Powerwall 2: Installation

[Ampster]

It's almost as if the local utility is attempting to sabotage your efforts to powering your hom

It probably depends on where you live. I am in Sonoma and found out the building officials and PG&E are still catching up. The County Building Department was closed for 3 weeks because they had no power. My solar install went quickly on 12/18 but coordinating with PG&E for the panel upgrade has taken a long time. I won't get the PTO until then. At least my home didn' burn down. I can adjust my expectations.

 

[arnolddeleon]

Thanks to @Kren I was able to self-register using the following link: https://www.tesla.com/powerwall-self-registration. However after a couple days I still get "Registration Incomplete" when I look at the webserver on the gateway directly. See attached image. I called Tesla support and they told me to just ignore it. The person at the other end sends they see the powerwall and the registration was done with the email address that I used for the self-registration process. How do I know that my install is really complete?

arnold

 

[arnolddeleon]

A couple of pictures showing something close to the final Configuration.

New main panel is installed (sorry didn't have it open, it only has a main breaker). It feeds the gateway to the left of it. The gateway feeds the main load panel that below it and the generation panel which is on the other side of the wall.



Here is the generation panel. The powerwalls are connected here and so are my two PV system. There are a couple of unused breakers because the installer didn't have any blanks to cover. Note the the large CTs that Tesla uses to monitor the output of the PV systems. Also visible are smaller CTs that I was in the process of installing for my monitoring system. BTW you will notice in this photo that the one Powerwall breaker was on and the other was off. It appears that there is nothing in the system that will alert if you one of the Powerwalls has it's breakers off. I ran in this configuration a little while before I noticed that breaker was off in the photo.

 

[Kren]

Thanks to @Kren I was able to self-register using the following link: | Tesla. However after a couple days I still get "Registration Incomplete" when I look at the webserver on the gateway directly. See attached image. I called Tesla support and they told me to just ignore it. The person at the other end sends they see the powerwall and the registration was done with the email address that I used for the self-registration process. How do I know that my install is really complete?

arnold

View attachment 273442

Hey, how did you get that Stop Powerwall button??? Is your app running on android or ?. I think you are running this directly through your PC. How does it look when you run the app on your phone? Do you still see the message Registration Incomplete? Do you still have that nice red button on your phone app?

 

[sreehyd]

Didnt see anyone reporting "zombie" power drain from powerwall2's during night.

My setup is 2 powerwalls on ver 1.10.2 with Solar Panel integrated , After day-long use for charge (solar) and discharge during TOU times (2pm - 9pm) (PGE). The quite period is between 9pm to 8am , I have noticed during quite period, I am losing about 2 to 3% charge which is roughly accounts for 1Kw , This was consistent and noticed for about 4 to 5 days, So i decided to "Turn-off" physical switch and see what happens. I did this continuously for about 3 days Didnt see any loss . Though I didt account every "watt" but didnt see any %points difference .

Does anyone else notice this ? . I never tested this in past , because I recently got my Solar (2 weeks ago) and noticing this zombie drain.

thanks

 

[arnolddeleon]

Hey, how did you get that Stop Powerwall button??? Is your app running on android or ?. I think you are running this directly through your PC. How does it look when you run the app on your phone? Do you still see the message Registration Incomplete? Do you still have that nice red button on your phone app?

The image from talking to the gateway computer directly on my LAN via web browser. The red stop button doesn't show up in the app. The "Registration Incomplete" also doesn't show up on the phone app.

The other bonus today is I finally got Tesla to merge my two accounts and I was pleasantly surprised that it went seamless.

arnold

 

[eml2]

Hey, how did you get that Stop Powerwall button??? Is your app running on android or ?. I think you are running this directly through your PC. How does it look when you run the app on your phone? Do you still see the message Registration Incomplete? Do you still have that nice red button on your phone app?

The red Stop Powerwall button only shows up on Web UI with 1.9.x or later.

Didnt see anyone reporting "zombie" power drain from powerwall2's during night.

My setup is 2 powerwalls on ver 1.10.2 with Solar Panel integrated , After day-long use for charge (solar) and discharge during TOU times (2pm - 9pm) (PGE). The quite period is between 9pm to 8am , I have noticed during quite period, I am losing about 2 to 3% charge which is roughly accounts for 1Kw , This was consistent and noticed for about 4 to 5 days, So i decided to "Turn-off" physical switch and see what happens. I did this continuously for about 3 days Didnt see any loss . Though I didt account every "watt" but didnt see any %points difference .

Does anyone else notice this ? . I never tested this in past , because I recently got my Solar (2 weeks ago) and noticing this zombie drain.

thanks

I notice the small drain too. It is about 1% to 2% per day. Turning the physical switch seems to stop the drain.

 

[arnolddeleon]

The red Stop Powerwall button only shows up on Web UI with 1.9.x or later.


I notice the small drain too. It is about 1% to 2% per day. Turning the physical switch seems to stop the drain.

I have my own CTs on my 2 Powerwalls. I see them pushing around 70 watts when they are "idle" at night. I'm guessing they keep the inverter running so they can respond quickly to a power outage.

 

[sreehyd]

Ya, That makes sense. But 70 watts seems to be bit high for having Inverter to be in standby state . I will check today by turning off "soft-RED switch" on Gateway, Rather physical switch .

 

[jeep1979]

Don't forget all the phantom vampire devices that draw little but cumulative power, i.e. TV, amazon devices on standby, modem, wifi router, clocks (standalone, in microwaves, DVRs, stoves, etc.), chargers, refrigerator (even when compressor not running and more. I also have additional CT on a wifi device and whole house surge suppressor. So take all that into account as well as the Tesla switch.

