Hypermiling techniques?

[bxr140]

But the width of the box is less, so the area is the same. E=1/2 Mv^2 - it doesn't matter from a pure energy standpoint how fast or slow you get there.

Thats true, but a bit misleading in context. You're correct that the energy use for the acceleration events is basically the same, but the slower accelerating car will use less energy over equal timespans, which is how most people (I believe) are considering this scenerio with respect to hypermiling. If a fast car accelerates to speed in 10 seconds and a slow car does it in 20, the fast car uses more enery at the 20 second mark since its average speed over that 20 seconds is higher.

That higher average speed also contributes to more aerodynamic losses over the same period of time and as you point out the electrical losses from the faster acceleration are also higher. That all adds up into what most people will intuitively suspect, which is that you use less juice driving with a light foot.

 

[billarnett]

... That all adds up into what most people will intuitively suspect, which is that you use less juice driving with a light foot.

But the main reason why driving with a heavy foot uses more energy is that it usually results in a higher average speed. The "fair" way to think about this is to imagine two techniques of traveling a fixed distance in a fixed time (i.e. at a given average speed). The first driver accelerates rapidly up to slightly more than the required average speed and holds that for the remainder of the test. The second driver accelerates more slowly but in order to end the test at the required average speed he must end up at a slightly higher speed. Which one will use less energy? It's not so clear. Of course, that's NOT how people really drive. The terminal speed after the acceleration is usually set by the speed limit and that's the same for both drivers. So the first driver who accelerates faster also arrives at the destination earlier. And in the real world a driver who likes to accelerate faster probably also likes to drive faster in general. It's not the acceleration that eats the extra kWhs, it's the speed. (Mostly. There are extra losses for higher power, but my intuition is that they're relatively minor.)

A similar thing happens if the test involves stopping at the end. The first driver keeps going as long as possible then does max regen down to zero. The second driver starts coasting at the right time and just manages to stop at the finish line (hard to do But again, the second driver will have to drive the middle section of the test at a higher speed.

 

[MaturinNYC]

What's the explanation behind this?

I wish i could remember precisely (I'm all right brain), but the gist was that in neutral there are some vents that suspend their function - so the advice was to just rely on regen and keep all systems operating as designed. As to the logic behind throwing it in neutral, that was a habit i picked up trying to coax an extra few miles over a 90 mile commute in a MINI E that was forever threatening to shut down on the hills of the Northeast. It worked well in that beloved "mule", which i doubt had any venting sensitivity.

 

[bhzmark]

Coasting with as little drag as possible is far and away the most efficient way to hyper mile. The vent thing makes no sense. Usually coasting can't be for too long anyway because soon you have speed up or slow down so any vent issue wld be very short term.

 

[Cottonwood]

Coasting with as little drag as possible is far and away the most efficient way to hyper mile. The vent thing makes no sense. Usually coasting can't be for too long anyway because soon you have speed up or slow down so any vent issue wld be very short term.

Long mountain descents can often end with more energy in the battery with regen than with coasting. Take an extreme example, I can coast at a terminal velocity of 90 mph and add nothing to the battery, or I can add 3 rated miles to the battery regening at 60 mph. Ending a descent with 3 more rated miles in the battery is better for me.

 

[ken830]

Long mountain descents can often end with more energy in the battery with regen than with coasting. Take an extreme example, I can coast at a terminal velocity of 90 mph and add nothing to the battery, or I can add 3 rated miles to the battery regening at 60 mph. Ending a descent with 3 more rated miles in the battery is better for me.

I agree... A rule-of-thumb I use is -- Anything over 70 mph, and you're probably better off regen-ing down your speed. The loses due to drag are higher than the inefficiencies in the charging/discharging system.

 

[bhzmark]

Yes, the edge case proving the general rule.

When you are faced with a mountain descent that could accelerate you to a very high speed, the aero drag could overtake the battery regen inefficiency.

The rest of the time (i.e., nearly all the time) coasting is better.

With the corollary that you should only accelerate to a speed from which rolling resistance and aero resistance will decrease your speed sufficiently when you need to slow or stop (in other words -- coast). Accelerating under power beyond that speed can only result in energy loss (due to regen inefficiency or worse, making heat and brake dust).

