Low Production and Neurio Inaccuracy

[mbill002]

Hello,

My tesla solar system (no powerwall) was recently installed. Its a 7.6kW system with the 7.6kW tesla inverter. They also installed the Neurio device with 3 CTs - two on the 2 live wires going to the meter and one on one of inverter outputs. I also have a Eagle-200 device that connects to the meter wirelessly and tracks the net demand. I am using home-assistant and the forecast integration to estimate the production. It lets you enter the azimuth and pitch of each panel array and also considers the weather condition to make the estimate - a lot of people in forums report it to be pretty accurate. But I observed the following in the last few days-

1. My production is lower than the estimate home-assistant is making. A picture of the forecast and actual production is attached.

2. Neurio and eagle-200 numbers are different. Neurio numbers are lower most of the time.

I am located in San Jose, CA. I am wondering is this behavior expected? How accurate the Neurio numbers are in general?

Thank you.

 

[power.saver]

Hello,

My tesla solar system (no powerwall) was recently installed. Its a 7.6kW system with the 7.6kW tesla inverter. They also installed the Neurio device with 3 CTs - two on the 2 live wires going to the meter and one on one of inverter outputs. I also have a Eagle-200 device that connects to the meter wirelessly and tracks the net demand. I am using home-assistant and the forecast integration to estimate the production. It lets you enter the azimuth and pitch of each panel array and also considers the weather condition to make the estimate - a lot of people in forums report it to be pretty accurate. But I observed the following in the last few days-

1. My production is lower than the estimate home-assistant is making. A picture of the forecast and actual production is attached.

2. Neurio and eagle-200 numbers are different. Neurio numbers are lower most of the time.

I am located in San Jose, CA. I am wondering is this behavior expected? How accurate the Neurio numbers are in general?

Thank you.

The Eagle will be as accurate as your meter. The Neurios are almost as accurate, say within 1% of actual. Looking at your graph, it seems that maybe some loads are not included in the Neurio CTs. Do you know for sure the Neurio CTs are placed upstream (e.g. closer to the meter) than any other circuits? Are you able to plot the Neurio CT for the solar output? Does it look like a bell curve?

 

[mbill002]

The site CTs are placed after the main breaker and before the connection with the meter. Where the clamps are placed there is a metal plate shielding the live wires and the clamps goes over both the shield and wires - not sure thats gonna introduce any noise.

The solar generation looks like a bell curve, just lower than the forecast. But I also checked the pvwatts estimate, their estimate is much lower and aligned with my production.

 

[Vines]

The site CTs are placed after the main breaker and before the connection with the meter. Where the clamps are placed there is a metal plate shielding the live wires and the clamps goes over both the shield and wires - not sure thats gonna introduce any noise.

The solar generation looks like a bell curve, just lower than the forecast. But I also checked the pvwatts estimate, their estimate is much lower and aligned with my production.

The 100A CT's are accurate within 2%, and with them capturing only one line, this is accurate within 4% of total PV output. The other W1 and W2 neurios are more accurate.

 

[jboy210]

Hello,

My tesla solar system (no powerwall) was recently installed. Its a 7.6kW system with the 7.6kW tesla inverter. They also installed the Neurio device with 3 CTs - two on the 2 live wires going to the meter and one on one of inverter outputs. I also have a Eagle-200 device that connects to the meter wirelessly and tracks the net demand. I am using home-assistant and the forecast integration to estimate the production. It lets you enter the azimuth and pitch of each panel array and also considers the weather condition to make the estimate - a lot of people in forums report it to be pretty accurate. But I observed the following in the last few days-

1. My production is lower than the estimate home-assistant is making. A picture of the forecast and actual production is attached.

2. Neurio and eagle-200 numbers are different. Neurio numbers are lower most of the time.

I am located in San Jose, CA. I am wondering is this behavior expected? How accurate the Neurio numbers are in general?

Thank you.

Perhaps the home-assistant is wrong. Also, we are past the peak solar production for many homes the Summer Solstice, June 21. And depending upon the installation orientation, trees your peak may be even earlier. I am in the East Bay and my peak is closer to the beginning of June which is earlier than estimators I have used predicts.

 

[BGbreeder]

Everything looks ok to me. Production will be somewhat lower due to the details of the estimates and the smoke.

As @Vines pointed above, non revenue grade inverters aren't completely accurate, not measuring both phases will certainly impact accuracy, and your graphs don't look that far apart on average. Expect the numbers to be +/- 5% of "Truth".

All the best,

BG

 

[power.saver]

not measuring both phases will certainly impact accuracy,

Only for unbalanced loads. Inverters are balanced, so doubling one solar CT is the same as having two solar CTs.

 

[BGbreeder]

Only for unbalanced loads. Inverters are balanced, so doubling one solar CT is the same as having two solar CTs.

That's not exactly true. Yes, inverters keep 240V between the two hot lines line to line, but to do that, they do need to routinely have varying currents on the A/B phases to keep both phases 120V to ground for 120V loads. So, measuring one phases is an approximation of the total draw, but not an accurate one for any load that isn't 240VAC. Just think of a Powerwall hooked up to nothing but a 120V toaster. All of the power demand will be on one phase, and none on the other.

All the best,

BG

 

[power.saver]

That's not exactly true. Yes, inverters keep 240V between the two hot lines line to line, but to do that, they do need to routinely have varying currents on the A/B phases to keep both phases 120V to ground for 120V loads. So, measuring one phases is an approximation of the total draw, but not an accurate one for any load that isn't 240VAC. Just think of a Powerwall hooked up to nothing but a 120V toaster. All of the power demand will be on one phase, and none on the other.

All the best,

BG

No, I don't think so. My inverters have no neutral, and the current is balanced between the L1/L2 phases at all times. No current flows through the ground lead. It's just for equipment protection.

 

[mbill002]

So if the inverters produce balanced output irrespective of the load, that means the house will draw power from grid even if the solar is producing more than demand, right? Because one of the phase may have more load and 50% of the production may not be enough?

 

[mbill002]

Perhaps the home-assistant is wrong. Also, we are past the peak solar production for many homes the Summer Solstice, June 21. And depending upon the installation orientation, trees your peak may be even earlier. I am in the East Bay and my peak is closer to the beginning of June which is earlier than estimators I have used predicts.

Thats probably correct. I calculated the estimate using pvwatts and they are more aligned.

 

[power.saver]

So if the inverters produce balanced output irrespective of the load, that means the house will draw power from grid even if the solar is producing more than demand, right? Because one of the phase may have more load and 50% of the production may not be enough?

Correct. The grid will supply energy to cover the difference between inverter generation and load demand, balanced or unbalanced.

 

[BGbreeder]

Correct. The grid will supply energy to cover the difference between inverter generation and load demand, balanced or unbalanced.

No, I don't think so. My inverters have no neutral, and the current is balanced between the L1/L2 phases at all times. No current flows through the ground lead. It's just for equipment protection.

I can't speak to your particular wiring, but I suspect that you have a neutral.

What do you think happens when you are off grid? The Powerwall can't borrow power from the grid.

There is a neutral, as well as a protective earth, aka ground. You might want to check the nameplate on your Powerwall. I would call your attention to the box title "Phase" below, wherein all versions of it include a neutral and a protective earth.

All the best,

BG

 

[taphil]

My inverter has no neutral. The PW does.

 

[power.saver]

My inverter has no neutral. The PW does.

Same. No neutral to inverters, but PWs have them, and they do the load balancing when off-grid.

 

[BGbreeder]

@power.saver My apologies. I was thinking that you meant the inverter within the Powerwall!

Sorry for any confusion as well.

All the best,

BG