Poor Regeneration when starting off down a large grade.

[Rshephorse]

I spend a fair amount of time at a ski area, and when leaving will start off with a nearly 2000' decline. As a new user I've gotten a bit addicted to watching the 'Energy' graph and, when trying, can usually drive the car with much better efficiency than predicted. But, when starting off with this large descent, the car seems to regenerate only about a quarter of the predicted amount. The graph predicts I'd pick up about 5-6% battery capacity on the descent and I often get less than a 2% bump which is disappointing. The care seems to be performing quite well in other ways, and large descents in the middle of a drive on other mountains are as predicted by the energy graph.

My questions are:

How does tesla come up with these predictions? Just elevation change or data from Tesla cars?- and if so does model Y behave differently from other models? Are they just guessing how a Y will do even though I'm the first one to drive a Y there? Is it possibly that the prediction is just wrong? Is the prediction just S, X and 3 data at this point?

Does driving smoothly down work better, or letting the car accelerate at '0 throttle' and then pulling off on the accelerator to let regen really surge work better?

How much of a difference would preconditioning the battery make? I'm not seeing the dreaded 'dots' showing limited regen capacity, but I'm a little worried that this will be 10 times worse in the winter with a cold battery. How are people getting their battery preconditioned- it seems like charging can work if you time it right, using Tesla App to 'heat the cabin' works, or are their other apps that are better at just heating up the battery?

 

[user212_nr]

The Tesla estimates for the amount of power that it expects you to use often fail to accurately take into account:
a) your speed
b) your climate setting
c) the outside environment

The problem is that they guess these inputs from your average wh/mile and from the navigation data. In some cases they will be very accurate, if your driving wh/mile is constant.

Lets say you've been driving a few hours in traffic, 55mph, then get to an open road, then it will take a while for the nav. computer to change its estimates.

You don't say how long the drive is, or if you desperately need the power to get to your destination, but if not then you have a perfectly normal situation. If you need the power, ex. in the winter, then lowering your speed can help.

 

[Rocky_H]

I spend a fair amount of time at a ski area

= cold

the car seems to regenerate only about a quarter of the predicted amount.

Yeah, regen is charging, and cold will limit how much power it will allow for recharging.

How much of a difference would preconditioning the battery make?

Eh, some. If you do warm your cabin first, it will also do some battery preheating if the battery temperature is pretty near freezing or below.

I'm not seeing the dreaded 'dots' showing limited regen capacity,

Heh--well...they do a little bit of hand waving with the user interface there to try to cut down on people constantly freaking out. It will usually not show you any dots even when it is definitely a little bit limited. When you start to see some dots, it's limited even more.

I'm a little worried that this will be 10 times worse in the winter with a cold battery.

It will be worse. Normally, I would say that's just the way things are through the winter, and it's not too bad dealing with partially limited regen all winter long, but your situation of starting off a drive with a 2,000 foot decent is going to be kind of annoying.

How are people getting their battery preconditioned- it seems like charging can work if you time it right,

Yes, charging very noticeably heats the battery, or as I mentioned, if it's near freezing, cabin heating should also make it do that.

 

[user212_nr]

= cold

He does say toward the end of the post that he is doing this in the summer. Ski lifts can be repurposed for hiking in the summber.

I'm a little worried that this will be 10 times worse in the winter with a cold battery

 

[camalaio]

My reference: 2019 Model 3 LR AWD. The below statements all assume no use of friction brakes on the descent.

I too find that the nav highly overestimates how much energy it will recover from regen, at all times of year and at all reasonable speeds (50km/h through parks as posted, or highway speeds). I've wondered many times about making the same thread. I also find it's at least as efficient as expected outside of the descents in summer (winter is a different story). If you really want to start questioning Tesla's methods, have it recalculate the route multiple times while parked - you'll notice the plan is different every time! I've seen it swing +-3% either direction, completely crazy. So how they estimate this is beyond my comprehension.

I'd have to dig it up in my post history (hopefully it's somewhere), but the energy conversion efficiency was abysmal for descents. Certainly way lower than people's estimates here, but I didn't find any real data I could actually compare to.

I've wondered about your surging question. Assuming regen efficiency is static over all speeds (it's not), the best thing to do would be to keep a steady pace - going faster for any period would result in more aerodynamic losses. I know the regen efficiency changes on a number of variables (including speed), but I do not know the grand picture of all that.

Preconditioning, since you don't have any noticeable regen loss, would make zero difference. There's no specific command to do this, it just seems to decide to do it when the battery is "cold enough" now and the climate control is on. This can happen at any reasonable state of charge, whether or not it is plugged in. Being fresh off of a charge would usually prevent this behaviour (since it is above some temperature anyways for charging), but it's still possible to occur.

