Power company not accepting production power

[jandrese]

consumption-hookup.jpegPower does get exported. We saw the credits in the last bill, just not as many as we were expecting.

Here are the pictures of the CTs. The only thing I didn't do is trace which cables goes back to which port on the consumption meter, because they are buried a bit in all of the other wires in the box and take the long way around going all around the outside perimeter of the box before exiting out the top. I've included a top and bottom view as well as a shot of the hookups on the consumption meter itself. If necessary I can trace the wires to verify they are plugged into the correct ports, but if that can't be the issue then I'd prefer to avoid it.

Andy,
Thanks for the clarification about the grid. What I didn't mention is that the hypothesis of jandrese is about what happens on the utility side of the meter and whatever happens on that side of the meter should not affect the meter reading. A more likely scenerio is the meter is not programmed correctly.

My thought was if the consumption meters are just measuring the flux on the wire going out to the meter, then if the meter is choking back the power they'll read it as low flux and assume that it must be self-consumption. My theory depends mostly on the inverter only measuring the indicated voltage of each panel and using that to work out the production figure, not measuring the flux actually going out to the house like the consumption meter does.

Also, thanks to everybody in the thread taking the time to help me out with this.

[Ampster]

I have never heard of a meter having the ability to."choke back power". I would focus on calibration issues between your two sources of data. I actually have several methods of checking the data on my system. Learn how to read your meter and see if the power (kWs) it is registering corresponds with the power net of consumption recorded by your instruments at that same moment. Try it during different times of day.
You may also want to recalibrate your expectations based on real world data not some theory about flux power. Watts law would be a good place to start.

[oregon_phil]

From what I can tell from the photos, jandrese has an energy meter hooked up to measure power to and from the grid. The inverter infers consumption because it knows solar power produced and power to/from the grid.

The photos are fuzzy, but the CT's appear to be the correct part number, in the correct orientation and wired to the energy meter correctly. There is a big wad of black electricians tape where the L1, L2, N wires are connected to the energy meter.

I don't have a SolarEdge system, but if it was me, I would 1) check all connections to make sure they are tight, 2) make sure the RS-485 is wired correctly & terminations are set correctly, and 3) when you think the system is acting up, force your consumption to zero to see what the meter reads. Oh and if there are any firmware updates, that should be investigated too. There are a lot of inverter settings that need to be correct too for communication.

It is possible the energy meter is bad or the assumption on energy consumption are incorrect.

Just my thoughts.

Added thoughts,

jandrese , I just realized that my original warning about working with electricity is on the first page of this thread. Your CT's are upstream of your main breaker and will be live no matter what. If there is any additional electrical work to be done, I think your installer should perform the task. CT's might look harmless, but it is an inductor that might harm you when you make or break connections in the wrong order. Look up "what happens when you unplug an inductor".

After I proved the meter is not performing correctly by forcing my consumption to zero during suspect periods, I would ask the installer to verify the installation.

[jandrese]

Sorry about the quality of the images, the space is pretty tight and there isn't a good place to angle the camera in there. I have been careful not to touch anything inside of the box.

I have gone around the house and verified that no major appliances (water heater, washing machine, dishwasher, refrigerator, hvac, microwave, toaster, stove, oven) are on when the system is showing 10kW of draw in mid-day. Some low draw appliances are still on, including the internet router so the inverter and consumption monitor can update the cloud service.

One more observation: I just had another normal production day, where it seems that all of my excess power went back to the grid as expected. It happened on the day after 2 net-negative days (it was cloudy all weekend).

[oregon_phil]

Moderators: In post 20, I made the mistake of correcting grammar without giving a reason and now it's unapproved. You can delete post 20 since I'm posting the contents below. Thanks and I will remember from now on to always give a reason for editing so the machine is happy.


-----------------------------------------
jandrese ,while this document isn't exactly your system, CT installation guidelines would still apply. "Application Note – HD-Wave Inverters with Built-In RGM & Consumption Monitoring for North America Version History Version 1.0, March 2020 – first version". I suggest you read this document.

Problem: SolarEdge power reporting system doesn't appear to match actual power usage . What possibly could explain your power discrepancies?

1) Equipment malfunction
2) Incorrect installation
3) Data reporting errors
4) SolarEdge power reporting system is correct; actual power usage pattern assumptions are incorrect.

In post #19, you give two qualitative observations. Could you provide pictures of the 10kw day, cloudy day and "normal" production?

