Emergency Battery Saving (2,500 Mile Vacation from Wisconsin to New York and back)

[gad1976]

Hi All,

Taking the longest trip yet in my new Model Y LR. Loving the car.

Today we left the beaten path and I had my first scare. We left the freeway to see Fort Ticonderoga in NY. Very nice, by the way.

The issue quickly became battery. The computer told me when I left the freeway I would be able to make it to the Fort and then to my final destination and still have 30% battery when I arrived. This is my first time in mountains with my MYLR and the battery began to drop rapidly. I might have been able to be more careful, but I drove as I usually drive (the mountains being the new part of the equation). I anticipated that the computer would be able to calculate for me better than it did.

With 60 miles to go it was continuing to drop fast and said I would arrive with 6% battery (when an hour before it had predicted 30%). It had been dropping so fast I didn’t trust that. It was 85+ outside. I set the A/C to 72, kept the windows up. Drove at 50 mph. Turned “Chill Mode” on. Once I did all of these things the battery leveled out and I arrived at the charger with 7%.

Sorry for the long story, but my question is, has someone figured out a formula or algorithm that would eek out the greatest battery life in a situation like that? Or is it as simply as drive slower. 5mph and you can go forever? Or if you go below a certain speed, say 35 mph, you actually use more energy? Did dropping my speed to 50mph help or is it all the same if I’m below 70mph for example?

Glad I didn’t get stuck, but curious about best practices the next time my battery is dropping more rapidly than I expect.

Thanks in advance for your thoughts.

 

[Tronguy]

Well, a few things.

Regarding windspeed past the car: Goes as the cube of the velocity. Below 40 mph or so it's all about moving the mass of the car and friction. Above 50, it's noticeable, and the difference between 55 and 70 is rather strong.

Next: The car runs algorithms. In my experience, those algorithms tend not to be, for lack of a better term, "abrupt". Say one is going to Boston from NYC. When one starts, the algorithms in the car will take a kind of a look at the outside temperature, some rough speed limits, possibly altitude changes (both up and down), and then put up an estimate of how far you'll get. If, halfway through the trip, one starts practicing abrupt halts and jackrabbit starts, the mileage is going to get worse; but it's not like five minutes of this is suddenly going to be applied to the next two hours of the trip. Do it for half an hour and the estimated range numbers will drop. Now, if one had started doing the brake/accelerate games right off, one would have gotten a better range estimate in the beginning.

So, the short answer is, it's possible to fool the range estimator, as you've just figured out. The one thing that I know that works when one is running out of range and one is on interstates and the like is to slow down; the forums have lots of stories about, at the car's urging, to slow down to 55. And I've seen those, "Slow down, darn it. You're running out of charge!" messages myself.

Now, when the SO and I have gotten those messages, with us being engineers, we do some quick triage to see how things are going. Pull up the energy screen and get the current average W-hr/mile; figure out how far it is to our destination; do a little multiplication and subtraction and work out Just How Bad Is It, Anyway? (This tends to be especially fraught when traveling in cold weather in the 2018 M3, where it's possible to get a 350 W-hr/mile rather than the 225 W-hr/mile one sees on warm summer days.) In those cases, we figured we didn't need to panic, although we got into the next charging stop with one to several percent charge left.

As an example: Suppose we left for Boston with the outside temperature at 25F. On that particular M3, the initial W-hr/mile would be in the 350 range. As the car warmed up this would drop; after an hour of driving, it would be down to 300 or so, a considerable range improvement. But then it would freak out about not making it to the next charging stop - but it was basing the panic on the average W-hr/mile, not the new, lower value.

(Mind you, your car and all newer Teslas have a heat pump built in which makes these kinds of problems not as much of an issue: The car warms up and gets low W-hr/mile within ten or fifteen minutes of getting on the road.)

Actually.. If you didn't get the, "Slow down, already!" messages, I suspect that all your messing about with the A/C, chill mode, and what-all didn't really change your energy usage. The car knew the distance to your charging stop; knew what you were doing in terms of driving; and likely didn't see a problem with you showing up with 7% charge. The "rapid decay" you saw was the range estimator algorithm slowly catching up with reality.

Now, if you want to see the Reality: Suggest that you look at the Energy display. That display is, if I'm not mistaken, a snapshot of how far the car really thinks you're going to go.