On another note, Tesla and the county were out at my house today and signed off on the final inspection of my installation. One step more to completion.

 

[Kren]

Don't forget all the phantom vampire devices that draw little but cumulative power, i.e. TV, amazon devices on standby, modem, wifi router, clocks (standalone, in microwaves, DVRs, stoves, etc.), chargers, refrigerator (even when compressor not running and more. I also have additional CT on a wifi device and whole house surge suppressor. So take all that into account as well as the Tesla switch.

On another note, Tesla and the county were out at my house today and signed off on the final inspection of my installation. One step more to completion.

I agree. I took my app and went to every room when everyone was asleep. So many devices drawing power. Turned off outside lights,T.Vs, tablets, Notebooks, etc. and my draw went from .5 Kw to .2 Kw. I have gotten to the point where I check my Tesla app before bed. If it is more than .5Kw I know someone left a closet light, outside patio light or something of that nature on.

 

[arnolddeleon]

I have my own CTs on my 2 Powerwalls. I see them pushing around 70 watts when they are "idle" at night. I'm guessing they keep the inverter running so they can respond quickly to a power outage.

View attachment 273714

I was wondering if the Powerwalls were also responding to what they thought was solar output. As it happens one of PV systems consumes some power at night because it is charging some old lead acid batteries. As an experiment killed the power to my inverters so they wouldn't draw power. It looks like this dropped the the amount of power that the Powerwalls are putting out down to 30 watts while are in "standby".

 

[arnolddeleon]

I was wondering if the Powerwalls were also responding to what they thought was solar output. As it happens one of PV systems consumes some power at night because it is charging some old lead acid batteries. As an experiment killed the power to my inverters so they wouldn't draw power. It looks like this dropped the the amount of power that the Powerwalls are putting out down to 30 watts while are in "standby".

View attachment 274321

And sure enough as soon as I turned the solar inverters (Outback GVF3648 if you are curious) on again the Powerwalls started "discharging" to approximately match what the inverter (actually charger) was pulling. This makes suspect that they are simply reading the absolute value of the power flow on the CTs they have attached to the PV system.

So in summary I think each powerwall pushes about 15 watts when it is idle. The powerwalls also turn on if they think there is solar but they can be confused by consumptions from the solar system (e.g. one that has an inverter/charger).

 

[miimura]

Shouldn't it know the current is going the other way when the charger is going as opposed to solar production output? Instead is it trying to zero out the flow through your utility meter? You should be able to separate the inverter/charger load from your household loads.

 

[mongo]

Shouldn't it know the current is going the other way when the charger is going as opposed to solar production output? Instead is it trying to zero out the flow through your utility meter? You should be able to separate the inverter/charger load from your household loads.

The PW knows the current is going into the PV feed, since the PW is increasing its output. If Tesla didn't consider this use case, then they likely didn't add an call to abs() for the solar input.

Edit: or it is operating as intended for use with self-melting PV panels/ solar shingles.

 

[MP3Mike]

Shouldn't it know the current is going the other way when the charger is going as opposed to solar production output? Instead is it trying to zero out the flow through your utility meter? You should be able to separate the inverter/charger load from your household loads.

CTs can't tell you which direction the energy is going, they just tell you how much current is going through them.

 

[mongo]

CTs can't tell you which direction the energy is going, they just tell you how much current is going through them.

On their own sure, but if ithe GW monitors the phase voltage, then you can get direction. The CT has polarity, so if the current polarity/phasing is opposite the voltage polarity/phasing, then there is a source instead of a load.

 

[arnolddeleon]

On their own sure, but if ithe GW monitors the phase voltage, then you can get direction. The CT has polarity, so if the current polarity/phasing is opposite the voltage polarity/phasing, then there is a source instead of a load.

Yes the the CTs I use most definitely have polarity. The gateway clearly tracking polarity/phasing since can tell the different between buying and selling to grid. It also displays if the inverters are charging or discharging (although it could get information through other means). I suspect they actually use abs() for the solar CT, it simplifies things in most cases. It means the installation engineer doesn't have keep track/calibrate for direction. OTOH if they supported measuring net flow they could support more esoteric installation configurations, like ones were they can't put the CTs on the solar output lines but instead just look at the net output of a sub panel that has solar and loads.

I would be thrilled if Tesla handled my corner case but I would be surprised if they did. My hope is someday to replace my Outback inverters with something better. That would likely easily give me 5% boost in output and possibly over 10%. Newer inverters have significantly higher efficiencies and I would lose some of the losses that i'm seeing.

arnold

 

[mongo]

Hi

I suspect they actually use abs() for the solar CT, it simplifies things in most cases.

According to your test, didn't the PW put out more power when the inverter was drawing power? That is why I thought the PW was directly acting on the CT reading: charge during PV output, discharge during inverter draw.

And I misspoke, I was not thinking of an ABS(), but rather a MAX(0, sensor) for never negative.

 

[arnolddeleon]

Hi



According to your test, didn't the PW put out more power when the inverter was drawing power? That is why I thought the PW was directly acting on the CT reading: charge during PV output, discharge during inverter draw.

And I misspoke, I was not thinking of an ABS(), but rather a MAX(0, sensor) for never negative.

Yes the PW increased its output when I my solar inverter/chargers started consuming power. I just realized the the PW may actually know the direction of the power flow with the solar CT because the behavior also resembles an attempt to zero it. I originally was thinking it was trying to "charge" itself by seeing "positive" flow from the solar but the ends up discharging. There may not be a practical difference between those two behaviors. What I'm sure of is that the PW responds to the state of the solar CT even if that state indicates power flowing in instead of out.