
 
Thanks!
This system has been operational for about a month, so I am learning and testing.
You are correct, the workshop (a separate building from the house) loads are not very significant,
but it is giving me a chance to see how things operate and behave.

There are manual transfer switches in place to connect the house and workshop to the solar.
We recently did this as a test, just to see that everything works in the house.

House loads ran fine, w/ the heat pump, water heater, range, etc. disconnected.
But I can see from the sensors I have on the grid mains that we will have to be very careful when running the house on solar.
iotawatt.png
It is going to be another month or two before the big hurricanes arrive, so we are on schedule to be ready.

Here is yesterday’s battery summary data.
battery_summary.png
 
This looks odd.
I wonder if maybe the battery monitor is not wired properly. If you look at my graph, you can clearly see when the current goes positive, charging, the battery voltage goes up. And when the current goes negative, discharging, then the battery voltage goes down. Looking at your graph, this trend is not clear, and in many cases, it looks backwards. I think the wires on the current shunt may be reversed. From midnight up to after 7 am, you have a steady current of about 3 amps, with some device pulling about 3 amps which appears to cycle off fo 30 minutes, and then stay on for a bit over an hour. From 7 am to 9:30 AM the solar started to produce charge current. But the current graph goes down while the voltage climbs up. The cycling around 9 am may be the charge controller as well as some loads causing some fairly large quick voltage/current swings. Just before 10 am the battery voltage reached fully charged and the charge controller shuts down. But the battery voltage seems to quickly fall from 58.5 volts down to 53.6 volts. That is a pretty big drop, especially with the current settling to basically zero. You should certainly check all of your connections to your battery bank, I think you may have some resistance in a bad connection somewhere. When you start getting loads after 1:30 pm, just 15 amps is causing the battery voltage to drop over 3.5 volts. OUCH! I know lead acid batteries are not as stiff as Lithium, but this is not good. How long is the wire run from the battery bank to the inverter? What is the wire gauge? This calculates to about 0.23 ohms. On my graph, I have a load hit 68 amps, and the voltage only drop 0.8 volts, or a resistance of just 0.011 ohms. That is a huge difference. It also looks like just before 8 pm, whatever was cycling in the early morning came back on again. Maybe a heater with a thermostat? But it is a small amount of current, maybe just 2 amps.
 
This looks odd.
I wonder if maybe the battery monitor is not wired properly. If you look at my graph, you can clearly see when the current goes positive, charging, the battery voltage goes up. And when the current goes negative, discharging, then the battery voltage goes down.

I was thinking the same thing but, didn't have time to dig in.

To emphasize the possible reversed current sensor:
Overnight, you are pulling 30-40 amps from the battery and the SOC stays pegged at 100%
Once the battery starts charging the SOC drops. Most BMS's count current only to move SOC until it resets at low or high voltage.


Also, is your constant background load 2,000 watts? All night? I'm at like 600 watts, and that's with 2 PC's that I don't power down.
 
Thanks very much for the review & comments!
I agree something is wrong, after looking at the wiring again, I have “A”, “B” reversed.
Will fix that in the morning.
Wiring the battery monitor was complicated.
I had to guess about where to connect “C” “”D” on the BATTERY SHUNT drawing.
I am confused about the drawing that shows + to +, and - to -.
This is supposed to be 8-6v in series, WTH .
If you can see where “C”, “D” should be moved to, LMK.

Separate from all this, the inverter is being replaced under RMA next week.
Originally, the LED’s on the front panel showed battery charge level, now they are all dark.
The on/off button no longer allows the unit to turn on, only “off” works.
2vdc is all the AUX U1, U3 can provide, I need 12vdc to run a battery box fan.
reversed_prescaler_connections.jpg
battery_prescaler_connectionsv2.jpg
battery_monitor_overview.png
pre-scaler_wiring_harness.png
pre-scaler_schematic.png
battery_shunt_prescaler_connections.png
 
Hmm
I think C and D go to the mid point (@24v) but the drawing is confusing because they show two 48v strings (4 12v batteries in series x2)

You should check the directions, but I'd guess either C and D go on the same point or you leave one off because you have just a single string.

