I received a visit from my utility company today asking me to disconnect my solar panels system. I told him it was impossible because my system was not designed to export back to grid, it just help charge battery and support for my loads, because my system is off-grid. He told me my smart meter was giving an error, he was courteous indicating that this is just a request that but because I don't have a contract agreement with them they will lock my meter until I disconnect. They don't want to get that way because right now is extremely hot here in Texas, but they will if I don't comply.
Can anyone confirm that the error code D7 is related to feedback to grid?
Anyways, as soon as I put the breaker that supplies electricity to my growatt as off the error code disappeared.
I am contacting the vendor (watts247.com) about this issue, or my growatt is defective or it just not designed correctly (at least on the internal programming). I am guessing growatt shares the same hardware between offgrid and grid tie inverters.
Edit days later:
I completed my investigation and have more data from my growatt records and from my emporia VUE that records my consumptions at main panel. Thanks to @hwy17 for the definition of the D7 error code that means one of the phases the current was inverted.
The conditions when it happened could be seen on the following graph
Being SUB, I had the system imported from grid to support my loads at night, and when the sun rises solar starts taking over. On previous days was not an issue, because battery was not fully charged and grid was always being imported a little bit to help charge it but the afternoon before the battery became fully charged and no more grid assistance to charge it was needed. The AC input to the device is connected to a dedicated socket wired to a 40A breaker from main panel. The loads are on a transfer switch, feed by the growatt. I still maintain big loads such as AC, stove, dryer, and a couple of small circuits on the grid main panel, but the 3000W from the device are enough to cover most of my small loads. The biggest load is just 1200W from an electric grill. I am on a 240V panel from grid, were 2 lines feed the main panel, and each of them become 120v when neutral is used. The stars were aligned at 11:10am, so on the same main panel circuit/line were I have the growatt none of the other remained loads at my hose were ON, so the only consumption was the growatt, and at that specific time morning shades freed my panels so it started to produce more power to fully cover my loads, so importing from grid was not needed so it went to zero, and probably beyond that.
Accuracy on the VUE is not that perfect, but gives you in idea. Main_A and Main_B are the 2 lines coming from the utility smart meter, you can see that at 11:21am it was negative Main_B... but probably the others small numbers were considered negative by the smart meter. The hall sensors used on those 2 wires Main_A and B are read bidirecionally, but the hall sensors for each of the circuits are only setup for 1 direction. So those red numbers on the "Growatt" circuit is not possible to confirm if positive or negative. Anyways, the Utility guy arrived to my home around 11:20AM, may be he is a field technician already in my neighborhood when the meter alerted them. When you call them it takes them hours to arrive, but if someone is feeding without a contract, it is just a couple of minutes to show up. I believe the error does not stay permanently once activated, because as soon as I turned off the breaker for the growatt, I looked at the meter and the error was not longer there. So I believe my device was backfeeding enough time to get their attention. Be careful, those are assumptions since I don't know who their system works.
Before you asked, circuit "Light-luz patio" was consuming at that time, but it was consuming from Main_A.
The following table shows the data coming from the growatt, and how it was PV&Grid Charging+GridBypass at the time the back feed happened.
It shows how the battery was not being used to support any loads until I switched off the breaker for the growatt.
As some had already mentioned here, these devices supported the loads at the same time grid is used. In order to do that they are basically matching the grid frequency from the grid with the inverter, and tie them together. And if their circuits are not faster, or accurately sensing currents, when the grid import is transitioned out, there is the possibility of backfeeding through the AC input if the cut off is not fast enough.
edit 6/5:
More testings performed today to continue the investigation. I was brave enough to connect the AC input again. Before that, I moved 2 circuits with constant charge to the same line the growatt breaker is installed, the fridge and modem/routers. Modems/routers are around 30 watts constant, while fridge is around 11 watts without compressor on. Here are the results:
Again there was a back feed at one point, in 1 second it flowed back around 220 watts. It happens at the point where the Grid import on the growatt was reduced 500w to 70W, which also correspond to a ~1000w load (dishwasher) was getting off. Growatt data does not have that good resolution so I can be sure.
I went out to see if there was any error from the smart meter, and fortunately it was not. Since the screen from it is rotating different information, one of them said "Test", well actually all LED are on, so I am not sure if that is a test or not.
but it lasted for about 5 seconds. So it is possible that if your backflow is in around 1 seconds, it will not trigger the error since the test sampling time of the meter is not that fast. May be the constant loads helped to maintain that grid line wattage farther than zero, so the meter did not considered that blimp worth of the Error D7. This situation requires more testing, I mean it is not economical to keep a 300W constant load if that is basically the saving I am getting from the system.
