Powerwall Install - PowerFlow Incorrect

[tcsm]

Hi,
A couple of days ago we got our Powerwall 2 installed, we have 5.2KW of Tesla Solar also.

Our house has a 400AMP service so we protected one of the panels (200AMP) with the Powerwall for backup load.

I’m noticing that when we draw from the 200AMP panel that isn’t part of the backup load all the data becomes incorrect.

I attached an example of what I see.
Here everything looked normal then I started charged the car (not part of backup load) the Powerflow show my solar is doing 13KW (impossible) and putting into the grid, but that was not the case at all.
Is this normal? Or have the installers done something wrong?

Hope this makes sense.
Thanks for the help.

 

[power.saver]

The installers have done something wrong, likely misconfiguring or putting the CTs on the wrong wires. You should call them to come back and fix it.

 

[markus478]

Exact same thing happened to me and it was an installer fault. They did not add a circuit breaker and remove the mains connection to one phase in my case.

 

[markus478]

To clarify this can happen if one of the CT clamps is not working correctly. The CT should be installed in a specific way and if installed incorrectly it can cause false readings. From what i read it feels like the 200 AMP panel CT is not working correctly. Ask the installer to use his probe and check the CT. The app uses the CT info and works out some of the readings using calculation. so if one of them is incorrect, it will impact all values and if a particular CT is the problem, then it makes sense that you see issue when that circuit is used.

 

[tcsm]

Thanks for your feedback!
I have been on the phone with them today and so far they have made two changes which made it worse.
The 2nd change they made ended up dumping all the stored energy from the Powerwall @ 5KW. (see attached)
Hopefully it at least went back into the grid, as the meter was showing power going to the grid at a pretty quick rate.

They have now escalated to Tier 2.

 

[markus478]

The image shows the house drawing 18 KW; are you saying that this is incorrect?

 

[tcsm]

Yep - that is incorrect.
Last night it showed I was producing solar all night - now that would be nice! - see attached.

 

[tcsm]

Powerwall also hasn't charged after it dumped 75% of its energy...just sitting in Standby Mode. It was charging until they made the change to try and fix my first issue.
Guessing it stopped at 25% as that is the limit I have it set at.

 

[miimura]

Your CT placement and/or configuration is really messed up.

 

[reddy]

The installer should have caught this before they left. My installer initially had two CTs swapped in the wrong locations. During commissioning it was immediately obvious, and the Tesla people on the phone told him how to change the location to give correct readings.

I was impressed on how the Tesla people on the phone could rapidly identify a problem, and provide instructions to the installer on how to resolve it. We were back online and working properly in 5 minutes.

This sounds like the installer needs to come back and fix it. It should be an easy fix.

 

[markus478]

i think one or more of your CT is connected wrong way around and this is causing the calculations to be incorrect. The CT need to be installed facing the house (all need to be the same direction over the cable). otherwise you will get negative reading; but you can also change negative value to a positive one in the installer portal. Also in the installer portal you should be able to see the current (Amp) levels each CT is reporting. When there is an error you will see one of them is reporting in correct or spurious value. So you know which one to fix / replace.

 

[pdxgibby]

I'm in a very similar situation as the OP. My powerwalls were installed on Tuesday. My power flow is completely wrong. Has been escalated to Tier 2 support as of yesterday. I'm waiting to hear back from them. Hoping that they can send someone out to debug the gateway CT placement, as it sounds like that is the root cause.

As an aside, I also have my own Sense Energy monitor. After the installation, it was really flaky too. I ended up having to do a data reset and let Sense re-calibrate with the new CT placements for Sense. I also realized that 2 CT pairs for Sense in insufficient in a tri-source power configuration (grid, solar, powerwall). I think I'm just going to have Sense attached to solar and grid only -- ignoring when the battery is powering the house. We'll see if that works for what I want to accomplish, although when I switch to advanced time based control, everything will likely be wonky. Anyone on here have a recommendation for how to mitigate that missing CT pair for Sense?

 

[NuShrike]

I'm in a very similar situation as the OP. My powerwalls were installed on Tuesday. My power flow is completely wrong. Has been escalated to Tier 2 support as of yesterday. I'm waiting to hear back from them. Hoping that they can send someone out to debug the gateway CT placement, as it sounds like that is the root cause.

