Everything Solar Forum

communities_1.jpg

 View Only
Expand all | Collapse all

Challenging Issue

  • 1.  Challenging Issue

    Posted 11-26-2021 11:45 AM
    Hi all,

    I've been dealing with a measurement issue for a couple months that I've been unable to determine root cause of.  I have a unique set up as I have 2 - 200 amp panels (total of 400 Amps), feeds from electric company are 200 amps.  I have a SolarEdge StorEdge single phase inverter with 8.9 kWh of panels behind it with a LG 10Kwh battery.  I have a SolarEdge meter with CT's to both panels hooked up in parallel to the meter.  The meter is correctly connected to the power system and CT's but when measurement calculations are done it is essentially backwards.  The SolarEdge app shows export to the grid when it should be import from the grid and draw from the grid is shown but summed with the production and shown as consumption.


    The only number that is right above is the production number.  The import from the grid should be export to it and consumption should be the difference between production and export so in the example above the house should be consuming ~1.28Kw.  

    So, the question is what setting is wrong or are the CT's facing the wrong direction?  Right now the settings have the meter as Export/Import but I wonder whether it should be set as a Consumption meter.  The CT's are facing the grid and are connected on the feeds to each panel.  A complicating factor is that I am not sure that L1 in both panels is truly L1 or whether they got swapped during wiring of the house (I'm not the original owner) and I'm not sure how you validate that the L1 side of each panel is the same from the utility.  If they are indeed swapped it should be as easy as switching the CT's in one box.

    Now you are probably wondering how I got to this point after 3 years of operation and that is a valid question.  My utility only read once a month on a variable schedule and I could get daily readings from Solaredge so I took the consumption and export readings from the utility and the production numbers from SolarEdge.  When the utility swapped the meter I could get daily readings and in the process of reconciling the two I found very high variance both high and low and after quite a bit of investigation discovered numerous issues in wiring the meter and connecting the CT's in the two panels together that are now corrected but the math the inverter is doing is wrong but I suspect that is because the configuration is wrong.

    Some of you would probably say "make your installer fix it" but my installer is 200 miles away so its not as simple as saying "fix it", they are primarily in the commercial side and no longer operate a residential business and are willing to help but like many things, if I can figure it out and fix it on my own everything would be easier.

    Any idea's are welcome.  My next step is to change the meter to a consumption meter in the inverter settings and see what happens.  I am running out of ideas….

    Regards,

    John

    ------------------------------
    John Gillispie
    AARP
    johnp250@yahoo.com
    ------------------------------


  • 2.  RE: Challenging Issue

    Silver
    Contributor
    Posted 11-29-2021 11:03 AM
    Hi John,

    Thank you so much for raising this question! I am hoping @John Burke, the ASES PV Chair, may be able to offer some guidance on this.

    Thank you,
    Ella




    ------------------------------
    Ella Nielsen
    Membership & Online Community Intern
    American Solar Energy Society
    Boulder CO
    enielsen@ases.org
    ------------------------------



  • 3.  RE: Challenging Issue

    ASES Life Member
    Posted 11-29-2021 11:08 AM
    How do you know that your house load isn't really 11.3kW?

    ------------------------------
    Mike Curran
    Retired from (dare I say it?) Fossil Plant Mgt.
    joacchim57@gmail.com
    ------------------------------



  • 4.  RE: Challenging Issue

    Posted 11-29-2021 06:44 PM

    Schematic attached below.

    Although in the winter the house can draw 11Kwh, not this time of year.  Right now it appears to me that the math in the inverter is wrong.

    There is one 400Amp feed separated in the meter setting feeding to the two panels.

    Yes, there is only solar feed to one panel.

    House was originally constructed as a main house and attached apartment with separate HVAC and Hot Water with separate panels for each.   When we remodeled we put in zoned geothermal to eliminate one HVAC system.  Apartment hot water replaced with tankless system for west end of house.  



