Enphase Solar and Powerwall integration

[Will792]

I also have enphase micros (and an enphase envoy) feeding Tesla gateway/Powerwalls.

I’m curious if your Tesla CT and Enphase CT measure the same energy/power. In my experience Tesla will report 10% more solar generation compared to enphase.

For example on a given day, Tesla could say I generate 33 kWh of power.

When I log into Enphase / enlighten it will say my solar production is 30 kWh.

I’m just curious if you also see reporting discrepancy between the two ecosystems.

In my case Tesla reports PV production that is typically 5% higher than Enphase readings and consumption that is 10% higher than Enphase. Unfortunately my grid voltage is a bit too hot. Most of the time it is 248V-249V. I speculate that Tesla calculations include voltage in calculations while Enphase is using 240V nominal voltage.

I also have GE RGM for PV production. One day I will compare aggregate RGM power difference for one day to figure out if Tesla or Enphase readings are closer to reality.

 

[miimura]

I have M250 Enphase micros from 2012. There are no Enphase CTs in my system because it only measures production. The Enphase system collects data directly from the inverters over their powerline communication system with the Envoy Ethernet connected device. There is a difference between the Tesla measured production and the Enphase reported production but it's not consistently off in one direction.

 

[charlesj]

I have M250 Enphase micros from 2012. There are no Enphase CTs in my system because it only measures production. The Enphase system collects data directly from the inverters over their powerline communication system with the Envoy Ethernet connected device. There is a difference between the Tesla measured production and the Enphase reported production but it's not consistently off in one direction.

I have the M215 and no problems as you describe. My daily difference between the two is less than 0.5%.
I can read any and all my panels performances which is great as I had on inverted that went belly up. Without this knowledge Who knows when it would have been replaced, if at all.
System works great with the two Tesla PWs.

 

[miimura]

I have the M215 and no problems as you describe. My daily difference between the two is less than 0.5%.
I can read any and all my panels performances which is great as I had on inverted that went belly up. Without this knowledge Who knows when it would have been replaced, if at all.
System works great with the two Tesla PWs.

My system works great too. Per-panel output and production is very good for identifying issues. I can even see Winter shadows move across my array. The fact that the numbers don't match is an extremely minor issue. That's why I didn't even mention the actual size of the difference that I saw.

 

[aus1ander]

I also have enphase micros (and an enphase envoy) feeding Tesla gateway/Powerwalls.

I’m curious if your Tesla CT and Enphase CT measure the same energy/power. In my experience Tesla will report 10% more solar generation compared to enphase.

For example on a given day, Tesla could say I generate 33 kWh of power.

When I log into Enphase / enlighten it will say my solar production is 30 kWh.

I’m just curious if you also see reporting discrepancy between the two ecosystems.

Did you ever determine the cause of this overreported solar and could your installer correct it? My Enphase system suffers the same problem with Tesla reporting 7% more solar than I actually produced.

 

[holeydonut]

Did you ever determine the cause of this overreported solar and could your installer correct it? My Enphase system suffers the same problem with Tesla reporting 7% more solar than I actually produced.

Nope, for all my own data purposes, I apply a knock-down factor to anything Tesla reports (whether it's solar generation, home loads, or exports/imports to the grid). This is the only way I can align Tesla data with data from the PoCo, my Emporia Vue, or Enphase.

 

[BGbreeder]

Did you ever determine the cause of this overreported solar and could your installer correct it? My Enphase system suffers the same problem with Tesla reporting 7% more solar than I actually produced.

I've monitored the Enphase reporting on our system and it was 100.% accurate compared to the utility meter. Frankly, it is the solar production that I really care about. Home use, Powerwall status, are both order of magnitude concerns for me. Solar sets the energy budget.

And yes, neither the grid meter nor Enphase reporting agrees with my Tesla reporting either.

 

[aus1ander]

Nope, for all my own data purposes, I apply a knock-down factor to anything Tesla reports (whether it's solar generation, home loads, or exports/imports to the grid). This is the only way I can align Tesla data with data from the PoCo, my Emporia Vue, or Enphase.

Interesting thanks. I find the grid import/export CT is very accurate and aligned well with my utility meter. Its just the way Tesla measures solar on these Enphase systems that's messed up and I don't have a good explanation as to why! I too just depend on my Enlighten data for the most accurate solar numbers and adjust down my monthly consumption by the amount of overreported solar Tesla shows.

 

[BGbreeder]

Interesting thanks. I find the grid import/export CT is very accurate and aligned well with my utility meter. Its just the way Tesla measures solar on these Enphase systems that's messed up and I don't have a good explanation as to why! I too just depend on my Enlighten data for the most accurate solar numbers and adjust down my monthly consumption by the amount of overreported solar Tesla shows.