In post #16, you show the CT's and comment the wires are buried in the box and go around to the box to the energy meter.

It appears the unshielded CT wires exit the power panel in the same conduit as some AC wires. Is this true? If yes, what AC circuit are those wires attached to?

Read the solaredge document above. A better practice would be to run the CT wires out of the electrical box in a knockout by itself without AC wires.

[bcroe]

Moderators: In post 20, I made the mistake of
correcting grammar without giving a reason and now it's unapproved.

My experience here is that immediately going back to correct those things
will put it into moderation. So I just leave it till the next day, though maybe
a couple hours or even less is enough to avoid moderation approval.
Bruce Roe

[jandrese]

The wires from the CTs are the black and white ones twisted together. As you can see from the full frame shot they are routed out to the left, then go down to the bottom around to the right, and back up to the exit on the top of the box. The exit is shared with the power feed for the consumption meter itself (wires from that run into a pair of breakers on the left side of the box).

full-box.jpeg

[jandrese]

Here are some more examples of my listed solar output. Here are my thoughts on each day:

• 5/26: Mostly sunny day, consumption tracks production closely and is much higher than expected
• 5/27: Another mostly sunny day, consumption tracks production closely and is much higher than expected
• 5/28: Mostly cloudy, consumption tracks production again, save for a little bit in the evening. My son apparently discovers that you can leave Twitch streaming all day to earn some kind of twitchbucks leading to high power usage for a few days.
• 5/29: Cloudy, A little hard to say on this one because the base load is so high, but I think production was working normally.
• 5/30: Cloudy, Again it seems to be working normally
• 5/31: Sunny, production is working normally.
• 6/1: Mostly sunny, production is working normally.

If we get a couple more sunny days I'm expecting to go back to the behavior seen on 5/26.

[solardreamer]

consumption-hookup.jpeg

That consumption meter looks like it has 3 CT connections and you only have 2 CT's connected. Are you sure it's not a 3 phase consumption meter? If it is then the unconnected CT input is floating and could be causing the strangeness you have observed. Some meters can be configured to use only 2 of the 3 inputs so you want want to check that.

[jandrese]

The strange thing is the third terminal jack is labelled "N/A" where the other two are labelled with the polarity. I had thought maybe the company used the same hardware for all of their consumption meters and adjusted the internals depending on the model.

[oregon_phil]

jandrese . Even when you think the system is operating normally, I don't believe it is operating normally. I have highlighted a section of your May 29th production/consumption.

My interpretation of SolarEdge graph at the highlighted point in time.

Solar production = Green

If all solar production is consumed, then green solar production curve becomes blue (self consumption).

If consumption is larger than solar production, then graph would stack red over blue (see second image that I pulled from google).

For the point highlighted in the attachment, how can the graph show green, blue and red? approximately Red = 4.2kw, green = 3.2kw and blue = 2.3kw.

In other words, there should be no green showing because all of the solar power should be self consumption.

I think your installer needs to verify the installation. There is that big wad of black electricians tape that has the power wires under it..

Maybe the power reported can give us a clue on the solution, but it isn't coming to me right now.

[oregon_phil]

The RS485 communication cable is suppose to be shielded.

The AC connections under the black electricians tape starting from the connection closest to you to the wall is : No Connect, L2, L1, Neutral, Ground.

In the three phase model of the consumption meter, the No Connect AC terminal is L3 and the N/A terminal is for CT3.

If the AC wires are not connected correctly, you could have problems. Neutral and ground wires should be traced to make sure they are actually connected.

[jandrese]

I don't think it is necessarily an indication of a problem to see both production and consumption in the same sample. The meter aggregates data in 15 minute chucks before uploading to the cloud service, so it is possible to have say 8 minutes of production and 7 minutes of consumption in a single sample.

At any rate, I have scheduled an appointment with my installer to come out tomorrow and verify all of the hookups. I will update this thread with any findings for future reference.

[jandrese]

Update: The solar contractor came back and verified that all components of the system were installed correctly. He also called SolarEdge and had them verify the configuration of both the inverter and consumption meter. No configuration issues were found. In the end they decided to try replacing the consumption meter, but none are currently in stock so it will be a couple of weeks before we know if that solves the problem.

[oregon_phil]

Thanks for the update. I'm curious, I know the installer confirmed the installation was correct. Did they unwrap the black electricians tape for the L1, L2, N and Ground connections to the energy meter?