Surprised not to see a discussion thread about the brand new Energy screens / Charts in Tesla app [7.15.2023]

[aesculus]

@wwhitney and @getakey are on two different themes: @getakey doesn't want to track his energy sources, he is merely pointing out that whatever solar energy he produces should be going to the grid for full credit. In fact he is ONLY credited for solar exports when they are produced and is not supposed to exporting energy he acquired from some other source. So in the simple peak example all the house load would be satisfied by PW discharge and all solar energy produced would be going to the grid for credit.

The Tesla algorithms (have been?) doing a good job of this. But in the latest release it appears that either the powerwalls are not covering the house load as they are supposed to, or Tesla's graphing engine is not showing the correct discharge for the devices.

@getakey do you have the ability to host Powerwall-Dashboard ? If so it may be able to show you what is happening locally. It takes data directly from the GW vs taking it from the Tesla servers. It is showing what the CTs are reading.

 

[getakey]

@wwhitney and @getakey are on two different themes: @getakey doesn't want to track his energy sources, he is merely pointing out that whatever solar energy he produces should be going to the grid for full credit. In fact he is ONLY credited for solar exports when they are produced and is not supposed to exporting energy he acquired from some other source. So in the simple peak example all the house load would be satisfied by PW discharge and all solar energy produced would be going to the grid for credit.

The Tesla algorithms (have been?) doing a good job of this. But in the latest release it appears that either the powerwalls are not covering the house load as they are supposed to, or Tesla's graphing engine is not showing the correct discharge for the devices.

@getakey do you have the ability to host Powerwall-Dashboard ? If so it may be able to show you what is happening locally. It takes data directly from the GW vs taking it from the Tesla servers. It is showing what the CTs are reading.

yes, your first para is exactly it. I'm on Time based control with No grid charging. It looks to me that the graphs are showing the wrong colors in some places. Look at this for the PW during Peak. It is showing grey for grid in portions of the usage. Just like portions of the home load is showing yellow instead of green

 

[wwhitney]

I'm on Time based control with No grid charging. It looks to me that the graphs are showing the wrong colors in some places.

My point is that it doesn't matter (electrically) whether the graphs show at some point in time that PW is 1 kW grey (grid) and the rest blue (house), with PV 1 kW blue (house) and the rest grey (grid), or if it shows that PW all is blue (house) and PV is all grey (grid). The distinction is arbitrary.

In other words, the graphs are trying to show a distinction that is fundamentally undefined.

Having said all that, for the case of export PV only, the latter coloring is simpler and would apparently be less confusing, so perhaps the app designers should switch to that choice.

Cheers, Wayne

 

[getakey]

My point is that it doesn't matter (electrically) whether the graphs show at some point in time that PW is 1 kW grey (grid) and the rest blue (house), with PV 1 kW blue (house) and the rest grey (grid), or if it shows that PW all is blue (house) and PV is all grey (grid). The distinction is arbitrary.

In other words, the graphs are trying to show a distinction that is fundamentally undefined.

Having said all that, for the case of export PV only, the latter coloring is simpler and would apparently be less confusing, so perhaps the app designers should switch to that choice.

Cheers, Wayne

I don't export any PW to grid. Sorry for the volume control in the pic, but you can see the PW usage shows 100% to house so the graph is wrong

 

[wwhitney]

I don't export any PW to grid

That statement is not actually possible from physics. You can say "the PW sets its export level to match the house consumption" and the "net grid export will never exceed the PV output." So those are kind of like your statement, but not exactly the same. [And all of these statements are only true as an average over minutes or longer, the PW doesn't adjust its power output instantaneously, so there will be milliseconds or seconds where the limits are temporarily violated.]

Think of it this way: suppose we have a bank account, and you deposit $500 (PV) and $300 (PW), then withdraw $500 (grid export) and $300 (house loads). Did the $500 deposit go to the $500 withdrawal, and the $300 deposit go to the $300 withdrawal? Or did the $300 deposit plus $200 of the $500 deposit go to the $500 withdrawal, with the balance of the $500 deposit going to the $300 withdrawl?

You can't actually say, as money is fungible. All you can say is that the sum of deposits = the sum of withdrawals.

Cheers, Wayne

 

[getakey]

That statement is not actually possible from physics. You can say "the PW sets its export level to match the house consumption" and the "net grid export will never exceed the PV output." So those are kind of like your statement, but not exactly the same. [And all of these statements are only true as an average over minutes or longer, the PW doesn't adjust its power output instantaneously, so there will be milliseconds or seconds where the limits are temporarily violated.]

Think of it this way: suppose we have a bank account, and you deposit $500 (PV) and $300 (PW), then withdraw $500 (grid export) and $300 (house loads). Did the $500 deposit go to the $500 withdrawal, and the $300 deposit go to the $300 withdrawal? Or did the $300 deposit plus $200 of the $500 deposit go to the $500 withdrawal, with the balance of the $500 deposit going to the $300 withdrawl?

You can't actually say, as money is fungible. All you can say is that the sum of deposits = the sum of withdrawals.