    ------------------------------
    John Gillispie

    johnp250@yahoo.com
    ------------------------------



  • 5.  RE: Challenging Issue

    Silver
    Contributor
    Posted 11-29-2021 12:24 PM
    First a full schematic would be nice. However in lieu of that, how many utility pole transformers or slab mounts do you have considering two main panels?
    Second, I assume the solar connections are on only one main panel and not two?
    Why two main panels?
    Is all the house load on only one main panel? Is it the same panel as the solar.
    Hard to trouble shoot without all details.

    ------------------------------
    william fitch
    Owner
    www.WeAreSolar.com
    fcfcfc@ptd.net
    ------------------------------



  • 6.  RE: Challenging Issue

    Silver
    Contributor
    Posted 12-01-2021 05:27 AM
    Edited by william fitch 12-01-2021 01:25 PM
    Thanks for the Sche. My god, that is a confusing drawing. The whole two panel thing (Two panels in parallel) makes something relatively simple into complex. Current in a conductor is everywhere the same. Lets say you were gening 5 amps to panel 1 and 5 amps to panel 2. Any current measuring device I have ever seen (Direct, shunt or magnetic field) will show 5 amps on meter 1 and 5 amps on meter 2. If you put them in series on a single line, both will still show 5 amps each, but the 5 on the two meters will reflect the single 5 amps on the single line. To me from what I can see in the drawing, I would take panel 2 out of the equation, use 1 CT on panel 1 and run everything through there. Its not like you are going to overload a 200 amp panel (48KW max peak). Any production goes AGAINST GRID LOAD, which reduces the panel stress anyway. And when determining RE total load via the 20% rule, you have up to 240 amps on a 200 amp panel for bus bar load.

    EDIT: To be more specific regarding panel 2, since you have only one utility meter anyway, disconnect number 2 panel from the meter base. Take a #2 awg copper two wires and a #4 copper green (GRND) and run from panel 2's main feed lugs (L1 & L2 & GRND) to a 100A 240vac breaker in panel #1. In short, convert panel number 2 into a 100A subpanel off panel #1.
    Then rewire the solar for a normal single panel install. Piece of cake.

    ------------------------------
    william fitch
    Owner
    www.WeAreSolar.com
    fcfcfc@ptd.net
    ------------------------------



  • 7.  RE: Challenging Issue

    Posted 11-29-2021 08:37 PM
      |   view attached
    Some points to check:
    • All L1 and L2 measurements, CTs and voltage sense, must be truly be on their respective line.  One check is to measure the AC voltage from the L1 terminal in the inverter to the buss where you installed the L1 CT to measure current close to the meter feed.  Should be close to zero, if 240V you have swapped lines somewhere, inspect where you landed the L1 on the circuit breaker, it is easy to assume the top pole is connected to the left buss. 
    • The proper orientation of the CTs and polarity (white & black wires) is important.  
    • Sounds like you have a service entrance that supports two main breakers on one meter.  The CTs need to be located on the buss from the meter in order to measure the same power as the meter. The standard included CTs may not fit this buss and you may need larger or flexible CTs.
    • Each energy meter has an address on the RS-485 buss that needs to be in accordance with the inverter setup.
    SolarEdge has a se-energy-meter-with-modbus-connection-manual-na.pdf 2-page document that is helpful.
    SolarEdge also has a series of on-line courses, set up a free account at https://accounts.solaredge.com/solaredge-webapp/
    Your installer should be able to access the inverter setup remotely, but I have not tried this.

    ------------------------------
    William Kaszeta
    President
    Photovoltaic Resources International
    Mesa AZ
    bill@kaszeta.org
    ------------------------------



  • 8.  RE: Challenging Issue

    Silver
    Contributor
    Posted 11-29-2021 09:48 PM
    Hi John,

    John Burke emailed me and had the following answer for you:

    "The 'rated' out-put of any PV module is only limited by the 'latitude' of the installation site, the weather conditions, as well as the time of day! These factors will determine the 'actual' output of your PV array ... not the 'rated' output, (which is how the modules are 'sold' ... based on 'Air Mass One' ... Sahara Desert, high noon"

    Thank you so much @John Burke!
    Ella​

    ------------------------------
    Ella Nielsen
    Membership & Online Community Intern
    American Solar Energy Society
    Boulder CO
    enielsen@ases.org
    ------------------------------



  • 9.  RE: Challenging Issue

    ASES Life Member
    Posted 11-30-2021 01:24 PM
    Hi John

    Sounds like your CTs are installed backwards. Double check them.