Now, winter. In the case you do have regen loss, the math works out such that preconditioning the battery so that it can regen more doesn't gain you anything. It takes a ridiculous amount of energy to raise the battery temperature the amount needed for significantly more regen. You'd probably have to drive straight down Everest to have a net gain, and even then you'd actually be regen limited part way down anyways since it starts to restrict the amount of energy taken in by regen at once. The "unlocked" regen provided by preconditioning in the cold is purely to preserve driving behaviour, and is actually a gigantic waste of energy in the reality of existing roads. Quote me. I'll defend that with numbers when we get to winter. I'm pretty sure even the software version that introduced the battery heating made no claim that this was to make it more efficient.

He does say toward the end of the post that he is doing this in the summer. Ski lifts can be repurposed for hiking in the summber.

It can be quite a bit cooler up on the hills even in summer, enough that it's "cold" to the car (especially if staying overnight, not sure that's the case here though). The local ski hill is forecasted about 8-10C lower right now, for example, and it's not a particularly high elevation ski hill. But since they don't have any regen dots, this ain't the case anyways.

 

[Rocky_H]

But since they don't have any regen dots, this ain't the case anyways.

Heh--well...they do a little bit of hand waving with the user interface there to try to cut down on people constantly freaking out. It will usually not show you any dots even when it is definitely a little bit limited. When you start to see some dots, it's limited even more.

 

[Rshephorse]

Thanks for all your answers. I'll add to this post if I figure anything else out. I'll have to pay more attection to kwh/m and to actual change of battery charge on descents and not count on the 'predicted performance' as much.

 

[Rshephorse]

So I figured out a few things. Get the battery warmed up helps, BUT I've discovered something. Regeneration is a nervous old grannie when it comes to going around corners! My max regeneration on ScanMyTesla gets up into the 60's fairly quickly with a warmed up batter at 60% charge, BUT as soon as I hit a tight sweeping turn it immediately drops the max regen down to 30! My guess is that it is interpreting the diff in wheel rotation between inside and outside tires as spin, and drops regen to help prevent slipping... this is when I am really sure there isn't any actual slippage taking place. It is very slow to come back, and with every tight turn - regen pops back down to 30.

 

[PagodaY]

Battery re-charge is not all there is to be considered. Going down a 2k drop has to impact your brakes. Having the regen work will lessen the brake wear. In the long run that might be more valuable than a few kWH it takes to pre-condition the battery.

 

[user212_nr]

So I figured out a few things. Get the battery warmed up helps, BUT I've discovered something. Regeneration is a nervous old grannie when it comes to going around corners! My max regeneration on ScanMyTesla gets up into the 60's fairly quickly with a warmed up batter at 60% charge, BUT as soon as I hit a tight sweeping turn it immediately drops the max regen down to 30! My guess is that it is interpreting the diff in wheel rotation between inside and outside tires as spin, and drops regen to help prevent slipping... this is when I am really sure there isn't any actual slippage taking place. It is very slow to come back, and with every tight turn - regen pops back down to 30.

Actually sort of interesting.

If you've done any racing, or played any racing games, you'd know that you should not brake during such a turn. There is some allowable braking, but since you don't know exactly what that is, its better not to brake at all. Reduce your speed before the corner, such that you can coast.

Surely it is Tesla recognizing the danger of unintended hard braking during a turn, and that is a good thing. Though corners and winter driving are a weakness for Telsa's one petal driving.

Additionally, unless you brake during the corner, you have not lost any energy due to the lost regen. You loose energy due to the corner, and Tesla nav may not properly assess that in their calculations.

 

[Rshephorse]

Actually sort of interesting.

If you've done any racing, or played any racing games, you'd know that you should not brake during such a turn. There is some allowable braking, but since you don't know exactly what that is, its better not to brake at all. Reduce your speed before the corner, such that you can coast.

Surely it is Tesla recognizing the danger of unintended hard braking during a turn, and that is a good thing. Though corners and winter driving are a weakness for Telsa's one petal driving.

Additionally, unless you brake during the corner, you have not lost any energy due to the lost regen. You loose energy due to the corner, and Tesla nav may not properly assess that in their calculations.

I understand what they are doing, and I definitely understand the dynamics that you are describing (I also drive a '80s 911 - even lifting off the accelerator in corners is quite an issue there too) but the regeneration is reduced not only in that corner, but stays down for the next 10 minutes. For stability - they would only need to reduce regen during 'critically' hard cornering, but could bring regen back up as soon as your lateral g's drop back down... but that's not the way it works. My guess is they don't want the amount of regen to be too dynamic- it would make it a bit too unpredictable? -you would think regen is there if it came back right away, but it would always fade away through the hard corners- which- given that the car often understeers could be a problem... I find that I'm getting a bit slow to but on the brake since I count on regen to do so much.