Cheers, Wayne

Ok, I get the physics, but the App is telling me:

• 100% of the PW usage went to the house, yet the graph is showing otherwise.
• 0% of the PW went to the grid.
• 100% of the exports to the grid came from solar
• 0% of the grid went to the PW

 

[wwhitney]

Ok, I get the physics, but the App is telling me:

OK, looking closer at the PW graph you posted (post #29), the bug I see is that it is sometimes showing e.g. 99.9% home (blue) + 0.1% grid (grey) as just a single grey pixel, with the faint blue shading under it. It would like to show it as a a blue pixel with a grey pixel just above it, but maybe in the rendering they end up being the same pixel. It should ensure that the pixel is blue in that case, or put the sharp blue line just under the grey line. Or maybe just round the inputs to the graph the same way it is rounding the summary at the bottom, so it just graphs 100% blue. Or maybe just put the grid at the bottom of the graph instead of at the top of the graph.

Cheers, Wayne

 

[wwhitney]

P.S. Looking at the solar graph in post #29, you can see the small blue spikes between 6 and 9 p.m., where the app has attributed some of the solar as going to the house. Those should correspond to the places in the PW graph where the line is grey with faint blue shading under it, showing a small amount of PW going to the grid. That choice of allocation is what I was referring to as arbitrary.

 

[getakey]

OK, looking closer at the PW graph you posted (post #29), the bug I see is that it is sometimes showing e.g. 99.9% home (blue) + 0.1% grid (grey) as just a single grey pixel, with the faint blue shading under it. It would like to show it as a a blue pixel with a grey pixel just above it, but maybe in the rendering they end up being the same pixel. It should ensure that the pixel is blue in that case, or put the sharp blue line just under the grey line. Or maybe just round the inputs to the graph the same way it is rounding the summary at the bottom, so it just graphs 100% blue. Or maybe just put the grid at the bottom of the graph instead of at the top of the graph.

Cheers, Wayne

But in the Home graph, the amount of the line showing solar is much greater than a few pixels

 

[wwhitney]

But in the Home graph, the amount of the line showing solar is much greater than a few pixels

Not from 5 pm. to ~7:15 p.m., which corresponds to the time the PW graph is showing the PW discharging. What it's trying to show is one pixel of solar above 99% faint green for PW.

The problem in the rendering is when it wants to show Color 1 above Color 2, and Color 1 is only one pixel tall, and the line showing the top of Color 2 gets overwritten by Color1. Below the lines is the faint shading.

Cheers, Wayne

 

[getakey]

ok, here it is zoomed in. I see what you are saying, but I think it is poor UI design. It gives you a totally wrong impression or it did to me

 

[aesculus]

After reviewing my own use case I can unequivocally state that the new graphs are totally useless for me. It is going to take forever to reset my brain to this new format. An example of "improving" something that was not broken IMHO. That is unfortunate as there are so many other things broken that Tesla could be working on.

 

[getakey]

yesterday is an even better example of how the graph looks wrong during Peak. It look like a lot of solar is powering the house instead of the PW which is really powering the house

 

[wwhitney]

yesterday is an even better example of how the graph looks wrong during Peak. It look like a lot of solar is powering the house instead of the PW which is really powering the house

Have you filed a bug report? IMO for the case where one component is very small, they should either (a) graph it at the bottom or (b) ensure that if the lines overlap, they shift the lower (on the screen) color line downward so you can still clearly see both lines, or (c) just shade the area under the curve closer to the color of the line, so the area under the curve is more clearly green, and you aren't misled by the yellow line.

Cheers, Wayne

 

[ArizonaJon]

I find when I turn on the eye to show all activity, it is difficult to see the individual elements. I prefer the eye off and to only look at one element at a time. Then the graph seems to track what is happening.

 

[aesculus]

I find when I turn on the eye to show all activity, it is difficult to see the individual elements. I prefer the eye off and to only look at one element at a time. Then the graph seems to track what is happening.

The "eye" is a disaster.

 

[charlesj]

So are the roof panels. They don's change color to yellow when sun is up, and change color shade as production increase or decrease.

 

[MorrisonHiker]

After reviewing my own use case I can unequivocally state that the new graphs are totally useless for me. It is going to take forever to reset my brain to this new format. An example of "improving" something that was not broken IMHO. That is unfortunate as there are so many other things broken that Tesla could be working on.

This is how most of us felt years ago when they threw out the old graphs and replaced them with graphs that were totally useless for us. We were no longer able to see trends that we'd tracked for years. While there are still a few bugs with the updated graphs, we're grateful they give us back the data we lost years ago.

 

[RHinCT]

...we're grateful they give us back the data we lost years ago.

I haven't had my system very long (early December). I've been exporting the data and saving it into an Excel spreadsheet. What I noticed with the latest update to the app is that they changed the sign on some columns, specifically the sign of what goes to the grid. This was pretty confusing when I pasted August data after July data that was the old way. In any case, since I keep my own copy I don't intend to let that data be at the mercy of Tesla.

 

[swedge]

I see some have Vehicle charging displayed. I have a vehicle, but no such display on the PowerWall graphs.

But here is my question: Where is the PW getting the Vehicle data?

There are current transformers on the grid, solar and PW. House consumption is calculated as the balance of the others. Most installations have the car charging included in house consumption. So to display house and vehicle, perhaps they take info from the car or charger, display that, and subtract the car from the home to display that. Is this what they do?

SW