It has been speculated that the Tesla CTs aren't tied to a narrow performance specification, so some CTs over or under report. If it bothers you, I would raise the issue to Tesla and see if they would swap the CT(s) on your solar. Alternatively, is it possible that you have some load in your Enphase panel that drains off some power before it gets to the Tesla CTs?

All the best,

BG

 

[miimura]

It has been speculated that the Tesla CTs aren't tied to a narrow performance specification, so some CTs over or under report. If it bothers you, I would raise the issue to Tesla and see if they would swap the CT(s) on your solar. Alternatively, is it possible that you have some load in your Enphase panel that drains off some power before it gets to the Tesla CTs?

All the best,

BG

The Tesla system is registering higher production than the Enphase and utility meter. A load cannot cause an error in that direction. Tesla just doesn't care to calibrate the system before shipping it.

 

[charlesj]

For those who do not have Enphase, the inverters on each panel sends the performance signals to Envoy that sends the data to Enphase company by by Internet. There is not a CT that would be similar to the one Tesla attaches to the Ac wires at the solar breaker. This is where the issue is between the two.
Enphase claims it to be revenue grade, 0.2% accuracy.
When I compare the two systems report, it varies during the day as Enphase has a 15 minute integrations whereas Tesla does it at 5 minutes but at end of day I am usually at 0,5% and at times less difference.

 

[el-guapin]

I've monitored the Enphase reporting on our system and it was 100.% accurate compared to the utility meter........ And yes, neither the grid meter nor Enphase reporting agrees with my Tesla reporting either.

FWIW, my Enphase reporting is also spot on with my utility meter.

 

[aus1ander]

For those who do not have Enphase, the inverters on each panel sends the performance signals to Envoy that sends the data to Enphase company by by Internet. There is not a CT that would be similar to the one Tesla attaches to the Ac wires at the solar breaker. This is where the issue is between the two.
Enphase claims it to be revenue grade, 0.2% accuracy.
When I compare the two systems report, it varies during the day as Enphase has a 15 minute integrations whereas Tesla does it at 5 minutes but at end of day I am usually at 0,5% and at times less difference.

A good CT should be able to be this accurate. Our issue is that the Tesla CT in the Tesla Gateway routinely reports 7-10% more power than is actually there. There is something specific to the way the Enphase system is set up that is creating this problem, and I have no idea why. Voltage spikes?

 

[BGbreeder]

A good CT should be able to be this accurate. Our issue is that the Tesla CT in the Tesla Gateway routinely reports 7-10% more power than is actually there. There is something specific to the way the Enphase system is set up that is creating this problem, and I have no idea why. Voltage spikes?

Operative word here "should". I think that many of the complaints originate from the difference between "should" and "as installed".

I can think of no reason why the 240V sine wave, grid quality power from Enphase microinverters would be a problem. Even hypothesizing that somehow 10% of them are some how mysteriously defective and outputting something that isn't "clean" power and somehow is misread by the Tesla CTs as added power is improbable. The way microinverters function will automatically cause the other 90% to compensate for deviations from the "erring" 10%. It is intrinsically clean power by design.

Alternatively, one could hypothesize that there is a single point of failure in a device widely reported here to have issue, namely the Tesla CTs.

If it bugs you, I would call it into Tesla service.

If you want to dig in yourself, with all the caveats about these panels being live, high power electrical areas and the dangers that result from that, you might check your gateway for a couple of things;

• one you could also double check the CT wiring to make sure the two wires are tightly twisted all the way to the gateway input
• two, that the CT wires are fastened tightly, (well seated)
• three, that the CT wires aren't lying next to some other power wires that might be inducing current in the CT wires, and
• four, check to see if you have one or two CTs measuring your solar production. If you have two CTs, you could try running on just the A or the B CT (doubling the output) and see if one of them is problematic. (Although I confess that I can't think of a reason why a loose wire would cause erroneously high readings. @miimura is right on the money that a load in the Enphase panel won't cause a Tesla reading to be high either.)

I have seen panels where the CT wires were lying neatly along high current wires which shouldn't matter in principle, but isn't good practice due to the potential for induced currents.

All the best,

BG

 

[skepticcyclist]

My Enphase Envoy always reports about 1-2% more production than my Tesla Gateway. Most likely just loss from the extra wiring and hardware between the microiverters and combiner box to the gateway as well as the gateway processing

 

[charlesj]

Operative word here "should". I think that many of the complaints originate from the difference between "should" and "as installed".

