Conext ComBox reporting incorrect power values

I have a pretty basic off grid system made up of Conect XW components.  XW+6848, dual MPPT60, ComBox, SPC, and AGS.

I have the Combox send me a daily email report, and there is a consistent discrepancy between the energy used by the loads,  and the energy used from the batteries.  The load energy should be equal to or less than the energy drawn from the battery.  That's physics.  But the ComBox reports the opposite, showing more energy consumed by loads (6.5kwh) than drawn from batteries (4.9kwh).  There is either a bug here, or I'm interpreting the report wrong.  This becomes clearest on rainy days when there is zero PV power produced, and the only non-zero power values are the loads and battery draw.

Below is an example report showing my magic system that creates power from nowhere.  Note the "Today" values for power consumed by loads and power used form battery.  Huh?

Energy Harvested from PV


Energy (KWh)Duration [hh:mm:ss]
Today0.00000:00:00
This Week41.75019:24:38
This Month116.48083:44:54
This Year753.847559:35:16
Lifetime753.847559:35:16

Energy Used by the Loads


Energy (KWh)Duration [hh:mm:ss]
Today6.51623:47:56
This Week40.935140:38:12
This Month169.962523:30:50
This Year633.8242453:36:58
Lifetime633.8242453:36:58

Energy Stored in the Batteries


Energy (KWh)Duration [hh:mm:ss]
Today0.00000:00:00
This Week41.75019:24:38
This Month210.710101:43:22
This Year934.732588:05:50
Lifetime934.732588:05:50

Energy Used from the Batteries


Energy (KWh)Duration [hh:mm:ss]
Today4.97723:47:56
This Week32.827140:38:12
This Month133.329503:21:24
This Year499.4352416:02:42
Lifetime499.4352416:02:42





«1

Comments

  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    You mentioned AGS. Where is Kwh produced by the generator accounted for?
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    Estragon said:
    You mentioned AGS. Where is Kwh produced by the generator accounted for?
    In this report (I omitted it with my cut and paste) it shows as generator run time and kwh produced.  On this particular day there was no gen run time.  The actual measurement and accounting must come from the XW+6848 which measures input voltage and current from generator operation.

    In the example I have posted, I believe all measurements come from the 6848.  As a stand along device, it measures DC voltage and current, and output AC voltage and current.  From those you can easily calculate load kwh and battery draw kwh.  It's almost as though it has mixed them up when reporting.  Due to losses in the inverter, I would expect battery draw to always be larger than the loads powered, but it reports the opposite which of course is categorically impossible.

    One thing worth noting is that when the panels do produce, by way of the MPPT60s, the solar power produced kwh correctly matches the kwh reported as going into the batteries, so that part makes sense.  And the same is true when the generator runs.  Its kwh output matches the energy going into the batteries, and is typically around 30kwh for a gen run.  It's the draw out of the batteries relative to the loads powered that is wacky.
     

  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    I'm a little suspicious of the reported pv. Mine will produce a bit of power unless it's really gloomy all day, or the array is covered by thick snow. It might only be a couple hundred watts, but not zero. I'm wondering if some power is coming from pv and going to loads, getting included in the AC load number from the inverter output, but not showing up in the DC numbers. Maybe the pv and battery numbers are from a shunt (as opposed to the controller output)? If so, the pv power wouldn't be crossing the shunt, so wouldn't get included in DC numbers.

    The battery in/out ratio looks roughly right (~10% charging losses + maybe 20% inverter loss).
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • MichaelK
    MichaelK Registered Users Posts: 234 ✭✭✭
    Is this an example of not accounting for "power factor" in electrical calculations.  In my own case, my submersible well pump uses 10amps at 240VAC, which equals 2400VA.  The display on the XW reads 1.7kw.  That works out to be a power factor of ~0.7.
    System 1) 15 Renogy 300w + 4 250W Astronergy panels,  Midnight 200 CC, 8 Trojan L16 bat., Schneider XW6848 NA inverter, AC-Delco 6000w gen.
    System 2) 8 YingLi 250W panels, Midnight 200CC, three 8V Rolls batteries, Schneider Conext 4024 inverter (workshop)
  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    I don't think PF is the issue. It's likely included in the inverter AC output number, as the power gets made even though it doesn't end up as useful "work".