    ------------------------------
    Robert Foster
    Assistant Professor
    New Mexico State University, College of Engineering
    Las Cruces NM
    rfoster@ases.org
    ------------------------------



  • 10.  RE: Challenging Issue

    Posted 11-30-2021 07:33 PM
    There are two sets of CT's, one for each of the breaker boxes and are wired to the meter in parallel.  The CT's are pointed at the grid per the arrow on the inside of the CT openings.

    The local utility will not allow any customer equipment in the meter setting as the installer wanted to put them there.  Installer asked for options from Solaredge and they said to put the CT's in both boxes and wire them in parallel.  Utility has tested their meter and its within .0010-.0015 on both legs both import and export.

    I need a new volt meter to get measurements.  How do I check that L1 is the same in both panels?  L2?  I've suspected that the feeds got swapped in the boxes and that could cause anomalies.

    I've read the manual (https://www.solaredge.com/sites/default/files/se_energy_meter_with_modbus_installation_guide-na.pdf) and I am not 100% sure whether my meter would be defined as Load Connection Point (Consumption?) or a Grid Connection Point meter (Export/Import?) (see page appendix C on page 42 in the manual).   Right now the inverter thinks the meter is a Export/Import. 

    ------------------------------
    John Gillispie
    AARP
    johnp250@yahoo.com
    ------------------------------



  • 11.  RE: Challenging Issue

    Posted 11-30-2021 10:56 PM
    John,
    Your panel CTs need to be in series. That way currents in each panel add to each other. The two L1 CTs are in series, etc.  If they are in parallel you will get the average, not the sum.

    To check that  L1 is the same in both panels, simply measure the voltage between the two L1 busses, should be close to zero.  If you get something about 240Vac they are not on the same line. They need to match L1 in the inverter.

    CAUTION: You do not want to disconnect the CT leads and try to measure the voltage, it can be very high without loading resistor that is internal to the energy meter.

    These CTs that are on Panel 1 and Panel 2 are Grid Connection Point meter (Export/Import)

    Your drawing does not show any circuit breakers or disconnecting switches.  If your "Utility feed & Meter" includes two 200A circuit breakers (so that you can disconnect the cables to the panels safely) and if the CT hole is large enough for two of the wires (3/0 or 4/0 CU?), you can simply route both L1 wires through one CT.

    I hope this helps.

    Bill


    ------------------------------
    William Kaszeta
    President
    Photovoltaic Resources International
    Mesa AZ
    bill@kaszeta.org
    ------------------------------



  • 12.  RE: Challenging Issue

    Posted 12-01-2021 05:26 PM
    No, the CT's should not be in series. To combine readings of two CTs measuring current in two different conductors, you put them in parallel as shown in the link here, from the manufacturer of the SolarEdge CT's that John has in his system:

    https://ctlsys.com/support/paralleling_current_transformers/

    ------------------------------
    Ken Nadsady
    AviSun Renewable Energy
    Hudson OH
    Knadsady@avisun.net
    ------------------------------



  • 13.  RE: Challenging Issue

    Posted 12-01-2021 08:12 PM
    Attached is the picture of the meter with the wiring I recently corrected.   I am pretty confident that the wires are connected correctly but will check tomorrow.  There are 4 CT's, 2 for each 200A box.  The breaker panels do not adjoin each other so I had to run about 10 feet of wire to combine the CT's.  It's twisted 18g wire.  One of my concerns is that the main feeds are swapped in the boxes.

    For an electrician out there, please help clarify something for me.  When documentation talks about L1 and L2, I have been assuming that is the left and right sides of the main feeds to the panels themselves.  See attached picture.  I've also been assuming that when the meter calls for connection to L1 and L2 (the meter is 240V) I ran wires from the left and right banks of the breakers from one that had 240A wired.  Clarification appreciated.

    I've ordered a new volt meter and should get it in a couple days to check that L1 is indeed L1 and not swapped in the boxes.  Will also check whether L1 from the inverter is correct.  There is an emergency shut off between the inverter and the panel that AC feeds into not included in the schematic.  An earlier questions was whether there were 200A breakers for each panel.  Yes, each panel has a 200A breaker.