The two wires to the current sensor are the ones to swap to change the current flow indicated.
In the drawing A or L- goes to the load side. I think in your set up you have that connected to the battery side.
 
The "C" and "D" mid point wires are to monitor if the battery bank is staying in balance. Since you only have one string, I would connect both of them together at the buss bar between your 4th and 5th batteries. This way it can detect if the upper or lower half has more or less of the total voltage. As long as both halves are equal, they assume all of the cells are close to equal.

When I was looking at making an LFP pack, I had an interesting idea. I would still have a stand alone BMS for the batteries, but to make the Schneider Battery Monitor a bit more effective, I thought about using an Arduino to monitor all of the cells and output the highest cell voltage on one of the mid point wires, and the lowest cell voltage on the other mid point wires. That way, the Schneider software could show the spread from highest to lowest cell on the mid point voltages.
 
It worked!
I fixed the reversed sensor connection, rebooted the battery monitor.
It took a while for the system to make some adjustments, but now it seems to have settled down to what I hope is a normal “Battery Summary” display.

I’ll move the two mid-point wires during the week, when I have the system down to swap inverters.
Thanks very much!
Screen Shot 2021-06-13 at 12.33.17.png
 
FLA battery life is limited, affected by DoD and temperature, among other things. I estimate cost of ownership at $0.25 to $0.50/Wh over life.
After calculating expected cycle life, it it looks like it would wear out in maybe 3 years rather than 10 to 15 years, consider using grid power at night to avoid draining overnight. If cycles are shallow, then not a problem.
 
That does look much better, but it still looks like you have a resistance issue somewhere in the battery bank or wiring. When the charge current stopped, the voltage fell from over 58 volts, to under 54 volts. I am no expert on Lead Acid batteries, but I would think a 4 volt drop from a 30 amp change in current seems excessive. That is still over 0.13 ohms of effective resistance. At just 30 amps, that is a loss of 120 watts. That should be making enough heat to easily feel where it is happening. Load the system and check for any warm connections, and measure across each connection to look for voltage drops. Also measure the voltage of each battery both at no load and at a decent load to see if one is maybe not charged. Balancing may not be as critical with lead acid, but if one battery is far different from the others, it can still cause issues.
 
Thanks, I will conduct those measurements and see what appears.
I have never equalized these batteries. They have been in service for a month.
Can't find any consensus about whether to run equalization on new batteries.
(Most articles/posts say equalization is for older batteries)
 
"Equalize" can be hard on batteries, over-charging full cells to bring low cells up. Apparently regular maintenance for FLA.
My SI-6048 would do a periodic equalize for FLA after a certain Ah amount of cycles (but disable for AGM.)

With 6V batteries (or 12V or if 2V cells have exposed terminals), you can charge each independently to same SoC.
One might hope several new batteries are fully charged, but can't count on it.

So check voltage of each battery, and if needed get a charger and top each one off. 6V/12V charger would be good for other batteries.

When a charged battery settles to a lower voltage than normal, equalization might help. Because FLA you can check specific gravity, but for people with AGM all they have is voltage. I have some older AGM car batteries. I fully charged with an automatic charger, then with 14.5V/1A CV/CC supply, then 15.5V/1A for 8 hours. Voltage came up a bit.

Make sure your absorption and float voltages and charge times are set according to battery recommendations, which do vary.

21 kWh battery, 5.6kW PV, about 0.25C
How does that compare to preferred rate for the batteries? I would have thought they should be lower.
Those batteries should give very long life. I think they are happiest with a particular regulated charge current, enough to stir electrolyte but not heating too much.

Does your system let you program a set battery charge current, adjusting what MPPT delivers to supply inverter load plus charge current?
 