    Assuming it isn't included though, it would only increase the apparent discrepency between watts produced or taken from storage (which appear to be less than) vs those consumed.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    Does anyone else with a ComBox see this phenomenon?  I think the first diagnostic step is to understand if it's unique to my system, or an issue with the ComBox.
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    I don't have XW charge controller, so I can't make the same measurements,  Generally, the XW gear, has honest reporting. It may gloss over it's internal self-consumption.
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    Although I haven't tried it with the MTTP60s disconnected, the issue would appear to be strictly between the XW+ inverter and the ComBox.  The XW+ is the only thing pulling power out of the batteries, and it's the only thing serving loads.  And the ComBox is doing all the data gathering and display.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    The XW is reporting the amount of energy that it uses to charge the battery to full and the amount that is used by the load.
     It has shunts internal to the XW and uses data from the mppt also.  I think you are overthinking this and there are better charts to use than what you posted here for offgrid.

    If you look at the data at just before sunset on a good day. The Mppt harvest will be equal to the loads during the day, the amount to charge the battery, and the loss in the system.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    edited December 2017 #11
    The XW is reporting the amount of energy that it uses to charge the battery to full and the amount that is used by the load.
     It has shunts internal to the XW and uses data from the mppt also.  I think you are overthinking this and there are better charts to use than what you posted here for offgrid.

    If you look at the data at just before sunset on a good day. The Mppt harvest will be equal to the loads during the day, the amount to charge the battery, and the loss in the system.

    Right, and that part tallies out just fine.  But with no sun and no other charge sources - just inverter-powered loads and inverter draw from the battery - shouldn't the two be the the same, give or take conversion efficiencies?  And how could inverter-powered loads ever be more than the power draw from the battery when the battery is the only source of power?
  • mike95490
    mike95490 Solar Expert Posts: 9,583 ✭✭✭✭✭
    >  And how could inverter-powered loads ever be more than the power draw from the
    >  battery when the battery is the only source of power?

    I don't think the inverter can count it's own internal losses
    Powerfab top of pole PV mount | Listeroid 6/1 w/st5 gen head | XW6048 inverter/chgr | Iota 48V/15A charger | Morningstar 60A MPPT | 48V, 800A NiFe Battery (in series)| 15, Evergreen 205w "12V" PV array on pole | Midnight ePanel | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph ) on a timer for 3 hr noontime run - Runs off PV ||
    || Midnight Classic 200 | 10, Evergreen 200w in a 160VOC array ||
    || VEC1093 12V Charger | Maha C401 aa/aaa Charger | SureSine | Sunsaver MPPT 15A

    solar: http://tinyurl.com/LMR-Solar
    gen: http://tinyurl.com/LMR-Lister ,

  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    mike95490 said:
    >  And how could inverter-powered loads ever be more than the power draw from the
    >  battery when the battery is the only source of power?

    I don't think the inverter can count it's own internal losses
    If the discrepancy were due to an internal loss, then the load power would be less than the battery power.  If that's what I was seeing, It would be normal and I would never have created the post.

    But my load power is greater than the battery power.  According to the ComBox, 1.5kwh is appearing out of nowhere and powering my loads.  That would be pretty cool, except I don't believe it.
  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    > @Dave Angelini said:
    > The XW is reporting the amount of energy that it uses to charge the battery to full and the amount that is used by the load.
    >  It has shunts internal to the XW and uses data from the mppt also.  I think you are overthinking this and there are better charts to use than what you posted here for offgrid.
    >
    > If you look at the data at just before sunset on a good day. The Mppt harvest will be equal to the loads during the day, the amount to charge the battery, and the loss in the system.

    I think the answer is here... "shunts internal to the XW". I don't know this gear well, but the Combox looks like a pretty sophisticated (and potentially complex) piece of gear to me. My guess is the data you're pulling doesn't include charge contoller output at it's terminals to the extent that current powers loads. Pulling a different chart/report set would likely pull a register value from the controller, and give you a >0 value for pv.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    Estragon said:
    > @Dave Angelini said:
    > The XW is reporting the amount of energy that it uses to charge the battery to full and the amount that is used by the load.
    >  It has shunts internal to the XW and uses data from the mppt also.  I think you are overthinking this and there are better charts to use than what you posted here for offgrid.
    >
    > If you look at the data at just before sunset on a good day. The Mppt harvest will be equal to the loads during the day, the amount to charge the battery, and the loss in the system.

    I think the answer is here... "shunts internal to the XW". I don't know this gear well, but the Combox looks like a pretty sophisticated (and potentially complex) piece of gear to me. My guess is the data you're pulling doesn't include charge contoller output at it's terminals to the extent that current powers loads. Pulling a different chart/report set would likely pull a register value from the controller, and give you a >0 value for pv.
    Whatever the bug is, it could well be along these lines.  When the combox accounts for solar production, is also shows it 100% as energy going into the batteries.  The same is true for generator power production - it is accounted for 100% as battery charge energy.  Then loads are shown as draw on the batteries.  This is a perfectly reasonable way to do it.