    I appreciate everyone's thoughts, ideas. and clarifications. I agree with whoever said it's complicated, the whole system is complicated.  Where I live we have no code or local inspections for residential, as such, I am leery of everything about the panels, grounding, feeds from the utility, etc. 

    The pictures were taken at different times so as I've been debugging and wire chasing. things changed.  The picture of the meter is the latest configuration, the picture of the panel that has the meter in it is older.  For those interested, pics of the system are attached that can matched to the drawing.




    ------------------------------
    John Gillispie
    johnp250@yahoo.com
    ------------------------------



  • 14.  RE: Challenging Issue

    Posted 12-01-2021 08:27 PM
    Looks like I made a mistake, the CTs are in parallel if they include a burden resistor.  Good reference by Ken.

    ------------------------------
    William Kaszeta
    President
    Photovoltaic Resources International
    Mesa AZ
    bill@kaszeta.org
    ------------------------------



  • 15.  RE: Challenging Issue

    Posted 12-01-2021 05:34 PM
    As I wrote elsewhere in this thread (sorry I did not read this specific post before replying), The CT's orientation looks correct, pointing to grid, the meter configuration as Import/Export is correct when monitoring the service conductors so leave that the same, and combining two CT's that measure L1 from two panels is correctly done by connecting the output of the two CT's in parallel.  The issue seems to be that you may have the leads from the CT's reversed at the meter terminals.  Black needs to go to the blak dot terminal, white needs to go to white. Check that wiring and reverse at the meter if needed and it will fix your issue.

    ------------------------------
    Ken Nadsady
    AviSun Renewable Energy
    Hudson OH
    Knadsady@avisun.net
    ------------------------------



  • 16.  RE: Challenging Issue

    Posted 12-01-2021 05:08 PM
    Edited by Ken Nadsady 12-01-2021 05:17 PM
    Hi John,

    It looks like from the photo the twisted pair leads from the CTs were extended by splicing additional wire onto them.  In the photo of the CT we see the black and white wires they come with from the factory. In the photo of the meter, it shows red and white wires terminated at the CT terminals.  If the red is spliced to black, and white spliced to white, then you have the CT leads reversed on the meter. The label shows where the white and black wire go in each CT input, and reversing those connections will result in export and import measurements being transposed. Flip the black and white leads at the meter if they are not correct and leave the CT orientation as is.

    Also, how many CTs are installed?  Are there four CTs?  Two CTs in Panel 1, one for L1 and one for L2?  and Two in Panel 2, one for L1, and one for L2?  Just trying to make sure you are measuring both L1 and L2 current in both 200A panels, because you cannot assume L1 = L2 balanced, as some unbalanced current will flow in the Neutrals back to the utility.

    And, leave the meter configured for Import/Export, and don't change it to consumption.  This configuration tells the inverter/monitoring system what point your CT's are measuring. With CT's on the serice conductors going to the utility, you are measuring Import/Export.  In order to measure Consumption, you would need to isolate the solar connection out of the mix, i.e. if you had a supply side solar connection and put the CT's between that connection point and the breaker panel, you would be measuring consumption. But that is not what you have, so leave it at Import/Export.

    Sunny Regards,
    Ken

    ------------------------------
    Ken Nadsady
    AviSun Renewable Energy
    Hudson OH
    Knadsady@avisun.net
    ------------------------------



  • 17.  RE: Challenging Issue

    Posted 12-08-2021 09:07 AM
    Hello all,

    Thanks for your help thus far.  Sorry I've been busy with the holidays.  I have made progress but am still off 10-15% on both import and export between what the meter says and what the utility says.  The issue I found was that the feeds from the panel to the (SolarEdge) meter were both on the same leg.  When I corrected that I started getting measurement that made more sense but is still off.  I did get a Volt meter that is functional and I'm trying to understand how to measure a couple things.

    Using the volt meter how do I check two things? 

    First that the L1 and L2 off the panel to the Solaredge meter are correct? 

