Hyper-miling, regen, and Neutral

[AudubonB]

In another thread today, wycolo wrote:

Hyper-Milers pop it into NEUTRAL all the time in pursuit of 'better than regen' efficiency. Free wheeling down hills can be had easily this way (now that cold morning starts won't return until December).

This confuses me (easy task, there). To give a real-road example, if I start with "X miles" remaining on my battery level, and then regen my way down the several miles of the Yarnell ramp, my battery level will indicate "X+3" at the ramp's bottom. But if I were to throw the S into Neutral at the top, besides having to brake many times on the way down (it's not a straight run and not even Monster Tajima could just coast-accelerate to 162mph or whatever the final velocity would be), then I certainly wouldn't be able to coast those 3 gained miles before losing speed.

So:
1. Is Wycolo correct, but that the Mexican overdrive (NPCSpeech for throwing into Neutral) beats regen only in limited conditions (straight ramp; modest duration)?

2. Is Wycolo incorrect - or rather, are those hyper-milers acting against their interest?

3. Is my example faulty?

I've owned almost exclusively mannytrannies for the past 50 years - nothing like hitting 90mph down a good long Alaskan or northern Canada ramp in a pickup truck with 26,000 lbs of cargo, campershell, trailer....., in hopes of saving a few tablespoonsful of diesel. But over the decades, it does add up

 

[David99]

The important part is not neutral vs regenerating. The issue is speed. The faster you go, the more you lose in wind drag. It is irrelevant whether you go downhill or just coast. The faster you go the more losses you have. Going downhill slower and getting regen is better in terms of energy than putting it in neutral and going faster.
Yes, regen isn't 100% efficient. But in this case it doesn't matter how efficient it is. Not doing any regen is always worse. Going slower is always better when trying to consume less energy.

Think about it this way:

if you go down hill in neutral, the car will pick up speed to a point where the wind drag balances out gravity pulling the car down hill.

If you go slower, regen is what balances the car at a certain speed. You are getting free energy that would be 100% lost in wind drag if you were to go down in neutral.

 

[AudubonB]

I agree. That is my point, but my question stands: is there any merit under any reasonable circumstances to throwing an S in Neutral rather than into regen? Do these hypermilers know something I don't?

 

[ZBB]

By the way -- here's a pic I took a couple weeks ago at the bottom of the Yarnell Grade that Audie mentions. The Yarnell Grade is an elevation change of about 2000 feet over 4 miles. Speed limit is mostly 35 mph until the bottom (I snapped the pic right as the limit went back to 65).

Notice the "since last charge" stats... After charging a bit in Yarnell, we drove about 2.5 miles around Yarnell prior to heading down the grade. The kWh readouts were going backwards during the decent -- and the Wh/mi was negative over 6.5 miles.

I'll vote for letting regen do its thing -- especially on steep descents...

 

[arg]

I agree. That is my point, but my question stands: is there any merit under any reasonable circumstances to throwing an S in Neutral rather than into regen? Do these hypermilers know something I don't?

It's extremely difficult to generalise, due to the need to consider both time and energy.

Suppose you have a long, steep hill such that freewheeling down it achieves a terminal velocity higher than your cruising speed.
If you instead keep constant speed, you will have accumulated more energy in the battery the longer the hill is, while the freewheel case just has enough stored kinetic energy to carry you a constant distance no matter how long the hill was.

Seem like a clear win for regen if the hill is long enough. BUT the freewheeling car has covered the course faster than the regen car. So to achieve the same overall journey time it could have cruised at a lower speed at some earlier time, and so the energy saved then could mean it has more in the battery than the regen car.

So to work out even this simple case needs to know the speeds throughout the journey. Given the energy cost is proportional to the square of speed, the answer may well be different depending how long you are prepared to take for the whole journey, and also whether you are actually prepared to drive the 'ideal' speeds (maybe too fast for your license in places, too slow for the annoyance of other traffic in others).

I concluded that it's just too hard to come up with a meaningful theoretical basis; the empirical results from hypermilers are probably true, but only if you meet their assumptions.

 

[SlyWombat]

The point was that when you start in the morning with a full charge (or it is too cold to regen), regen does nothing, there is nowhere for the power to go, so why "slow down", instead coast in neutral.

 

[MikeC]

This is the question I pondered driving from Vegas to Barstow starting with 175 rated miles in heavy winds. I think the tipping point is: at what point does the increased inefficiency from drag outweigh the inefficiency of regen. This is of course up until you reach the maximum speed you are willing to go - obviously at that point you will want to regen. I don't know the answer and I suspect it is a pretty complex equation with a number of unknown constants (e.g., what is the inefficiency of regen? Is it constant or does it depend on speed?) and variables (e.g., slope of hill, etc).