    If under some conditions the Combox isn't counting solar production, that would of course explain the appearance of 1.5kwh.  But it would be a bug just the same.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    edited December 2017 #16
    You should call schneider if you still have warranty. Possibly an internal calibration is off or something in your installation. I have never seen this and none of my clients have ever mentioned this.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    Even if not under warranty, it could well be a bug specific to a particular version of firmware, so you might want to check if there's an update.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    Estragon said:
    Even if not under warranty, it could well be a bug specific to a particular version of firmware, so you might want to check if there's an update.
    It does look like the next step is to call Schneider.  I checked firmware revs the other day and confirmed that everything is current.

    Also, for what it's worth, I looked back a year ago and see the same phenomenon back then.  At that time, I had a 6048 and a different combox.  So this issue has presented itself across two different ComBoxes and two different inverters of slightly different generations.

    In my experience, problems like this often exist for years with nobody noticing or taking the time to raise it as an issue.  I am directly guilty of this myself considering the issue has been around since I first installed a Combox in October of 2016, and I have not taken the time to address it until now.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    I don't see this problem. I have had lots of XW's back to 2007. I have worked closely with the engineers at Schneider, Xantrex and Trace and never heard anything negative except the 3 second speed of the xanbus in reporting changes. It is very accurate however.
    If after you start a service request and strike out I will give them a link to this thread. Good Luck!

    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • ColoradoSolar
    ColoradoSolar Registered Users Posts: 63 ✭✭
    I have a similar issue in that early in the morning when the PV is not producing much the "System Performance" page on the ComBox will show no solar and the loads will be larger than what is coming from the battery. If I go to the "System Devices" page then each individual charge controller will show some production (in the range of a couple hundred watts). So somewhere in the ComBox it is ignoring low solar production for the main page calculations. I don't worry about it because once the sun is up higher and PV production really starts for the day it seems to balance itself out. So I tend to use the data from the ComBox as a general indicator of what is happening in the system and not precise numbers.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Realistically, most measurement systems are going to be, at best, about 1% accurate. If you have a large system (say 10,000 Watts), +/- 100 to 200 watts is within the margin of error. Let alone the 3-5% (or more) losses in the MPPT controllers, inverters, chargers...

    It is possible they are simply "ignoring" these small values in the display system as they are really meaningful.

    In times past, there have been reports that some MPPT charge controllers consistently (for example) reported closter to ~5% more harvest that actually measured with lab grade instruments. In general, it did seem that the Xantrex units were closer to ~1% accuracy for there high end MPPT controllers.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    BB. said:
    Realistically, most measurement systems are going to be, at best, about 1% accurate. If you have a large system (say 10,000 Watts), +/- 100 to 200 watts is within the margin of error. Let alone the 3-5% (or more) losses in the MPPT controllers, inverters, chargers...

    It is possible they are simply "ignoring" these small values in the display system as they are really meaningful.

    In times past, there have been reports that some MPPT charge controllers consistently (for example) reported closter to ~5% more harvest that actually measured with lab grade instruments. In general, it did seem that the Xantrex units were closer to ~1% accuracy for there high end MPPT controllers.

    -Bill

    I agree with you and if the difference was only 1%, or even 5% I wouldn't have bothered to mention it.  But I'm looking at close to 25% error.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    Bill is definitely correct. The error is in the noise. The controllers will often ignore less than 100 watts when they have thousands of watts in the array. I do believe that it is getting added in but I could be wrong as I ignore it.

    So what did you find Twisted Tree? You called Schneider right?
    One of the things I mentioned to you was your installation of the system. You are following good network practice and not subjecting the bus to interference and RF susceptibility. It is a fragile older way to run a network on long distances. Other things too but I will let Schneider speak on that.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    I just sent a note to Schneider and will report back.  1.5kwh error out of a total of 6.5kwh strikes me as more than just noise, but maybe I'm asking too much.

    Re the installation, you are probably referring to the extra long Xbus that I had.  At one point is was being blamed for an AGS bug that I found, but that was indeed a bug and fixed in a FW update to the AGS.  I was as surprised as anyone that it went unnoticed for so long.