Does your system let you program a set battery charge current, adjusting what MPPT delivers to supply inverter load plus charge current?
I was curious about this as well. I do know you can set the maximum current and it will back off to hold that current, but I think that is just the output of the charge controller. I can't find anything in the Schneider manuals if it knows to add the current being used by the inverter to hold true constant charge current. This won't be a concern on my system, as my batteries can easily take the full power of any solar I can throw at it and should not go into any power limiting until the batteries are full, and then it will try to hold constant voltage so it will essentially be using the solar to directly power the inverter as the batteries rest. But in a case where you need to limit the solar power to be under the battery max charge current, it could be a concern. I wonder if it can adjust the charge current based on the battery monitor current shunt? I will see if I can find any application notes on the Schneider MPPT and Battery Monitor when setup together.
 
Test equipment: Fluke 177 DMM, Fluke 68 infrared.
voltage_temp_readings_june132021 copy.jpg
Battery spec label:
rolls_battery_specifications copy.jpg

schneidersystemdevicespage.png

I wonder if the Schneider system is adjusting the battery voltage?
One of my questions to Schneider was about all the different battery and charger settings in three of the devices.
“make them all match” was the answer.
Except that the battery and charging nomenclature is different on each device, for many of the settings.
And each device has some battery and charging settings that the others do not have.

It would be great if Schneider had an actual Product Manager that would get all the different device/system teams in a room and……

I also contacted Rolls to get their opinion as to how to configure the Schneider battery and charge settings. This is what they sent:
rolls_suggested_settings copy.png

They did say that the 100A charge controller output was too much for these batteries.
However, in monitoring the PV and charge controller activity, this has not happened yet.
When this off-grid system uses the batteries for a period of time, and I see the Sun finally come out and hit the PV, the watts goes over 5kw, and the amps go to 90+ for a few seconds, and ramp down as the batteries absorb the input. Thus I have not changed this yet, until I have a better understanding of the effects of all these different settings.

Why would Rolls want these 445Ah batteries configured as 325Ah batteries in the Schneider devices?
(I left it at 445 until a better explanation appears)

Here is the inverter configuration. I have the grid AC in breaker off, so there is no charging from the inverter.
I don't know if the BMS settings do anything when the system is configured for FLA, that might only be for Lithium?

inverterbatterysettings.png
inverterbatterymanagement.png
Here is the MPPT charge controller configuration. It is doing all the charging, and may or may not be fighting with the other devices for battery management.
mpptchargecontrollerchargersettings.png
mpptchargecontrollerbatterysettings.png
Here is the Battery monitor configuration.
batterymonitorchargersettings.png
Screen Shot 2021-06-13 at 16.43.55.png
 
Last edited:
I finished making the last two changes to get officially through the inspection and thought I would share some pictures of the set up.

The trench needs to be filled and I've got to pour the walk way back in but, the inspector didn't want/need to see that. For the interior of the battery box, it's not as clean as I would like, but the unexpected wood cost increase means that I used whatever scrap I had around. I'm planning to double capacity, maybe I'll redo the box lining then.
 

Attachments

  • PXL_20210611_173133054.jpg
    PXL_20210611_173133054.jpg
    139.8 KB · Views: 30
  • PXL_20210613_192350638.jpg
    PXL_20210613_192350638.jpg
    165.8 KB · Views: 30
  • PXL_20210613_192421934.jpg
    PXL_20210613_192421934.jpg
    284.1 KB · Views: 29
  • PXL_20210613_192338658.jpg
    PXL_20210613_192338658.jpg
    243.2 KB · Views: 28
Very neat and organized installation!
Nice battery box - looks like one of those giant job-site toolboxes.
 
I will look over the Rolls battery specs and the Schneider settings, but I have to work today, so it may take a bit to get back on it. The battery voltages do look good, so see if you can check for voltage drops under a decent load. I still think there may be a resistance problem somewhere. 30 amps should not be causing a 4 volt drop.
 
Thanks!
It will be a week or so, replacement inverter was delivered this morning.
Shutting down the PV tonight.
I have to take the system apart to get the old inverter out, safely, and then hope the new inverter works!
schneiderinverterreplacement.jpg

inverterreplacementcrane_cradle.jpg
 
Back
Top