 

[David99]

... at what point does the increased inefficiency from drag outweigh the inefficiency of regen...

That's the wrong way to look at it. Regen can be as inefficient as you want it, even if it could only recuperate 1% it would still be better than getting nothing back.

When speed is priority, then yes going slower and doing regen is not good. If you are into hyper-miling then the priority is to use as little energy as possible. Then using regen and going slow downhill is the best thing to do.

The only reason why some people still think neutral would be good is because in ICE cars it would disconnect the engine which would just cause drag and friction and slow you down. That isn't the case with electric motor.

 

[GSP]

Neutral can be more efficient than regen when you use it to coast farther at normal or slow speeds. Instead of staying on the accelerator and then lifting to CT so that regen bring s you to a stop at an stop sign, shift to N (at a much earlier point) and coast up to the stop sign. It will take longer (make sure no one is behind you!) but ALL of your kinetic energy will be used to overcome air and rolling resistance and transferred to distance traveled. If you had used regen, only about 75% of your kinetic energy can be recovered.

Typical hypermiler assumptions are to get the lowest wh/mi no matter how much extra time it takes. If you are not willing or able to take the time, someone else will, and you never will have the best number to brag about.

Shifting to Neutral is not recommended for safety, so it would be great if Tesla offered a way to quickly adjust CT regen from zero to max. D and L (or B) shifter positions are a good way, as are steering wheel paddles.

GSP

 

[physicsfita]

There was a similar thread a few months ago which looked at a slightly different issue. The takeaway from it applies, though -- the problem is seriously complex, and you are pretty much best off resorting to doing a numerical analysis for the particular conditions you have in mind. It turns out that in addition to drag, you also have to worry about rolling friction, and the whole thing becomes a mess, with the ultimate answer being "it depends". Sigh.

 

[yobigd20]

I love getting free miles while also gaining Wh into battery pack (rates miles increased) when going downhill.

 

[jerry33]

I agree. That is my point, but my question stands: is there any merit under any reasonable circumstances to throwing an S in Neutral rather than into regen? Do these hypermilers know something I don't?

1. There is no real difference between having the car in neutral and pressing on the accelerator enough so that there is neither orange nor green showing. I call this gliding.

2. If the car is in drive you won't ever have to fumble with the shifter if the unexpected happens.

3. Going slower always beats going faster, however wind direction is a big factor.

So the rules I use are:

1. Use gliding rather than regen under normal circumstances.

2. On steep downhills use regen to limit the speed rather than speeding up and letting wind resistance create the maximum speed.

3. Use regen rather than friction brakes to stop whenever possible.

4. I never use neutral. I don't see the need to ever do so.

 

[Cottonwood]

Here are some basics:

1. High speed for long periods wastes energy to aerodynamic drag, rolling resistance, etc. Speed drag.
2. Using the chemical energy in the battery to add kinetic or potential energy to the car and then regen that energy back to the battery loses something like 20-25% in the MS in each round trip cycle. Regen loss.
3. For very short rolling hills, there is far less loss to speed drag than would be lost to taking energy out of the battery and then putting it back in; regen loss. Do the roller coaster hypermile move here.
4. On very long descents, speed loss will use up almost all of the potential energy that could be put in the battery. As has been noted, you can end up at the bottom of a long hill with more rated miles in the battery with regen, or none by freewheeling. Use the regen hypermile move here.
5. The MS will not show more rated miles in the battery until you accumulate at least 1 kWh of energy (~3.4 RM in an 85). If you look at the energy on the "trips" display, it will show you the energy gained with 0.1 kWh resolution; it's there even if not shown in rated miles. Regen loss can be less than speed loss even if you don't regain 3 rated miles and don't see anything in the rated mi
6. You are safer by feathering the accelerator pedal to zero power, and just as effective, than putting the car in neutral. Just create a virtual neutral with your right foot.
7. Figuring out at what height of a climb/descent to do the roller coaster hypermile move, or the regen hypermile move is the question. I have not done that analysis yet; let's leave that as an exercise for the reader. My guess is that it is more than 100 feet vertical, and it is certainly less than 1,000 feet vertical.

 

[GSP]

Good advice from arg, Jerry33 and Cottonwood. This "simple" physics problem does get complicated fast.

On the subject of holding the accelerator pedal at exactly the zero power point, instead of shifting to neutral. This does have the exact same desired effect, but I find it extremely difficult to do exactly, say within +/- 100 W while driving my Volt in "L."

