[sciencegeek]

I logged four recent L2 charging sessions and looked at the correlation between temperature and charge rate.
Cabin temperature explains 72% of the variance in Level 2 charging speed. Correlation with battery temperature is worse (63%; not shown).
Correleation is not necessarily causation so we don't know whether the car reads the cabin temperature and directly applies some formula to throttle charging speed or whether it's something correlated with cabin temp.
The rest of the variance is probably due to cell balancing and other factors, including statistical noise.

 

[Humpbackbird]

was the battery SOC the same starting each session? mine charges at 32A for almost 95% of the SOC then it slows a bit. have not seen any temperature dependency.

 

[CyberSlug]

Out of curiosity why do you show the charge rate in amps instead of kilowatts? Also, it definitely seems weird to only be using up to 18 amps on L2.

 

[sciencegeek]

Out of curiosity why do you show the charge rate in amps instead of kilowatts? Also, it definitely seems weird to only be using up to 18 amps on L2.

It's the amps that go into the battery, post AC=>DC conversion. Voltage is between 380 at low SoC and 440 at high SoC.

 

[CyberSlug]

Based on what you see does it look like they are using constant current up to a point then switching to constant voltage? If so, doesn't that basically mean SoC also correlates w/ ambient temp at least for the first part of the charge?

 

[Qtown charger]

I logged four recent L2 charging sessions and looked at the correlation between temperature and charge rate.
Cabin temperature explains 72% of the variance in Level 2 charging speed. Correlation with battery temperature is worse (63%; not shown).
Correleation is not necessarily causation so we don't know whether the car reads the cabin temperature and directly applies some formula to throttle charging speed or whether it's something correlated with cabin temp.
The rest of the variance is probably due to cell balancing and other factors, including statistical noise.

View attachment 3928

Interesting.
Can you replot showing the 4 sessions as unique colors to give an indication of how the cabin temp varies across a charging session.
Could it be as simple as a closed vehicle sitting in the sun slowly heats up, whilst the charging rate slowly drops as the battery gets charged. Both time dependent, but totally independent.

 

[Ayepace]

I wonder if it is coincidental or an affect of the interrelationships of the cooling/heating systems. See the other thread where the Jaguar engineer explained them.

 

[sciencegeek]

Can you replot showing the 4 sessions as unique colors to give an indication of how the cabin temp varies across a charging session.
Could it be as simple as a closed vehicle sitting in the sun slowly heats up, whilst the charging rate slowly drops as the battery gets charged. Both time dependent, but totally independent.

I looked at that (but am too lazy to re-plot). The latest (coolest) session has the least overall throttling and the hottest session has the most, throughout. SoC-dependent throttling doesn't happen on L2 until you get to very high SoC (and I excluded that from the plots; it's a very clear abrupt drop that is governed by the max possible charge, which I can also see with the OBD tool).

To Ayepace's point, yes I think it reflects their design whereby they don't seem to cool the battery unless you turn on the AC/preconditioning or the battery gets super hot (which it wasn't in these charging sessions).

Oh one last thing. I'm just using linear correlation because I don't have a better model; I doubt whatever JLR does regarding throttling is linear (would be kinda dumb). In fact you can see in the plot that up to 80ish degrees it looks pretty flat with little throttling and past 80 it slopes down.

 

[Qtown charger]

I as

I looked at that (but am too lazy to re-plot). The latest (coolest) session has the least overall throttling and the hottest session has the most, throughout. SoC-dependent throttling doesn't happen on L2 until you get to very high SoC (and I excluded that from the plots; it's a very clear abrupt drop that is governed by the max possible charge, which I can also see with the OBD tool).

To Ayepace's point, yes I think it reflects their design whereby they don't seem to cool the battery unless you turn on the AC/preconditioning or the battery gets super hot (which it wasn't in these charging sessions).

Oh one last thing. I'm just using linear correlation because I don't have a better model; I doubt whatever JLR does regarding throttling is linear (would be kinda dumb). In fact you can see in the plot that up to 80ish degrees it looks pretty flat with little throttling and past 80 it slopes down.

I am totally confused by this. You said the charging rate is more strongly associated with the cabin temp than the battery temp, implying that this is not related to the heat generated by charging - unless this heat is transferred into the cabin.
I've never sat in the car while charging - do the air vents blow warm air while charging?

Are you charging during the day, or at night?