I can think of no reason why the 240V sine wave, grid quality power from Enphase microinverters would be a problem. Even hypothesizing that somehow 10% of them are some how mysteriously defective and outputting something that isn't "clean" power and somehow is misread by the Tesla CTs as added power is improbable. The way microinverters function will automatically cause the other 90% to compensate for deviations from the "erring" 10%. It is intrinsically clean power by design.

Alternatively, one could hypothesize that there is a single point of failure in a device widely reported here to have issue, namely the Tesla CTs.

If it bugs you, I would call it into Tesla service.

If you want to dig in yourself, with all the caveats about these panels being live, high power electrical areas and the dangers that result from that, you might check your gateway for a couple of things;

• one you could also double check the CT wiring to make sure the two wires are tightly twisted all the way to the gateway input
• two, that the CT wires are fastened tightly, (well seated)
• three, that the CT wires aren't lying next to some other power wires that might be inducing current in the CT wires, and
• four, check to see if you have one or two CTs measuring your solar production. If you have two CTs, you could try running on just the A or the B CT (doubling the output) and see if one of them is problematic. (Although I confess that I can't think of a reason why a loose wire would cause erroneously high readings. @miimura is right on the money that a load in the Enphase panel won't cause a Tesla reading to be high either.)

I have seen panels where the CT wires were lying neatly along high current wires which shouldn't matter in principle, but isn't good practice due to the potential for induced currents.

All the best,

BG

My Tesla CT is in my 2nd floor subpanel wrapped around the solar cables at the breaker. the CT box is in that subpanel. The CT's antenna is using a panel knock out into the wall cavity. This subpanel is a good distance from the Gateway and the rest in the garage. The wire is large enough to be on a 100A breaker near the Gateway. So, the CT could be influenced but the data is not there for me.
Interesting why others have such a large discrepancy.

 

[charlesj]

My Enphase Envoy always reports about 1-2% more production than my Tesla Gateway. Most likely just loss from the extra wiring and hardware between the microiverters and combiner box to the gateway as well as the gateway processing

Where is the breaker for the panels? Is there a Tesla CT coil around the wires at the breaker? How far away from the roof penetration to breaker CT?
The wire IR loss can be approximated.

 

[skepticcyclist]

Where is the breaker for the panels? Is there a Tesla CT coil around the wires at the breaker? How far away from the roof penetration to breaker CT?
The wire IR loss can be approximated.

Micros to Envoy as much as 20 feet run. Envoy to Gateway and Subpanel about 12 feet. One CT is in the subpanel with an antenna. Don't know where the others are; maybe in the gateway itself?

 

[charlesj]

Micros to Envoy as much as 20 feet run. Envoy to Gateway and Subpanel about 12 feet. One CT is in the subpanel with an antenna. Don't know where the others are; maybe in the gateway itself?

Thanks. If you only have 1 circuit breaker and an CT in subpanel, that will be it for Enphase CTs. Not sure how Gateway's CTs are installed in it or what it measures as Gateway already receives solar generation data from that antenna in your subpanel just like mine up on the 2nd story.

The Micros send data through a com signal on the power lines to Envoy so there is no power loss as it comes directly from the micros.
The Tesla CT measure current at the solar breaker and that short run would be the only very small wire loss from the inverters to the circuit breaker. I should have asked for that distance as the distance to Envoy has no power loss from production as it travels through a comm signal, totally separate and different from the power current.
I have a single 20A breaker. The wire run is about 10 ft to the CTs and used 10GA wire from the junction box on roof.
I just calculated the loss as 8 Watts for that run when it generates power to have 16A current, about 3800 watts generated. That is about 0.2%.
So, if your wire from roof to breaker is 20 ft one way, that would be 0.4%

 

[skepticcyclist]

Thanks. If you only have 1 circuit breaker and an CT in subpanel, that will be it for Enphase CTs. Not sure how Gateway's CTs are installed in it or what it measures as Gateway already receives solar generation data from that antenna in your subpanel just like mine up on the 2nd story.

The Micros send data through a com signal on the power lines to Envoy so there is no power loss as it comes directly from the micros.
The Tesla CT measure current at the solar breaker and that short run would be the only very small wire loss from the inverters to the circuit breaker. I should have asked for that distance as the distance to Envoy has no power loss from production as it travels through a comm signal, totally separate and different from the power current.
I have a single 20A breaker. The wire run is about 10 ft to the CTs and used 10GA wire from the junction box on roof.
I just calculated the loss as 8 Watts for that run when it generates power to have 16A current, about 3800 watts generated. That is about 0.2%.
So, if your wire from roof to breaker is 20 ft one way, that would be 0.4%

I don’t know why there’s always a few hundred watts difference with mine. Here’s Monday’s, my last completely clear day. 500 watts difference .