It is especially difficult in high-G turns, when I want to keep my eye on the road and planning my line through the corner, and the G-forces make it hard to hold my foot in the exact same position. I would never consider shifting to "N" in these conditions, but the Volt in "D" gives the desired effect. If maintaining speed is an option, instead of coasting through the turn, I can just apply a little accelerator in "D." It is also possible to transfer force to the front or rear wheels with the accelerator as necessary to maintain good balance and traction in the turn.

I am constantly shifting my Volt between "D" and "L" and would really like to have this capability in the Model S.

GSP

 

[Brass Guy]

I still argue that there is a slight difference between "gliding" in D and coasting in N. There is a small amount of overhead in the drive inverter that is still supplying power to the motor, matching the rotational speed. N turns off the inverter completely.
Most likely the total energy is not great, but in a case of a long gradual slope where the car would maintain the original speed in N, you would save some energy by using N. We are talking about hypermiling here. The effect can easily be observed on the dash display while stopped and switching between D and N - give it 2 seconds.

 

[Kipernicus]

I notice that in N you still see a little orange, especially if HVAC is on but even if not you will still see a sliver of orange.
If I feather the accelerator so that I see neither orange nor green, I must be regenning a little.

Also, when going downhill, I find the car accelerates faster if I pop it into N vs hold the accelerator at 0 (or at a sliver of orange). Have you felt this (I have no data to back up my claim)?

Usually though, I don't use N and try to glide as much as possible.

 

[MikeC]

That's the wrong way to look at it. Regen can be as inefficient as you want it, even if it could only recuperate 1% it would still be better than getting nothing back.

When speed is priority, then yes going slower and doing regen is not good. If you are into hyper-miling then the priority is to use as little energy as possible. Then using regen and going slow downhill is the best thing to do.

The only reason why some people still think neutral would be good is because in ICE cars it would disconnect the engine which would just cause drag and friction and slow you down. That isn't the case with electric motor.

If regen was only 1% efficient you'd lose 99% of your kinetic energy. Unless there's a stop sign at the bottom of the hill, presumably you want some of that kinetic energy to continue forward momentum instead of accelerating and draining the battery.

 

[AC1K]

Any speed above 25MPH | 40KPH and you coming out of the range of max efficiency, so when you are in neutral rolling down a hill above 40KPH, you are in fact loosing energy you could be recapturing with regen due to wind resistance

as you can see aerodynamics is the biggest factor over 50 MPH | 80KPH, if you can avoid going over that speed you should if you are trying to conserve power.

for hyper miling in an EV, you basically just drive as close to 40kph as possible (both up and down hill), gas is different though, you try and roll as far as you can without power because you cannot create gasoline or recapture your kinetic energy.

ive tried going down hills at 80-90kph and at 40kph (in the middle of the night with no one around ) and i captured WAY more energy back going slow because i didn't loose it to drag.

 

[jerry33]

It is especially difficult in high-G turns, when I want to keep my eye on the road and planning my line through the corner, and the G-forces make it hard to hold my foot in the exact same position. I would never consider shifting to "N" in these conditions, but the Volt in "D" gives the desired effect. If maintaining speed is an option, instead of coasting through the turn, I can just apply a little accelerator in "D." It is also possible to transfer force to the front or rear wheels with the accelerator as necessary to maintain good balance and traction in the turn.

If you are doing high-G turns then there's not a lot of economy to be had

However, I don't really understand what L buys you.

- - - Updated - - -

I still argue that there is a slight difference between "gliding" in D and coasting in N. There is a small amount of overhead in the drive inverter that is still supplying power to the motor, matching the rotational speed. N turns off the inverter completely.
Most likely the total energy is not great, but in a case of a long gradual slope where the car would maintain the original speed in N, you would save some energy by using N. We are talking about hypermiling here. The effect can easily be observed on the dash display while stopped and switching between D and N - give it 2 seconds.

1. Two seconds is a lifetime in an emergency.

2. While there may be a difference, I suspect you would need laboratory grade instruments to measure it.

 

[GSP]

If you are doing high-G turns then there's not a lot of economy to be had

However, I don't really understand what L buys you...

1) Not slowing down for turns means that I don't have to accelerate back up to cruising speed afterward. That provides economy (and fun).

2) "L" in the Volt provides pure regen. If you use the brake pedal you can get some friction braking as well. ("D" provides drive with less than 200 W of regen, which is good for coasting, and safer than "N.")

GSP