 

[gad1976]

Thanks this helps. Two questions for clarification:
1. Since the car didn’t give me a message telling me to “slow down” you think I was probably ok.
2. It sounds like if I’m needing to “save” I need to drop below 50. The slower I go, the more I save?

I was using the cars range estimator and it seemed to work correctly. I just didn’t know if I could trust it since I was suprised at how quickly the range was degrading. But my experience today suggests its trustworthy and it sounds like in your experience it has been useful.

Thanks!

 

[RHinCT]

My only thought is that going up the mountains is a killer, but coming back down can recover quite a bit. Just don't get to the point of needing to use the actual brakes; drive to keep regen slowing you down. Green at the top of the screen is good.

 

[gad1976]

Is there any value in this from Teslify? If I were to believe this traveling at 65 mph and 45 mph doesn’t make much of a difference. I like that data from Teslify, but not sure if it is really that useful.

 

[father_of_6]

With 60 miles to go it was continuing to drop fast and said I would arrive with 6% battery (when an hour before it had predicted 30%).

The range estimator was off by a quarter of your entire battery. I recently had a similar experience near ADK mountains in NY... pretty sure it was the elevation changes. The estimator said one thing, but reality was much lower. Usually it's fairly accurate. I slowed down a bunch, turned the climate controls down a bit.

Not certain about this, but I assume that the estimator doesn't have (or use) elevation data in it's original estimation, but perhaps the vehicle has an accelerometer to detect elevation changes... because it does show how much range was lost to elevation in the energy usage screen.

My take is... if it originally estimated that you'd get to your destination with 20%, you would not have made it (because it was wrong by 24%). This is a problem that Tesla needs to address.

Although I'm over the "range anxiety" and have not yet had the unfortunate experience of running out of juice (crossing fingers that never happens), it's much more complicated than running out of gas. Can't walk to a gas station and back with a can of electricity.

 

[father_of_6]

It sounds like if I’m needing to “save” I need to drop below 50. The slower I go, the more I save?

Yes. Here's a relevant quote:

The faster one travels, the more fuel is needed to overcome aerodynamic wind resistance. And wind resistance increases as a squared value, 45 mph using half the energy to maintain vs 62 mph. It takes almost triple the energy to maintain 75 mph as it does 45 mph.

Source:

A Guide To Fuel Efficient Driving - Part Two - CleanTechnica

You can drive your fossil fuel powered vehicle more efficiently to save fuel, money and reduce carbon pollution.

cleantechnica.com

 

[Tronguy]

Thanks this helps. Two questions for clarification:
1. Since the car didn’t give me a message telling me to “slow down” you think I was probably ok.
2. It sounds like if I’m needing to “save” I need to drop below 50. The slower I go, the more I save?

I was using the cars range estimator and it seemed to work correctly. I just didn’t know if I could trust it since I was suprised at how quickly the range was degrading. But my experience today suggests its trustworthy and it sounds like in your experience it has been useful.

Thanks!

As regards #1: Yep, you were probably OK. The "range estimate" that's up at the top of the screen is actually pretty rough in its accuracy. Yeah, if you're bombing down I-95 from Washington DC, to Miami, FL at a steady 70 mph (or something) it'll be pretty accurate. But start playing, "I'm on a freeway! No, I'm on local roads! No, I'm going up a mountain! No, I'm going down!" and that indicator is going to get fouled. In the scheme of Smart Things, it's on the dumb side.

A better one, as I mentioned, is the Energy display, available from that little batch of icons one can find by hitting that "one icon brings up a window of a bunch of 'em" at the bottom of the screen. It'll show your instantaneous W-hr/mile, the effects of your personal driving on range, and a half-dozen other things. If you wanna have better data for, "Am I going to get there?", that's your one-stop shop.

But even that one doesn't have your actual trip dialed in. And, interestingly and importantly, the main, "Am I going to run out of charge?" algorithm has elevation variations, outside temperature, headwinds, and your actual expected path (extracted, thank-you-very-much, from the Navigation) all factored in for your driving enjoyment. I.. don't think it has actual rainfall as a parameter, but give the maniac programmers over at Tesla some time .

Regarding #2: Say you're going 2 mph for some nutso reason. As the car rolls forward, ye tire sidewalls flex, energy is lost in copper wires, the heat from all that lost energy has to be gotten rid of somehow, bearings gotta turn, etc., etc. All that stuff tends to go with car velocity and, pretty much, is linear: Go 4 mph, and the losses will be 2x the losses at 2 mph. All these effects are termed, "rolling resistance".

