Volvo XC40 Forum banner

plotted data of temperature v. energy to recharge since the beginning of the year

3K views 26 replies 11 participants last post by  xc40rechargephilly 
#1 ·
I did a little data analysis and thought it might be interesting to some here. I pulled the daily temperature data (average, high and low) at the Philadelphia International Airport weather station (which is sort of the mid-point of my husband's daily commute). I graphed that against the amount of energy required to restore our XC40 Recharge to its charging set point after each there-and-back 52 mile total commute.
Slope Rectangle Line Font Red


footnotes:
1. Charging data taken from statistics reported by Wallbox Charging Station. Temperature data taken from NOAA.
2. I only included charging data points for dates when I am reasonable sure the only travel done was the 52 mile commute. Since the Volvo App for the AAOS system lacks the trip journal feature, I had to use my judgement about which dates to include and discard (I did compare against my calendar of events, and discarded any date when were likely to have used the XC40 Recharge to take an extra trip).
3. Dates where no charging took place obviously have no data point. Extended gaps in data represent periods in which we were not using the XC40 Recharge (days off work, extended period of travel, etc.).
2. I have decided not to include precipitation and windspeed data, which made the plot overly complex and difficult to interpret.
 
See less See more
1
#2 · (Edited)
energy consumed (over 52 mile commute) v. average temperature

linear line of fit (with 95% confidence band)
Y = -0.1523*X + 28.55
R^2 = 0.7324
I attribute the large scatter to secondary factors — wind, precipitation, snow and ice on roadways at lower temperatures, traffic variability, etc.
Slope Rectangle Font Parallel Plot
 
#3 ·
Fantastic data! Thanks for sharing!
 
#11 ·
OK, we've been keeping track of every charge/trip, and disappointingly I found no decent correlation between efficiency and temperature, or efficiency and speed. Couple plots below. We've driven just over 10,000 km since we bought the car (2022 XC40 Recharge Twin "ultimate" with 20-inch wheels) in late September 2021 in Australia. We had a mild summer and we live at an elevation of about 650 m in a cool climate region to the northwest of Melbourne. The plots below represent about 130 individual data points/charges. The average temperature was just below 15 C. Our overall average efficiency is 22.2 kWh/100 km (35.7 kWh/100 miles). We typically set the cruise control at the speed limit and drive conservatively. We always have OPD switched on. Most of our driving is rural 80-100 kph roads, with some 100-110 kph freeway driving. We recently did a 1400 km road trip on mostly 110 kph freeway (averaged about 24.7 kWh/100 km on the trip, so a bit worse than our long-term average). As you can see below the E-squared value on each of the trend lines is low, indicating a poor correlation. May be of interest! 2 engineer household so we love data, even if it doesn't really prove a point! ;)
Rectangle Slope Font Parallel Pattern

Rectangle Slope Font Plot Parallel
 
#13 ·
It's a bit tedious, but before each charging session, we take a photo of the instrument cluster trip statistics, and we reset that after each charge, and take a pic of the charge % when the charge finishes so we know what the starting charge is for the next trip. We also record the amount of kWh going into the car at each charge - either from the charge station or at home we use an energy meter. Very manual system and we enter it all into a spreadsheet. Examples of pics below to record the data.
Gadget Font Display device Technology Auto part

Light Gas Electronic device Circle Composite material

Automotive lighting Automotive design Gadget Font Tints and shades
 
#14 · (Edited)
I have converted my data to normalized consumption per distance v. temp (to align with @Greg S and because it's better reporting)

(in Americatown units)
Slope Font Rectangle Plot Parallel


(in sensible units for the rest of the world)
Slope Font Rectangle Parallel Plot
 
#15 ·
@Astus This particular commute is approximately 9% city, 91% highway (sometimes with congestion and stop-and-go traffic)

@Greg S Our car has a heat pump. What about yours?

@DreamensioN I pull the consumption data from the app for our wall charging station. It doesn't have an export option so it is a bit of tedium, but it's nice to have the numbers.
Product Rectangle Font Material property Communication Device
 
#16 ·
Yes we have a heat pump. It looks like you converted the wrong way when converting to kWh/100 km...should be divide by 1.6...looks like you multiplied by 1.6? But yeah, you have much better correlation. Not sure why our correlation is so poor...it may have to do with the type of driving we do is so varied and the temperature has been predominantly around 10C...with very few high-temp days as we had a cool summer.
 
#23 ·
Volvo released an api for accessing certain car functions. I need to have a look but maybe that will allow access to data from the car that we can use such as temp, bat temp, average speeds and power consumption and charging info. When I get my car (supposedly sep\oct) I have a good look to see if I can write an app for our phones that automatically collects this data and provides us with the metrics that we can all compare.
 
#24 ·
That would be awesome! The Sensus cars (pre-AAOS) had a lot more data available. My dad has an XC60 T8 plug-in hybrid and can get nice graphs of average fuel economy/efficiency etc)...sure hope all this data we've been generating in our cars over the past year(s) is stored somewhere and may be able to access one day!
 
#25 ·
I’m following this with interest but don’t have any data to contribute. Based on recent experience could I suggest possible additional explanatory variables.

Wet roads seem to have a big impact so either a dummy variable to represent a rainy day or rainfall data might be helpful.

Yesterday we did a short run from home (approx 20 metres elevation) to a restaurant (approx 500 metres elevation). The 60 km round trip used 16% - 11% up and 5% down. By the time we got to the bottom of the range we had more charge than we had at the top. So I’m thinking another possible variable could be maximum elevation gain or average gradient.
 
#26 ·
Yes, that's a good thing to consider (elevation change in particular). We notice when we drive down to Melb from home (~650M down to sea level) we get much better efficiency than coming home. I guess that's obvious! But our average efficiency for going down and back is about the same as our long-term average (~22.5 kWh/100 km). I think one of the most important things to record (or when discussing range) is the car's efficiency in kWh/100 km (or miles if you're in a non-metric country) as that is a much easier number to compare than things like "I drove 100 km and my range went from 72% to xx%", or "my range is 350 km, what's yours?" The former requires a secondary calculation, and the latter is basically meaningless unless more detail is forthcoming.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top