I've never seen this change of charging rate, nor have I felt these high cabin temps after charging. So in a sense, my data matches yours 🤔

 

[sciencegeek]

Car's climate is off, no hot air. My car is not garaged so even with a window open and sunshade on the windshield the cabin temps go way up on a sunny CA day. The last charging session was overcast, so it didn't get super hot in the car, and there was very little throttling (though still noticeable).

 

[Qtown charger]

Car's climate is off, no hot air. My car is not garaged so even with a window open and sunshade on the windshield the cabin temps go way up on a sunny CA day. The last charging session was overcast, so it didn't get super hot in the car, and there was very little throttling (though still noticeable).

Have you tried charging after sundown. Presumably the cabin temp would drop with no direct sun. Would the charging rated go up?

 

[sciencegeek]

I haven't because I need to be awake for the logging (it's manual, arg). But I would get the charger log, which might be enough. I'll give that a go one of these days.

 

[squishball]

Oh one last thing. I'm just using linear correlation because I don't have a better model; I doubt whatever JLR does regarding throttling is linear (would be kinda dumb). In fact you can see in the plot that up to 80ish degrees it looks pretty flat with little throttling and past 80 it slopes down.

Science geek, your R^2 is pretty good at 71%. So, it fairly close to linear. You could try a polynomial fit, perhaps a quadratic fit, since your plot looks concave.

 

[sciencegeek]

Aha! Ok yes the R^2 of a quadratic goes up to 0.78. But that doesn't prove anything, does it

 

[sciencegeek]

Have you tried charging after sundown. Presumably the cabin temp would drop with no direct sun. Would the charging rated go up?

Here's an overnight charging session ... as predicted by the mild ambient temperatures, no discernible throttling until the end. (The two dips are unlocking the car, when for inexplicable reasons charging stops and then restarts; Chargepoint smoothes their plots, which I find annoying, so it doesn't show it going to 0; Ayepace's Juicebox plots look much 'better').

 

[CyberSlug]

when for inexplicable reasons charging stops and then restarts

It's because unlocking unlocks the charge connector, and because they are focused on the EU market they decided to make having the charge port be locked a requirement for charging.

 

[Qtown charger]

Here's an overnight charging session ... as predicted by the mild ambient temperatures, no discernible throttling until the end. (The two dips are unlocking the car, when for inexplicable reasons charging stops and then restarts; Chargepoint smoothes their plots, which I find annoying, so it doesn't show it going to 0; Ayepace's Juicebox plots look much 'better').
View attachment 3954

That makes me feel a lot better. I was concerned that our cars were doing different things on charging. That profile looks identical to my usual charging profile.

So, now we are back to the correlation of charging rate and cabin temp. I wonder if L3 charging is affected in the same manner? My guess is most L3 charging is done during the day open to the sun.

 

[Trielectric]

I had a Ford Focus Electric. One 92 degree day I took a 130 mile trip in it. My second charging session it was hot. I decided to sit in the car and use the AC to cool myself. After an hour I looked at my charging port charge indicator and it hadn't changed, only one light out of 4(4 lights =full) was lit. I pulled up ChargePoint and I was getting a full charge of about 6.9Kw hour. My HVAC was sucking up that 6.9kw. On a hot day you hear my Jag HVAC running a lot when it is charging. The Ford was the same. The other night it was 62.degrees. Standing beside the Jag I heard no noise from the cooling system while the car was charging. I opened up the front hood and removed the plastic cover to look for this wiring harness fault some cars are finding. (My belief is the mod is done up under the dash and not under the hood.) When I leaned in the compartment I realized the pumps and compressors for cooling the battery were all running quietly. A 7 kw charger is going to use a whole lot more of the 7kw, on a hot day or night, to cool the battery and charger, than when it is cool outside and you are charging. Of course battery cooling is also going on just driving down the road on a hot day. My 7 kw garage charger uses 29amps when charging at the beginning of a charge.

 

[sciencegeek]

When I leaned in the compartment I realized the pumps and compressors for cooling the battery were all running quietly. A 7 kw charger is going to use a whole lot more of the 7kw, on a hot day or night, to cool the battery and charger, than when it is cool outside and you are charging. Of course battery cooling is also going on just driving down the road on a hot day.

The plot thickens! So you can hear the compressors, I have to check that out next time I charge on a hot day. It seems like L2 charging babies the battery nicely, which makes sense from an engineering perspective as most charging is done at home.