Next: Say you're going 2 mph. Pushing aside the air at that speed doesn't use up much energy. Likewise, 4 mph. But the People Who Really Know This Stuff (mostly, airplane designers, people who build guns for a living, rocket scientists, and the aerodynamicists at car companies) all claim that, roughly, the energy to push the air aside goes as the square of the velocity. I'm no expert: I just took the Amusing Undergraduate Physics that one takes as an engineer, where they touch on friction a bit and explains that one doesn't get parabolas when one tosses a rock into the air, but something that, well, approximates the shape of a female breast. With some of the math that goes with it. In any case, back to cars: Roughly at 40 or so MPH or so, air resistance gets up to the same magnitude as the rolling resistance. And, I'm a telling you, when something goes as the "square", it's steep. As others have pointed out on this thread, the difference between the energy required to defeat the wind resistance at 45 mph and 70 mph is 3x. So, yeah, slowing down helps.. a lot.

So, when one gets down to it, rolling resistance and wind resistance are both side effects of friction. Both go up with speed, rolling resistance more-or-less linearly, wind resistance as the square. So, you mentioned that you put your car on, "Chill Mode". Well... that has a bit of an effect, but probably not as much as you think.

The energy in a moving object (gad, I still remember this, it's been too many years!) is K = (1/2)*m*v^2. So, suppose the car's at a stop light (K = 0). Then, one speeds up to 100 kph (about 60 mph), which is about 27.8 meters/s. A model 3's mass is about 9000kg, so the energy in the car at 100 kph is about
27.8^2 * 9000/2 = 3.477 MJ. Ha. There's 3.6 MJ per kW-hr (A kW is 1000 Joules per second; an hour has 3600 seconds), so, the amount of battery energy to get from 0 to 60, for just the kinetic energy, is about 3.477/3.6 = 0.966 kW-hr's. Actually, it'll be more than that, since neither the battery, the motor, nor the electronics is 100% efficient; so let's say that every 0 to 60 mph screetch-the-wheels getaway uses up 1.0 kW-hr of your battery's storage. And it's probably more than that.

Now, when you stop, I dunno. Yes, the car uses regenerative braking, but that process isn't wildly efficient, either; maybe, I make a wild guess, 60% of the Kinetic energy can be recovered. On the Prius I used to drive, some smarter person than I was estimated that only 20% of the regen energy made its way back to the battery, so I could be 'way off. The energy that got put into rolling and wind resistance? Gone, gone, gone.

So, if you're on a road that has a stoplight per mile and a 60 mph speed limit, and you hit all the lights for 100 miles, your range is going to suffer. This is why jackrabbit starts aren't good for anybody. Regen helps, but it's not perfect. (Still better than an ICE which has no regen whatsoever, and the Laws of Physics Apply To Those Cars, Too.)

The next bit, though, is the question of Chill vs. Non-Chill. Non-Chill has.. more acceleration. The energy difference in a moving vehicle between two speeds, any two speeds, is Kdiff = (1/2)*m*(v1*v1 - v2*v2). Um. Do you see an "a = acceleration" term in there? I sure don't. There probably is some 2nd order effect involving acceleration, but first order effects, nada. (That's probably not true for an ICE: They're notoriously inefficient once one gets off their sweet spot, unlike BEVs, which maintain efficiency over a wide range of motor speeds.)

So, switching to Chill probably did little to nothing for you .

 

[Brian-MS90D]

Thanks this helps. Two questions for clarification:
1. Since the car didn’t give me a message telling me to “slow down” you think I was probably ok.
2. It sounds like if I’m needing to “save” I need to drop below 50. The slower I go, the more I save?

I was using the cars range estimator and it seemed to work correctly. I just didn’t know if I could trust it since I was suprised at how quickly the range was degrading. But my experience today suggests its trustworthy and it sounds like in your experience it has been useful.

Thanks!

Yes, #2. That's it. Everything else is insignificant (even mountains do not matter unless your destination is on top of one).

For range, speed is the only factor of any significance you have control over. And, as you go faster and faster then anything else REALLY doesn't matter.

It is true you will see a hit in range during winter, but you cannot control the weather. You can control your speed and slow down if needed.

 

[NYMY]

Yes, #2. That's it. Everything else is insignificant (even mountains do not matter unless your destination is on top of one).