    That said, I re-worked my xbus a while back to bring it well within specs.  Instead of locating the AGS at the generator in an out building, I moved the AGS into the house with the rest of the XW gear and set it up to control a set of slave relays at the generator.  So now the xbus won't become a distraction.  And for what it's worth, the ComBox discrepancies are the same both pre and post xBus rework.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    I missed the 1.5kWH vs 6.5kWH error--That is not insignificant... Sort of... Again, it is also time dependent.
    • 10 hours of charging * 200 Watts of "error" = 2,000 WH = 2 kWH of charging (or discharging) error

    I do not know near enough about these systems to make a determination if this is a significant/out of spec. error or not.

    -Bill

    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    Interestingly, it would appear the issue is in the XW6848.  Take a look at the AC load figures below.  They make no sense.

    It looks like they are saying the 240V current is the sum of the L1 and L2 currents?  That makes no sense.

    And 240V times the 240V current doesn't equal 363W.  Now power factor could be at play, but it reports 363 W and 363 VA which says pf=1.



    ParameterValue
    Device NameXW6848-01
    FGA Number865-6848-01
    Unique ID Number399166
    Firmware VersionV02.03.00-0008
    Modbus Address10
    Device PresentActive (data valid)
    Inverter EnabledEnabled
    Charger EnabledEnabled
    Sell EnabledDisabled
    Device StateOperating
    Inverter StatusInvert
    Charger StatusQualifying AC
    Active Faults FlagNo Faults
    Active Warnings FlagNo Warnings
    Battery Voltage52.1 V
    Battery Current-5.6 A
    Battery Power-291 W
    Invert DC Current5.6 A
    Charge DC Current0.0 A
    Invert DC Power291 W
    Charge DC Power0 W
    Battery Temperature23.1 ºC
    Charge Mode StatusPrimary
    Grid AC Voltage0.0 V
    Grid AC Input Current0.0 A
    Grid AC Frequency0.0 Hz
    Grid AC Input Power0 W
    Grid AC Current0.0 A
    Grid AC Power0 W
    Grid AC L1 Voltage0.0 V
    Grid AC L1 Current0.0 A
    Grid AC L2 Voltage0.0 V
    Grid AC L2 Current0.0 A
    Grid AC Input Power - Apparent0 VA
    Grid Output Voltage0.0 V
    Grid Output Current0.0 A
    Grid Output Frequency0.0 Hz
    Grid Output Power0 W
    Grid Output Power - Apparent0 VA
    Grid AC Voltage QualifiedQualifying
    Grid AC Frequency QualifiedQualifying
    Grid AC Qualified Duration0 s
    Generator AC Voltage0.0 V
    Generator AC L1 Voltage0.0 V
    Generator AC L2 Voltage0.0 V
    Generator AC Frequency0.0 Hz
    Generator AC Current0.0 A
    Generator AC Power0 W
    Generator AC L1 Current0.0 A
    Generator AC L2 Current0.0 A
    Generator AC Power - Apparent0 VA
    Generator AC Voltage QualifiedQualifying
    Generator AC Frequency QualifiedQualifying
    Generator AC Qualified Duration0 s
    Load AC Voltage238.3 V
    Load AC Current1.2 A
    Load AC Frequency60.0 Hz
    Load AC Power363 W
    Load AC L1 Voltage119.1 V
    Load AC L1 Current0.4 A
    Load AC L2 Voltage119.1 V
    Load AC L2 Current0.8 A
    Load AC Power - Apparent363 VA
    Auxiliary Output Sta

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Nominally, it is
    L1×V1=P1
    L2×V2=P2

    or roughly
    (L1+L2)×119 volts=Ptotal

    Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    BB. said:
    Nominally, it is
    L1×V1=P1
    L2×V2=P2

    or roughly
    (L1+L2)×119 volts=Ptotal

    Bill
    Yup, I agree.  But that's 145W, not the 363 that the device is reporting.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    There is a very large conversion loss on a 7KW inverter at low power. See what Schneider tells you.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • twistedtree
    twistedtree Solar Expert Posts: 135 ✭✭✭
    There is a very large conversion loss on a 7KW inverter at low power. See what Schneider tells you.
    Understood and agreed.  If the device were reporting 145W to the loads as hand calculated based on the L1 and L2 currents, and the 300W battery load that it accurately reports, I would totally believe it, and all would make sense.  Except that's not what it's reporting.  Conversion LOSSES, I understand.  Conversion GAINS defy physics, and it's reporting GAINS.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,730 ✭✭✭✭✭✭
    See what Schneider says... DC in and DC out + loss are equal in all of the systems I have seen. Maybe yours is somehow making the loss look like a gain. You hurt my head! You did get a reference number for your issue. Without it you are just spinning wheels.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net