    Second, how to check L1 and L2 are the same in both panels (I've suspected for some time that L1 and L2  were swapped in the panels(ie L1 in panel 1 is L2 in Panel 2))?

    Thanks in advance.

    ------------------------------
    John Gillispie
    AARP
    johnp250@yahoo.com
    ------------------------------



  • 18.  RE: Challenging Issue

    ASES Life Member
    Posted 12-08-2021 10:47 AM
    Don't know about 1st question, but for second, measure voltage between the two L1's.  If it's zero volts they are the same.  If it's 240 volts they are swapped.  If swapped, you would get zero volts measuring from L1 in one panel to L2 in the other panel.  Hope this makes sense.

    ------------------------------
    Mike Curran
    Retired from (dare I say it?) Fossil Plant Mgt.
    joacchim57@gmail.com
    ------------------------------



  • 19.  RE: Challenging Issue

    Posted 12-08-2021 03:42 PM
    John,

    Part of the problem I am having in helping you is that the photos you attached do not clearly show the wiring at the meter.  Wires and wire nuts are blocking the view, and in general, the wiring in that area is very messy.  I think it would be helpful for you to neaten up the wiring a little bit.  Once you identify what you are calling L1, wrap some red electrical tape around the conductor.
    The Solaredge meter needs the CT's that are meaasuring L1 in panel 1 and L1 in panel 2 connected in parallel. So the white leads of the two CT's measuring L1 should be spliced with a wire nut and an extension wire that is whit.  That white extension wire must go to the white terminal for L1 on the meter. Similarly, the black wires for L1 and the red extension wire (since your twisted pair extension wired does not have black) should go to the black terminal for L1 on the meter. Do the same for the L2 CT's.  Also, try and keep the twist in the wires, because that helps cancel out interfering signals that may couple onto the wire from nearby circuits.
    The SolarEdge meter needs to get the voltage input for L1 and L2 from a 2-pole 240V breaker like you described. I can't see the wire colors at the meter, but try and keep the wire colors consistent with your other L1 and L2 marking. Red for L1 and black for L2. You can use the same technique that Mike suggested to check the L1 and L2 wiring at the meter. Put one lead of your voltmeter on L1 at the main breaker terminal, and the other lead on the wire feeding L1 of the SolarEdge meter. If it is zero volts then both are L1. If it is 240V then the wire at the meter is actually connected to L2, so flip those leads at the SolarEdge meter.
    Just be careful. 240V is dangerous. Take your time and be wary of where live wires are.

    ------------------------------
    Ken Nadsady
    AviSun Renewable Energy
    Hudson OH
    Knadsady@avisun.net
    ------------------------------



  • 20.  RE: Challenging Issue

    Posted 12-12-2021 03:41 PM
    Thanks to everyone for assistance.  I've attached better pictures of the CT's and meter.  This afternoon I got a chance to validate a number of things including L1 and L2 on both panels.  I used the voltmeter to check between the two panels and indeed L1 is L1 and L2 is L2.  Validated the high voltage feeds to the meter were correct by checking between the breaker they are attached to and the main panel feeds.  Double checked that the CT's are properly wired and that also checked out.

    I am seeing a difference of ~20% on the import side (Solar Edge is under measuring) and around 17% on the export side (again, Solar Edge is undermeasuring).  I don't get daily readings from the Utility so I've been reading it manually and comparing to Solar Edge.  I can see a couple % difference but 20% is outside of what I would expect.

    As I thought about this, the emergency panel has a couple high wattage appliances (side by side refrigerator and an upright freezer on it along with TV/DVR/Cable modem/router/etc) on it.  There is no measurement on the emergency box but 20% for those things seems high when I consider that the whole house is electric and I charge a PHEV (Jeep Wrangler) as well.

    There are two potential issues for the measurement difference but I have no idea whether either one matters.  First, without measurement off the emergency panel that consumption is missing but that does not explain the export side, or the wire I've run between the boxes is introducing error.  You can see the difference between the wire on the CT's and the extension wire in one of the pictures.  I got the wire from this web site https://www.invertersupply.com









    ------------------------------
    John Gillispie
    AARP
    johnp250@yahoo.com
    ------------------------------