For range, speed is the only factor of any significance you have control over. And, as you go faster and faster then anything else REALLY doesn't matter.

It is true you will see a hit in range during winter, but you cannot control the weather. You can control your speed and slow down if needed.

One thing mentioning is that if you got nervous in the mountains, it might have been near the peak? When I'm driving in the mountains I often will actually gain percentages when I'm coming down and so my percentage can be frighteningly low on the way up. For this to be effective, however, you really have to use one-pedal driving and make sure you're doing regenerative breaking...

 

[kpanda17]

driving in winds, mountains, bad weather, high and low temps, only run batteries down to 20% or more

BMS should predict and adjust NAV results accordingly

 

[gtae07]

Had an odd experience going to the Atlanta area last week pulling a bike rack. We charged in Dublin, adding about 18% beyond the "you have enough to get to your destination" message so we'd have margin (I wanted 20-25% on arrival). About 60 miles later in north Macon, I had to pull over for about 3 minutes to resecure a bike, and when I got back in the car the estimate for percent remaining on arrival dropped to 11% from 22%. That was a bit alarming, and I wish I understood how it could change so dramatically like that. But, some traffic, slowing down a bit, and chill mode brought us to 15% at the end.

 

[kpanda17]

BMS has to learn to understand changes to your vehicle in order to predict the future range

Have you ever driven with a bike rack before?
Anything connected to the outside of rhe vehicle that is not in the original design affects range prediction by the BMS

 

[gtae07]

BMS has to learn to understand changes to your vehicle in order to predict the future range

Have you ever driven with a bike rack before?
Anything connected to the outside of rhe vehicle that is not in the original design affects range prediction by the BMS

Other than the 160 interstate miles I was driving right before that, no (I've only had the car a few weeks). But by the time we were 50 miles into the trip the prediction was apparently good, and was accurate all the way to the charge stop at 100 miles. I would think that by that point it would have gotten the idea.

Does putting it in Park cause a reassessment of the battery that doesn't happen while driving?

 

[gad1976]

As regards #1: Yep, you were probably OK. The "range estimate" that's up at the top of the screen is actually pretty rough in its accuracy. Yeah, if you're bombing down I-95 from Washington DC, to Miami, FL at a steady 70 mph (or something) it'll be pretty accurate. But start playing, "I'm on a freeway! No, I'm on local roads! No, I'm going up a mountain! No, I'm going down!" and that indicator is going to get fouled. In the scheme of Smart Things, it's on the dumb side.

A better one, as I mentioned, is the Energy display, available from that little batch of icons one can find by hitting that "one icon brings up a window of a bunch of 'em" at the bottom of the screen. It'll show your instantaneous W-hr/mile, the effects of your personal driving on range, and a half-dozen other things. If you wanna have better data for, "Am I going to get there?", that's your one-stop shop.

But even that one doesn't have your actual trip dialed in. And, interestingly and importantly, the main, "Am I going to run out of charge?" algorithm has elevation variations, outside temperature, headwinds, and your actual expected path (extracted, thank-you-very-much, from the Navigation) all factored in for your driving enjoyment. I.. don't think it has actual rainfall as a parameter, but give the maniac programmers over at Tesla some time .

Regarding #2: Say you're going 2 mph for some nutso reason. As the car rolls forward, ye tire sidewalls flex, energy is lost in copper wires, the heat from all that lost energy has to be gotten rid of somehow, bearings gotta turn, etc., etc. All that stuff tends to go with car velocity and, pretty much, is linear: Go 4 mph, and the losses will be 2x the losses at 2 mph. All these effects are termed, "rolling resistance".

Next: Say you're going 2 mph. Pushing aside the air at that speed doesn't use up much energy. Likewise, 4 mph. But the People Who Really Know This Stuff (mostly, airplane designers, people who build guns for a living, rocket scientists, and the aerodynamicists at car companies) all claim that, roughly, the energy to push the air aside goes as the square of the velocity. I'm no expert: I just took the Amusing Undergraduate Physics that one takes as an engineer, where they touch on friction a bit and explains that one doesn't get parabolas when one tosses a rock into the air, but something that, well, approximates the shape of a female breast. With some of the math that goes with it. In any case, back to cars: Roughly at 40 or so MPH or so, air resistance gets up to the same magnitude as the rolling resistance. And, I'm a telling you, when something goes as the "square", it's steep. As others have pointed out on this thread, the difference between the energy required to defeat the wind resistance at 45 mph and 70 mph is 3x. So, yeah, slowing down helps.. a lot.

