PHEV... EV range vs Speed and Temperature... your Emperor Needs You

[Nick72]

I'm not sure if Japan still has a Royal Family but intrigued if they do. I'm a mild Royalist by the way. Just don't mention Prince Andrew at the moment. 

 

Anyway, getting distracted. I've started scraping web data on EV range performance so I can create a model, charts, table of EV range for a given outside air temperature and average speed with and without aircon. EV range being full charge to empty (end of green on the dial).

As you can see there's a lack of data around 50mph and under 35mph as well as a lack temperature data below 18C ... in my scrapes so far at least.

If you got the data please send me here and I'll update the chart and eventually I'll be able to add series curves for each of the temperature ranges like -10 to 0, 0 to 10, 10 to 20, 20 to 30C etc. Note though the only sure way to get the correct data (not a rolling average estimate from the car algo) is to use the odometer at start on full charge and then again when empty. Subtract the two obviously for your miles achieved.

 

If I get enough data I'll create a machine learning model to suck in and use it to populate the attached table also for a quick look up. An hour job tbh. 

So far across the temperature range in question I'm estimating max range at 45mph. But this will change since the data I have is spubik at present. 

Good idea or am I just too bored and it is a stupid idea?

Inspired by my last PHEV from Peugeot. After arguing with them about the shockingly bad Battery mileage they finally gave me a speed, temperature, range table. Which gave me the evidence to say how me getting a third of their official spec (despite light footed acceleration attempts) was a clear indication the car was carbage. 

Just some fun and fascination tbh.

 

Ta

 

[ernieb]

Nick, I’ll send you data from time to time as it’s available.  The problem I have is the data I have in maybe compound in that it might be a couple of trips on different days so I’ll have to average data such as temperature but I’m sure it will be not a massive error as it’s likely to be consecutive dates.

I have a trip am tomorrow that will be continuous and will deplete the Battery so I can take that data set. Let me go over anything I have. OK?

[Nick72]

3 minutes ago, ernieb said:

Nick, I’ll send you data from time to time as it’s available.  The problem I have is the data I have in maybe compound in that it might be a couple of trips on different days so I’ll have to average data such as temperature but I’m sure it will be not a massive error as it’s likely to be consecutive dates.

I have a trip am tomorrow that will be continuous and will deplete the battery so I can take that data set. Let me go over anything I have. OK?

Sounds good Ernie. Suspect same for most folk. Rough average speed and rough average temperature across the few trips that runs the Battery out is as good as we can hope for I think.

The ML model will sort all that out. Amazing what it can do with enough data even if a chunk of it is not great.

Just going to trawl YouTube now for range test videos. 

I'm intrigued of temperature and speed sensitivity. Toyota seem to have done a good job on densentising things to give consistently good ranges. Heat pump and Battery cooler heater tech seems very good. 

From memory the Peugeot was optimised around 20mph 20C. Anything outside of that and range dropped off a cliff.

[philip42h]

If only I had a PHEV Nick ... 😉

Anyway, you are going to need a 3D plot to draw out the 'plane' of range, temperature and average speed ... I think ...

And some of you data points seem 'questionable':

• Two journeys with average speeds of 75 mph? I guess that it is possible but not very "EV"
• A journey at 25 mph with a range of 'only' 46 miles? Maybe a long slow climb up the foothills of Everest? Or someone in traffic alternately stamping on the go pedal and then on the stop pedal?

Wouldn't you expect range to be reasonably flat at moderate speeds for any given temperature - tailing off as speed increases and wind resistance becomes more significant? And, from general evidence from the EV world, tailing off as the ambient temperature falls?

[Nick72]

3 hours ago, philip42h said:

If only I had a PHEV Nick ... 😉

Anyway, you are going to need a 3D plot to draw out the 'plane' of range, temperature and average speed ... I think ...

And some of you data points seem 'questionable':

• Two journeys with average speeds of 75 mph? I guess that it is possible but not very "EV"
• A journey at 25 mph with a range of 'only' 46 miles? Maybe a long slow climb up the foothills of Everest? Or someone in traffic alternately stamping on the go pedal and then on the stop pedal?

Wouldn't you expect range to be reasonably flat at moderate speeds for any given temperature - tailing off as speed increases and wind resistance becomes more significant? And, from general evidence from the EV world, tailing off as the ambient temperature falls?

I think the rogue points will wash out in the analysis. With any luck!

Lots of Americans using the car in EV mode at high speed. I've not done the calcs but suspect it is still more economical to do that than use petrol. Except for perhaps in the deep winter.

We could do a 3D plot or just split the data into 10C temperature bands and have a single chart with multiple curves, each for a temperature band. That's probably easiest. But the machine learning model would sort it all out to be honest and allow me to populate the table.

There will be an optimal as you say. Drag is proportional to V squared.  45mph seems about right as does about 20 to 25C. 

 

[Nick72]

Added some more points. Thanks Ernie. 

Had to split the data into hot, mild, cold. Not enough mild and cold speed vs range data but some interesting observations so far. 

Peak range efficiency at around 40 to 45mph.

Circa 12 or so degrees temperature drop from hot to mild temperatures robs you of about 11 miles of EV range. Suspect this again when dropping another 10C towards freezing.

[ernieb]

Interesting Nick, the more data the better I think.