Efficiency/Range inaccuracy

[drklain]

#143
Describe the bug
The system appears to be using an incorrect value for efficiency when calculating range on the car. As a result, the ranges are all inflated. For my 2017 Model S, 90D it is saying that:

• at 60% charge, the car has a range of 210 miles (equates to a full battery range of 350 miles)
• degradation chart shows max range of around 348 miles (at normal battery capacity on a new battery it should be around 293 #miles)
• Efficiency chart says it is using an efficiency factor of .189 (which I believe is the model 3 factor). Derived efficiency factors are .142, 1.34 and .129 respectively. I believe if the lower efficency factors were used, the ranges displayed would be more accurate, but I'm not sure as I don't know exactly which range data the grafana charts in question are using to calculate the car's range.

Screen captures attached.

**Operating environment (please complete the following information):**
 - OS: [e.g. Synology]
 - Version [e.g. v1.8.0]

[adriankumpf]

I will look into it later. A little background information in advance for you to dig into if you want:

Rated Range  The car's estimate of remaining range based on a fixed energy consumption in Watt-hours/mile (whpm). The fixed whpm is different for different vehicles (85kWh / 60kWh) and in different markets based on the regulatory test for that vehicle in that market.

• US: The fixed whpm is based on EPA 5-cycle Standard
• Europe: The fixed whpm is based on NEDC range in accordance with regulation ECE101
• Canada: Unknown

Estimated Range  The car's predicted remaining range based on energy consumption during recent driving. This is based on a dynamic Watt-hours/mile (whpm) value that is based on recent driving habits. This is not country-specific.

Ideal Range  The car's estimate of remaining range based on a fixed energy consumption in Watt-hours/mile (whpm) at a specified average speed. The fixed whpm is a figure determined by Tesla and is not country specific. The average speed is a figure determined by Tesla and is not country specific.

Right now TeslaMate uses the ideal range everywhere.

[drklain]

No disagreement. I think the problem is that for these cars in question, Teslamate is still pulling data for a Model3 number instead of a ModelS number (hence the .189 efficiency rating). I know you said that Teslamate is now identifying the car as a Model S (90D) but are you sure that the charts in question are using that ID, or are they still keying in on the MDL3 code?

[adriankumpf]

https://github.com/adriankumpf/teslamate/blob/9c61150bd5614728447e21789ab6edc5169b631e/lib/teslamate/vehicles/identification.ex#L44

The base unit is Wh/km

[adriankumpf]

If you have a list of all the efficiency factors let me know.

[drklain]

OK so the problem isn’t that.  Somehow else the wrong range is thus being calculated, but I’m not sure why or how. From: Adrian Kumpf <notifications@github.com> Reply-To: adriankumpf/teslamate <reply@reply.github.com> Date: Friday, September 6, 2019 at 9:48 AM To: adriankumpf/teslamate <teslamate@noreply.github.com> Cc: "webmaster@klain.net" <webmaster@klain.net>, Author <author@noreply.github.com> Subject: Re: [adriankumpf/teslamate] Efficiency/Range inaccuracy (#149) https://github.com/adriankumpf/teslamate/blob/9c61150bd5614728447e21789ab6edc5169b631e/lib/teslamate/vehicles/identification.ex#L44 — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

[adriankumpf]

Okay, regarding your initial concerns:

1. The deviation of the range is due to this: Efficiency per outside temperature & updates #143 (comment)

2. I find the derived factor of .142 rather low. It's lower than the one used by the Model 3. Let's see how the estimation changes after a few more charge sessions. Plus, it's important to know what your mileage is. I believe the battery degradation falsifies the measurement a bit.

[drklain]

I wonder if we have a problem because I'm displaying range in miles vs. kilometers? IN response to your #143 comment, is there a way to force the grafana dashboards to use the other range numbers?

[adriankumpf]

You can change the units at any time via the web interface. Or if you want to override it for a particular dashboard you could update the dashboard variables.

[drklain]

I have the units selected as miles in the teslamate option and everything in grafana shows in miles. I was just wondering if the incorrect TR range for the car was coming from a math error there...

[adriankumpf]

Nope, it's just displaying the ideal range :)

[adriankumpf]

https://www.youtube.com/watch?v=eG2ZdYGDQOY

[drklain]

So instead of ideal, could we not display Typical (which is more realistic)?

[adriankumpf]

The different names are really confusing but that's exactly what I meant here #143 (comment)

[drklain]

I see the referencd, but how do we shift from ideal to rated?  I don’t see an option to do so… From: Adrian Kumpf <notifications@github.com> Reply-To: adriankumpf/teslamate <reply@reply.github.com> Date: Friday, September 6, 2019 at 4:26 PM To: adriankumpf/teslamate <teslamate@noreply.github.com> Cc: "webmaster@klain.net" <webmaster@klain.net>, Author <author@noreply.github.com> Subject: Re: [adriankumpf/teslamate] Efficiency/Range inaccuracy (#149) The different names are really confusing but that's exactly what I meant here #143 (comment) — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

[adriankumpf]

In the medium term, it will be possible to switch from ideal to rated range within TeslaMate. The rated range is already recorded. That means you won't lose any data until then.

with rated range I'm referring to the battery_range from the API which is the more realistic one in your case. Switching between the two is NOT possible at the moment.

[drklain]

Understood.  I’ll look forward to the update! From: Adrian Kumpf <notifications@github.com> Reply-To: adriankumpf/teslamate <reply@reply.github.com> Date: Friday, September 6, 2019 at 4:33 PM To: adriankumpf/teslamate <teslamate@noreply.github.com> Cc: "webmaster@klain.net" <webmaster@klain.net>, Author <author@noreply.github.com> Subject: Re: [adriankumpf/teslamate] Efficiency/Range inaccuracy (#149) In the medium term, it will be possible to switch from ideal to rated range within TeslaMate. The rated range is already recorded. That means you won't lose any data until then. with rated range I'm referring to the battery_range from the API which is the more realistic one your case. That is NOT possible at the moment. Closing this .... — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.