So, when one gets down to it, rolling resistance and wind resistance are both side effects of friction. Both go up with speed, rolling resistance more-or-less linearly, wind resistance as the square. So, you mentioned that you put your car on, "Chill Mode". Well... that has a bit of an effect, but probably not as much as you think.

The energy in a moving object (gad, I still remember this, it's been too many years!) is K = (1/2)*m*v^2. So, suppose the car's at a stop light (K = 0). Then, one speeds up to 100 kph (about 60 mph), which is about 27.8 meters/s. A model 3's mass is about 9000kg, so the energy in the car at 100 kph is about
27.8^2 * 9000/2 = 3.477 MJ. Ha. There's 3.6 MJ per kW-hr (A kW is 1000 Joules per second; an hour has 3600 seconds), so, the amount of battery energy to get from 0 to 60, for just the kinetic energy, is about 3.477/3.6 = 0.966 kW-hr's. Actually, it'll be more than that, since neither the battery, the motor, nor the electronics is 100% efficient; so let's say that every 0 to 60 mph screetch-the-wheels getaway uses up 1.0 kW-hr of your battery's storage. And it's probably more than that.

Now, when you stop, I dunno. Yes, the car uses regenerative braking, but that process isn't wildly efficient, either; maybe, I make a wild guess, 60% of the Kinetic energy can be recovered. On the Prius I used to drive, some smarter person than I was estimated that only 20% of the regen energy made its way back to the battery, so I could be 'way off. The energy that got put into rolling and wind resistance? Gone, gone, gone.

So, if you're on a road that has a stoplight per mile and a 60 mph speed limit, and you hit all the lights for 100 miles, your range is going to suffer. This is why jackrabbit starts aren't good for anybody. Regen helps, but it's not perfect. (Still better than an ICE which has no regen whatsoever, and the Laws of Physics Apply To Those Cars, Too.)

The next bit, though, is the question of Chill vs. Non-Chill. Non-Chill has.. more acceleration. The energy difference in a moving vehicle between two speeds, any two speeds, is Kdiff = (1/2)*m*(v1*v1 - v2*v2). Um. Do you see an "a = acceleration" term in there? I sure don't. There probably is some 2nd order effect involving acceleration, but first order effects, nada. (That's probably not true for an ICE: They're notoriously inefficient once one gets off their sweet spot, unlike BEVs, which maintain efficiency over a wide range of motor speeds.)

So, switching to Chill probably did little to nothing for you .

Thanks for this detailed answer. Very helpful!

 

[mongo]

Minor clarifications

A model 3's mass is about 9000kg, so the energy in the car at 100 kph is about
27.8^2 * 9000/2 = 3.477 MJ. Ha. There's 3.6 MJ per kW-hr (A kW is 1000 Joules per second; an hour has 3600 seconds), so, the amount of battery energy to get from 0 to 60, for just the kinetic energy, is about 3.477/3.6 = 0.966 kW-hr's.

Model 3 curb weight is 1850 kg, call it 2000 kg with passengers. So 0.77 MJ or 214 Wh net energy to accelerate.

Regarding windspeed past the car: Goes as the cube of the vvelocity.

Power goes up as the cube, energy per mile goes up as the square.

 

[Geech72]

Imagine how stressed out Ethan Allen and his boys were when they made their move on fort Ticonderoga; there were far fewer chargers to use back then.

 

[gad1976]

Imagine how stressed out Ethan Allen and his boys were when they made their move on fort Ticonderoga; there were far fewer chargers to use back then.

Haha. Not sure how they made it.

 

[avs007]

Other than the 160 interstate miles I was driving right before that, no (I've only had the car a few weeks).

This is the main thing going on... The car is still learning... When I first got my Y, I immediately went on a road trip to California... The car was pretty off with the range predictions when I was driving thru Shasta. But after a few months (I drive a LOT), when I drove to Canada, and back to California, the BMS was pretty spot on with the estimates, even when going thru the Mountains, which was pretty good, because I was driving around 80 going thru the mountains.

 

[bruce4000]

Before a road trip I use an ap like “A Better Route Planner” to get an idea of where to charge and how close I’m cutting it. Its usually pretty accurate. However the older version allowed you to add things like headwind to see how much would affect it but now just says will use “seasonal weather “?