Estimated charge time way off

[jsmay311]

I think y'all have the wrong take on this. Although the inaccurate time-to-completion estimate is indeed related to the cold, it's not a sufficient excuse for just how dramatically inaccurate the estimate is.

The estimate is still way off even after reaching a steady-state max power charge rate, and even with L2 charging.

For example: I charged my car last night in zero degree (F) temps in my detached garage at 240V 16A. The battery started bone cold (zero regen), so close to worst-case conditions, and yet once the battery was warm it charged at the same speeds I'd see in the summer. But the charge time estimates were still way off.

• 0 F ambient temps
• 240V, 16A charging
• First 20 minutes of charging session went to heating the battery.
• Once the battery was warm enough for full power to go into charging the battery, the time estimate didn't drop.
• After warming the battery, it estimated 14 hours to completion. It actually took less than 10 hours.
• The % error of the estimate never improved throughout the charge. It actually got worse.
  • It started out at ~45% error.
  • After 3 hours it was up to ~50%.
  • With 2 hours remaining, it was ~60% off.
  • With 1 hour remaining, it was off by ~85%
  • With only 1 minute from completion, the estimate was 30 minutes.

In summary, the estimate is terribly inaccurate in cold temps, and there's no good excuse for it being so terrible. Calculating reasonably accurate estimates should be a very simple task, particularly for L2 charging, but also L1 (albeit with a greater margin-of-error).

There's no need to make excuses for Tesla's failure to do this competently.

 

[jjrandorin]

I think y'all have the wrong take on this. Although the inaccurate time-to-completion estimate is indeed related to the cold, it's not a sufficient excuse for just how dramatically inaccurate the estimate is.

The estimate is still way off even after reaching a steady-state max power charge rate, and even with L2 charging.

For example: I charged my car last night in zero degree (F) temps in my detached garage at 240V 16A. The battery started bone cold (zero regen), so close to worst-case conditions.

• 0 F ambient temps
• 240V, 16A charging
• First 20 minutes of charging session went to heating the battery.
• Once the battery was warm enough for full power to go into charging the battery, the time estimate didn't drop.
• After warming the battery, it estimated 14 hours to completion. It actually took less than 10 hours.
• The % error of the estimate never improved throughout the charge. It actually got worse.
  • It started out at ~45% error.
  • After 3 hours it was up to ~50%.
  • With 2 hours remaining, it was ~60% off.
  • With 1 hour remaining, it was off by ~85%
  • With only 1 minute from completion, the estimate was 30 minutes.

In summary, the estimate is terribly inaccurate in cold temps, and there's no good excuse for it being so terrible. Calculating reasonably accurate estimates should be a very simple task, particularly for L2 charging, but also L1 (albeit with a greater margin-of-error).

There's no need to make excuses for Tesla's sloppiness here.

Your 240v 16 amp circuit charges at 3.8kW, which I am almost positive is less than the car will pull if it is trying to warm the battery, so the same thing applies to you as applies to the OP, just in lesser degrees.

if the car can pull more power than you are putting in at any one time because its cold and needs to warm the battery back up, that doesnt sound super simple to me to have "acccurate estimates", but I am not a programmer either.

 

[jsmay311]

Your 240v 16 amp circuit charges at 3.8kW, which I am almost positive is less than the car will pull if it is trying to warm the battery, so the same thing applies to you as applies to the OP, just in lesser degrees.

Correct... I have TesalFi, so I can see exactly what the charge rate is in MPH. For example, when the battery is cold soaked, I can see that at the beginning of the charge most of the power is going towards heating the battery rather than charging it.

But once it's warm enough to accept the full 3.8kW, it very rarely has to slow down the charge rate at all.

In fact, last night's charge was the first time I've ever seen it throttle back the charge rate in the middle of a charge. And even in that extremely cold scenario, it only activated the battery heater for 10 minutes total (twice for 5 minutes each time) over the course of a 10 hour charge.

So while the charge time estimates were off by upwards of 50%, the car actually only needed an extra ~1.5% of time to warm the battery. I'd hope we'd agree that that's a really bad estimate.

if the car can pull more power than you are putting in at any one time because its cold and needs to warm the battery back up, that doesnt sound super simple to me to have "acccurate estimates", but I am not a programmer either.

The car knows:

1. kW available from the charger
2. Battery temperature
3. Ambient temperature

With those 3 values, it should be relatively trivial* to estimate how much battery heating will be required during the charge and to factor that extra time into the charge time estimates as needed.

(*Extreme examples, like L1 charging in -20F temps, where the charger can't even keep up with the heated demands, are an exception. But the vast majority of scenarios like the ones that the OP and I are talking about should be very easy to estimate accurately.)

 

[jsmay311]

Your 240v 16 amp circuit charges at 3.8kW, which I am almost positive is less than the car will pull if it is trying to warm the battery, so the same thing applies to you as applies to the OP, just in lesser degrees.

One small correction which I overlooked in my previous reply: I have a RWD Model 3, and I believe the single motor pulls only ~3.5kW for heating the battery. So 3.8kW does allow for the full battery heating power to come from the grid with a small amount left over to go into the battery.

This aligns with what I see in TeslaFi. When the battery is so cold that it can't accept any charge, the car draws ~14A (~3.4kW +/-0.1kW) to power the battery heater. Then after awhile the grid current rises to the full 16A, but for the first 5-10 minutes the battery heater stays on, so the resultant charge rate going into the battery hovers around only 1-2mph. After this 5-10 minute period, the battery heater turns off and the charge rate jumps up to 14-15mph.

 

[jsmay311]

I was delighted at first when I saw in the 2022.8.3 release notes that charging time estimates were going to be improved by taking into account battery pack temperatures. But then I was disappointed to see that it only applied to Supercharging and other fast charging.

The inaccurate AC charging time estimates in cold weather is such low-hanging fruit for Tesla to fix.

Charging Time Estimation​

Estimated charging times are now more accurate by taking the current battery pack temperature into account when a vehicle is connected to a Supercharger or a third-party fast charger.

 

[Lephturn]

I was delighted at first when I saw in the 2022.8.3 release notes that charging time estimates were going to be improved by taking into account battery pack temperatures. But then I was disappointed to see that it only applied to Supercharging and other fast charging.

The inaccurate AC charging time estimates in cold weather is such low-hanging fruit for Tesla to fix.

Along with adding a “ precondition now” button in the app and having it notify when it is preconditionEd and ready.

 

[GtiMart]

@jsmay311 , I believe the estimation is a little bit less trivial than you think it is. You would also need to account for the wind, which affects the rate of cooling of the battery along with outside temperature, to estimate properly. And wind can change during charging, as well as outside temperature, especially if you charge over multiple hours. There's no escaping the fact that the estimate will change during the charge at least, it's estremely difficult to precisely estimate it upfront.

 

[jsmay311]

@jsmay311 , I believe the estimation is a little bit less trivial than you think it is. You would also need to account for the wind, which affects the rate of cooling of the battery along with outside temperature, to estimate properly. And wind can change during charging, as well as outside temperature, especially if you charge over multiple hours. There's no escaping the fact that the estimate will change during the charge at least, it's estremely difficult to precisely estimate it upfront.

It’s not less trivial than I think. This isn’t graduate-level thermodynamics here, and I’m not expecting accuracy down to the minute. But having charge times routinely overestimated by well over 100% is f’ing ridiculous. There’s no need to make excuses where none are warranted.