Normalised Range Estimate

[wattmatters]

Range estimate variability. Thought I'd post some actual data to show how much it can vary (and why).

The chart below shows the Normalised Range Estimate calculated at various times over the past four months:

Normalised Range Estimate is the range estimate if the car at the time had 100% state of charge (SOC).

e.g., if for instance the car reported it had an estimated range of 255 km and the SOC was at 66%, then that's 255km / 66%SOC = 3.79 km per %SOC. Multiply by 100 to give a normalised range estimate at 100% SOC, i.e. 379 km.

This is a calculation based on the data the Home Assistant SAIC integration automatically collects from our car. Data collection began in late March this year, so just over four months of data.

We can see normalised range estimate varies quite a bit, and for our car over the last four months between a minimum of 304 km to a maximum of 548 km.

This is for our MG4 Essence, 64 kWh battery. Northern NSW, Australia.

The estimated range the car reports is based on what the car's energy consumption rate has been recently, and that will vary depending on environmental conditions (cold weather reduces range), speed of travel (higher speed reduces range), type of terrain (gaining elevation reduces range), tyre pressures and rolling resistance of different road surfaces, way the car is driven (lot of hard accelerations reduces range), what accessories are being used (e.g. using aircon or heater reduces range) and so on.

In our case the lower range estimates would be associated with periods of highway driving, or including trips up the local mountain (~1km elevation gain). Higher range estimates when doing slower local driving, or include trips down the mountain. Being in rural/regional Australia there is a fair chunk of highway driving in the mix. Each trip to regional centre involves time on a 110 km/h highway, typically using cruise control.

Anyway, this is nothing new to most of you but thought it was novel to normalise the data and display it over a longer time period. It could equally have been an economy rate calculation.

 

[webhiker]

Brilliant piece of work.
You might be right that it only proves weekdays being said, but that the point, it proves and documents it ???

What did you do in details to have Home assistant and the lot capture this data?

 

[wattmatters]

What did you do in details to have Home assistant and the lot capture this data?

Start here:

API integration

I had a go at getting this integration installed and with a fair bit of help from the good people on the HA user forums* I have it running. Working on some dashboards, but the nice thing with this is I can have a dashboard operating on both my and my wife's phone whereas iSmart is limited to...

www.mgevs.com

 

[fnegroni]

Range estimate variability. Thought I'd post some actual data to show how much it can vary (and why).

The chart below shows the Normalised Range Estimate calculated at various times over the past four months:

View attachment 28451

Normalised Range Estimate is the range estimate if the car at the time had 100% state of charge (SOC).

e.g., if for instance the car reported it had an estimated range of 255 km and the SOC was at 66%, then that's 255km / 66%SOC = 3.79 km per %SOC. Multiply by 100 to give a normalised range estimate at 100% SOC, i.e. 379 km.

This is a calculation based on the data the Home Assistant SAIC integration automatically collects from our car. Data collection began in late March this year, so just over four months of data.

We can see normalised range estimate varies quite a bit, and for our car over the last four months between a minimum of 304 km to a maximum of 548 km.

This is for our MG4 Essence, 64 kWh battery. Northern NSW, Australia.

The estimated range the car reports is based on what the car's energy consumption rate has been recently, and that will vary depending on environmental conditions (cold weather reduces range), speed of travel (higher speed reduces range), type of terrain (gaining elevation reduces range), tyre pressures and rolling resistance of different road surfaces, way the car is driven (lot of hard accelerations reduces range), what accessories are being used (e.g. using aircon or heater reduces range) and so on.

In our case the lower range estimates would be associated with periods of highway driving, or including trips up the local mountain (~1km elevation gain). Higher range estimates when doing slower local driving, or include trips down the mountain. Being in rural/regional Australia there is a fair chunk of highway driving in the mix. Each trip to regional centre involves time on a 110 km/h highway, typically using cruise control.

Anyway, this is nothing new to most of you but thought it was novel to normalise the data and display it over a longer time period. It could equally have been an economy rate calculation.

Can you overlap the temperatures in the region?

 

[wattmatters]

Can you overlap the temperatures in the region?

Here it is with the exterior temperature as reported by the car. I adjusted the scales to separate the two data channels so as to make it easier to see any trends.

Given there is only 4 months of data I'm not sure there is enough weather variance to pick up an obvious correlation.

 

[FrankMG4]

Great data capture!

Yesterday I drove (mostly highway at 110km/h with cruise control) to my daughter and used 28% (from 92% to 64%) to get there, but I used 37% (from 64% to 27%) to drive back!?
So I wonder if the SOC calulcation that the car shows does show the SOC(capacity) or the SOC(energy). Great video about this: !
Note: The channel Stafl Systems has many interesting videos about lithium-ion battery technology and BMSs.

 

[wattmatters]

used 28% (from 92% to 64%) to get there, but I used 37% (from 64% to 27%) to drive back!?

Wind direction and elevation changes can make a big difference to consumption each way.

 

[wattmatters]

Yesterday did a 460km round trip, mostly 110 km/h dual carriageway.

Can see how at the start the range estimate was high, as it was assuming I was continuing with local driving. As we progressed can see how the range estimate was falling - and took quite a while to stabilise at the new highway speed consumption level. One stop for coffee on way there and a brief DC charge not far from our destination.

On the return journey we must have had more of a headwind as energy consumption rate was higher, indeed I needed to take some remedial action in order to improve our economy so as to make it home without need of a charge stop (I had a deadline to pick up our dog), so I dropped the speed back by 5-10 km/h and did some cruise control managed drafting of larger vehicles.

Can see how that improved range again, and once I was confident we had enough buffer I returned to regular highway cruise speed.

Got home with 5% left.

I think the main thing to notice is how much of a lagging indicator wrt changes in driving energy demands the range estimate is.