As an aside, I also have my own Sense Energy monitor. After the installation, it was really flaky too. I ended up having to do a data reset and let Sense re-calibrate with the new CT placements for Sense. I also realized that 2 CT pairs for Sense in insufficient in a tri-source power configuration (grid, solar, powerwall). I think I'm just going to have Sense attached to solar and grid only -- ignoring when the battery is powering the house. We'll see if that works for what I want to accomplish, although when I switch to advanced time based control, everything will likely be wonky. Anyone on here have a recommendation for how to mitigate that missing CT pair for Sense?

I've been talking to Sense over the last few days, and they recommended putting all the same phase PW and Solar lines through each of the Sense solar CTs. Then, things would read correctly.

I can't because I have lines in 3 different places so, it's physically impossible to pull them into a CT, per phase. I agree Sense needs an extension to support tri-source power. Maybe replace the solar-socket with a 4-head CT extension instead of the 2 that comes with solar.

I also have a separate Neurio installed which I use for my independent PVoutput monitoring -- same issues with battery-output creating weird consumption values, especially when the A/C comes on.

 

[pdxgibby]

Thanks @NuShrike for the tip on PW + Solar in the Sense Solar CTs. I think I'm in a similar position as you are in that I think the wiring is such that my feeds for PW and Solar are in different physical locations in the routing. I'll see if I can do it though.

 

[pdxgibby]

I've been talking to Sense over the last few days, and they recommended putting all the same phase PW and Solar lines through each of the Sense solar CTs. Then, things would read correctly.

I can't because I have lines in 3 different places so, it's physically impossible to pull them into a CT, per phase. I agree Sense needs an extension to support tri-source power. Maybe replace the solar-socket with a 4-head CT extension instead of the 2 that comes with solar.

I also have a separate Neurio installed which I use for my independent PVoutput monitoring -- same issues with battery-output creating weird consumption values, especially when the A/C comes on.

Googling tells me that there really isn't a 3-input (tri-source) commercially available monitoring device.

 

[NuShrike]

Googling tells me that there really isn't a 3-input (tri-source) commercially available monitoring device.

Too bad because this may be coming up more often as battery-storage installations rise.
Technically, a TED Pro may work as it comes with 6 to 8? pairs of CTs for individual device monitoring.

CTs can be summed which is why Sense suggested what they suggested.

We could also, just split the Solar CTs and only capture one phase of Solar and Battery. Then multiply the value by two, assuming the phases are balanced. This is a Neurio trick to monitor more than two PowerWalls, iirc.

 

[groovidad]

It appears the PW Issues are broad, across regions YET, it's not the PW, it's the technicians. My PWII, installed 19-20 April. Non-functional until local Elec. Co. approved install (a week later). Nothing worked right....NOTHING. It took 4 service calls and 4 different techs, over a week, to get things humming. I used the same installer that put PV on my home to install the PW. Cost me several days of missed work. All is well, now but, to have to go through that was a bit ridiculous.

 

[pdxgibby]

Installed on 6/11.
Called support on 6/11.
Elevated to tier 2 support on 6/12.
Magically started working middle of the day on 6/13.

I know they can assign/invert CTs at the mother ship. It looks like someone did that successfully. I'll keep a close eye on things .... well, forever. But it's very exciting to see things working properly.

 

[irpete]

Hi,
A couple of days ago we got our Powerwall 2 installed, we have 5.2KW of Tesla Solar also.

Our house has a 400AMP service so we protected one of the panels (200AMP) with the Powerwall for backup load.

I’m noticing that when we draw from the 200AMP panel that isn’t part of the backup load all the data becomes incorrect.

I attached an example of what I see.
Here everything looked normal then I started charged the car (not part of backup load) the Powerflow show my solar is doing 13KW (impossible) and putting into the grid, but that was not the case at all.
Is this normal? Or have the installers done something wrong?

Hope this makes sense.
Thanks for the help.

I had the exact same symptoms . In my case I swapped the CTs on the 200 amp panel that isn;t backed up (literally swapped the left and right CT's in the panel). This fixed all of my issues. It seems you need them not only installed label up but it also matters which one is on which side of the panel.

 

[wwhitney]

It seems you need them not only installed label up but it also matters which one is on which side of the panel.

I'm pretty sure that's not true electrically. So swapping them probably fixed your problem as a side effect. E.g. some conector you reseated, or perhaps there was a wiring difference between the two elsewhere in the circuit.

Cheers, Wayne