By the way, 1 hour after that 1 second backflow and nobody had knocked the door yet.
Can anyone confirm that the error code D7 is related to feedback to grid?
Anyways, as soon as I put the breaker that supplies electricity to my growatt as off the error code disappeared.
I am contacting the vendor (watts247.com) about this issue, or my growatt is defective or it just not designed correctly (at least on the internal programming). I am guessing growatt shares the same hardware between offgrid and grid tie inverters.
Edit days later:
I completed my investigation and have more data from my growatt records and from my emporia VUE that records my consumptions at main panel. Thanks to @hwy17 for the definition of the D7 error code that means one of the phases the current was inverted.
The conditions when it happened could be seen on the following graph
Being SUB, I had the system imported from grid to support my loads at night, and when the sun rises solar starts taking over. On previous days was not an issue, because battery was not fully charged and grid was always being imported a little bit to help charge it but the afternoon before the battery became fully charged and no more grid assistance to charge it was needed. The AC input to the device is connected to a dedicated socket wired to a 40A breaker from main panel. The loads are on a transfer switch, feed by the growatt. I still maintain big loads such as AC, stove, dryer, and a couple of small circuits on the grid main panel, but the 3000W from the device are enough to cover most of my small loads. The biggest load is just 1200W from an electric grill. I am on a 240V panel from grid, were 2 lines feed the main panel, and each of them become 120v when neutral is used. The stars were aligned at 11:10am, so on the same main panel circuit/line were I have the growatt none of the other remained loads at my hose were ON, so the only consumption was the growatt, and at that specific time morning shades freed my panels so it started to produce more power to fully cover my loads, so importing from grid was not needed so it went to zero, and probably beyond that.
Accuracy on the VUE is not that perfect, but gives you in idea. Main_A and Main_B are the 2 lines coming from the utility smart meter, you can see that at 11:21am it was negative Main_B... but probably the others small numbers were considered negative by the smart meter. The hall sensors used on those 2 wires Main_A and B are read bidirecionally, but the hall sensors for each of the circuits are only setup for 1 direction. So those red numbers on the "Growatt" circuit is not possible to confirm if positive or negative. Anyways, the Utility guy arrived to my home around 11:20AM, may be he is a field technician already in my neighborhood when the meter alerted them. When you call them it takes them hours to arrive, but if someone is feeding without a contract, it is just a couple of minutes to show up. I believe the error does not stay permanently once activated, because as soon as I turned off the breaker for the growatt, I looked at the meter and the error was not longer there. So I believe my device was backfeeding enough time to get their attention. Be careful, those are assumptions since I don't know who their system works.
Before you asked, circuit "Light-luz patio" was consuming at that time, but it was consuming from Main_A.
The following table shows the data coming from the growatt, and how it was PV&Grid Charging+GridBypass at the time the back feed happened.
It shows how the battery was not being used to support any loads until I switched off the breaker for the growatt.
As some had already mentioned here, these devices supported the loads at the same time grid is used. In order to do that they are basically matching the grid frequency from the grid with the inverter, and tie them together. And if their circuits are not faster, or accurately sensing currents, when the grid import is transitioned out, there is the possibility of backfeeding through the AC input if the cut off is not fast enough.
edit 6/5:
More testings performed today to continue the investigation. I was brave enough to connect the AC input again. Before that, I moved 2 circuits with constant charge to the same line the growatt breaker is installed, the fridge and modem/routers. Modems/routers are around 30 watts constant, while fridge is around 11 watts without compressor on. Here are the results:
Again there was a back feed at one point, in 1 second it flowed back around 220 watts. It happens at the point where the Grid import on the growatt was reduced 500w to 70W, which also correspond to a ~1000w load (dishwasher) was getting off. Growatt data does not have that good resolution so I can be sure.
I went out to see if there was any error from the smart meter, and fortunately it was not. Since the screen from it is rotating different information, one of them said "Test", well actually all LED are on, so I am not sure if that is a test or not.
but it lasted for about 5 seconds. So it is possible that if your backflow is in around 1 seconds, it will not trigger the error since the test sampling time of the meter is not that fast. May be the constant loads helped to maintain that grid line wattage farther than zero, so the meter did not considered that blimp worth of the Error D7. This situation requires more testing, I mean it is not economical to keep a 300W constant load if that is basically the saving I am getting from the system.
By the way, 1 hour after that 1 second backflow and nobody had knocked the door yet.
Last edited:
