Why would PW send power to the grid?

[Sonus]

Can someone explain this behavior?

 

[Sonus]

This is right now, so it looks like it stopped but still...

 

[charlesj]

Yes, that doesn't look right to me either. The panels should be sending power to the house, all or part, then to PW if PW is not full, and the grid if there is leftover, in my gathering the info here since reading it.
How is the Gateway set up for the PW, backup only or something else?

Perhaps sensor wires not connected properly?
When was it installed? Did you see this when the install crew was still there?

Also, did you check your PG&E meter if it was receiving all that power from you?
That may be a clue. Always good to compare to the meter outside what is happening with the grid.

 

[Sonus]

Actually I am on cost-saving mode so the house after 3 pm should use PWs power, that is ok, I am just confused about why the PWs send power to the grid that should never be the case in any mode.

also, it happened for a few minutes so i didn't get the chance to check the meter outside. It could be just a bug in the app I don't know.

 

[miimura]

Actually I am on cost-saving mode so the house after 3 pm should use PWs power, that is ok, I am just confused about why the PWs send power to the grid that should never be the case in any mode.

also, it happened for a few minutes so i didn't get the chance to check the meter outside. It could be just a bug in the app I don't know.

Did it last long enough to leave a spike on the intra-day chart in the app?

 

[bob_p]

I've noticed this happens periodically for brief periods before settling into the expected configuration.

The Gateway could be intentionally designed to slowly make adjustments between the house, solar, PowerWall and house power (may be due to requirements for equipment connected to grid power).

Doesn't appear to last long enough to have any significant impact on actual power usage/generation.

 

[Sonus]

Did it last long enough to leave a spike on the intra-day chart in the app?

No

 

[charlesj]

So, from what the others are saying, the switch over in such plans is not instantaneous but dragged out.
And, unless you are already in the garage when you see thit to check PG&E meter, you may miss it if it was a real issue.

 

[getakey]

I think it is a temporary issue in the display only. I see it sometimes very briefly, but then corrects itself

 

[bob_p]

It is real.

According to the smart meter data, we are sending a small amount of power back to the grid periodically - not enough to have any impact.

Suspect the slow switching is intentional - and shouldn't be considered to be a problem if it is only for brief periods.

 

[JayClark]

I think it is a temporary issue in the display only. I see it sometimes very briefly, but then corrects itself

Agreed. And I've noticed what the Tesla phone app displays can sometimes lag the actual Powerwall (if you check via the gateway local website) by any where from a few seconds, to minutes - which can be disconcerting. So what might just be a normal balancing blip on the Powerwall itself, can looks like it's lasting for minutes in the app. I see this regularly, but when I check the chart later, it shows it behaved as expected for the most part. There are occasional blips that do show up on the daily chart, mostly when my house loads are fluctuating "a lot" (like an expresso machine turning on/off quickly as it heats water/steam) or when our car first starts charging, but usually the PW seems to let some power back to the grid to try and get it zero our the incremental loads for the period, in the overall charts.

 

[getakey]

Is there a reason PW charge stops at 95% sometimes while other times going to 100%?

 

[aesculus]

And I've noticed what the Tesla phone app displays can sometimes lag the actual Powerwall (if you check via the gateway local website) by any where from a few seconds, to minutes - which can be disconcerting.

I can play the Watts up game (ie turning on and off appliances/devices) and I see about a second lag at most with my app.

Mine is connected via wi-fi and of course cellular.

 

[JayClark]

I can play the Watts up game (ie turning on and off appliances/devices) and I see about a second lag at most with my app.

Mine is connected via wi-fi and of course cellular.

Mostly it's only a lag of a few seconds on my system. I have some extensive home automation that can tweak PW settings via the APIs, so at times I'm shifting the configurations, and at those times I will be observing the app and local web reporting for extended amounts of time and it's usually then, only after hours of observing (fairly closely) that I'll catch those periods where it lags by more than a few seconds, but it takes a lot watching.

Normally I don't notice it. Maybe 1 day out of 15-20 days, an automated API configuration change pushed to the PW won't go through as scheduled, and often that's also when I notice a bigger difference in the sync of what the two views are showing. If I wasn't doing the automation, and getting alerts, I might not ever notice it.

 

[aesculus]

Maybe 1 day out of 15-20 days, an automated API configuration change pushed to the PW won't go through as scheduled, and often that's also when I notice a bigger difference in the sync of what the two views are showing.

I often find I cannot reach my PW gateway via my local net and browser. It may be my situation where the gw is attached to a bridge like device. I can see the pw gw in the bridge but for some reason it does not get the IP address from the access point. For this reason I cannot depend on using it for any API features until I get this figured out.

 

[jboy210]

I often find I cannot reach my PW gateway via my local net and browser. It may be my situation where the gw is attached to a bridge like device. I can see the pw gw in the bridge but for some reason it does not get the IP address from the access point. For this reason I cannot depend on using it for any API features until I get this figured out.

Are you using Ethernet or Wifi?

We have both enable on the gateway, since the Wifi is inside the house which is stucco and the GW is outside. Our stucco is held on by a wire mesh making it a Faraday cage that kills the Wifi signal strength outside the house.

 

[aesculus]

I am using WiFi but through a very powerful access point linked to the house router via ethernet.

The problem may be in that the "remote" AP is just passing through the IP address assignment to the router and there may be some hand off issues there. The signal strength is very good.

I may have to look at assigning the IP address of the GW permanently via the setup (if possible) rather than having the router set it's fix IP address via the MAC address.

 

[jboy210]

I am using WiFi but through a very powerful access point linked to the house router via ethernet.

The problem may be in that the "remote" AP is just passing through the IP address assignment to the router and there may be some hand off issues there. The signal strength is very good.

I may have to look at assigning the IP address of the GW permanently via the setup (if possible) rather than having the router set it's fix IP address via the MAC address.

Setting a permanent address outside the DHCP pool address space is what I do for most of my fixed devices (ex. NAS and other servers).

With wifi be careful increasing signal strength beyond the standard values. That can lead to issues where the sign appears strong at a distance, but communication is spotty.

 

[aesculus]

Somehow a bunch of edits on my last post disappeared. The net of it is that I could not find a way to assign a permanent IP address to the wifi connection via the gw and only the ethernet, that is not connected, supports that. And then when I canceled the ethernet setup then the gw no longer responded so I am now stuck in purgatory until the gw decides to retry connecting via wifi.

Anyway that's how I assigned the permanent IP address but for some reason it does not work 100% of the time.

In the GW browser the WiFi signal strength is 4/5 bars.

 

[aesculus]

And I've noticed what the Tesla phone app displays can sometimes lag the actual Powerwall

OK. I have seen this now. Interestingly the real time flow seems OK but the SoC seems to be off. Also interesting is that the SoC was 60% (my reserve level) when I checked after peak stopped the night before and the PW were no longer discharging. You will notice one is 54% and the other 57%. I guess this is either vampire drain and/or the latent temperature of the PWs.