 

[Cottonwood]

Yes, the edge case proving the general rule.

When you are faced with a mountain descent that could accelerate you to a very high speed, the aero drag could overtake the battery regen inefficiency.

The rest of the time (i.e., nearly all the time) coasting is better.

Depends on your definition of a mountain. My gut from watching lots of hills in my MS is that the trade off is at about 100-150 meters (333-500 ft) vertical. 500 ft vertical is just another hill around here in Colorado, and so in my driving, regen often wins over coasting. :biggrin:

 

[bp1000]

Sorry if this is going to sound really dumb but does hypermiling apply to all electric vehicles?

Gas engines are terribly inefficient, like 20-30% of all fuel combusted is converted into energy. So guys started measuring how to drive so you always maximised energy from fuel burn.

As I recall ev are nearly 100% efficient, but actually around 80% when you factor in charging losses.

So as long as you don't break traction and loose energy through the spinning wheels or request too much power for your target speed it's going to be the most efficient driving for your route.

I assumed that providing you don't mash the throttle, loose traction, request a big swell of power or completely ignore regen you are always going to be driving with max efficiency.

Eg I also assumed there would be no difference in energy consumed between accelerating slowly to 60 or normally to quite fast providing you dont break traction or overshoot your speed.

Hill driving I don't have knowledge on, I'd assume like with any other type of driving you would want to minimise changes in speed, keeping a steady speed. If you bleed off speed no where near the top you are only going to have to recover it until the top. I would assume it's more efficient to accelerate before a hill and keep a steady speed going up, bleeding off before the crest.

in all other driving scenarios you want to effectively keep a steady speed which will likely result in long periods of coasting on the flat or marginally downhill gradients below say 45-50mph. I would guess you don't want to aim for 0w consumption if you are only going to have to draw energy to make up the mph you loose, typically at higher speeds... Unless you slow down soon.

 

[Robert.Boston]

Absolutely, hypermiling applies to EVs and, in particular, to the Model S. Because the range is more limited than in typical ICE vehicles, and because it takes longer to refuel while underway, there's a potentially big payback to using your battery wisely.

An obvious and easy technique unique to EVs is using regenerative v. friction braking. Any time you have to touch the brake pedal, you've thrown away energy. So Hypermiling 101 says, learn to drive with one foot. The only times I've found when the friction brakes are necessary are (1) emergency maneuvers, (2) when regen isn't available because of cold, and (3) to slow the car from 4 mph to 0 for a full stop (and not even then, always; if there's a slight upward slope, you can usually stop without brakes and then simply use the brake to hold the position).

As the discussion above noted, there is a difference in energy used by accelerating quickly or slowly, simply arising from the simple fact that the average speed will be higher if you get to the cruising speed more quickly.

There are also higher internal losses in the power electronics when you push higher current flows. Notionally, you will optimize those losses by maintaining constant power to the motor, i.e. keeping the kW meter pegged and letting the speed rise or fall with the terrain. (As a practical matter, you won't be able to do this safely, but the principle is correct.) Thus, you can get better Wh/mile with the CC off by allowing the car to slow while driving up slopes and accelerate while going down. This "constant kW" strategy will be more energy efficient than the accelerate-and-coast strategy you suggest (assuming the same average speed) or driving at a constant speed using the CC.

 

[ItsNotAboutTheMoney]

Long mountain descents can often end with more energy in the battery with regen than with coasting. Take an extreme example, I can coast at a terminal velocity of 90 mph and add nothing to the battery, or I can add 3 rated miles to the battery regening at 60 mph. Ending a descent with 3 more rated miles in the battery is better for me.

You're now at 90mph at the bottom of the mountain and have more momentum than you would have with regen. You can now glide and use your momentum to travel farther. The problem with this is that your burning off more energy to drag than you would at 60mph.

To a hypermiler the preferred option is not regen, it's to be at a lower speed when you crest the mountain so you can glide and accelerate downhill.

 

[bp1000]

Thank you Robert, really interesting to learn about these new motors and how they work.

I hadn't realised there would be differences in energy used in the scenarios mentioned but as i read your explanation it made sense.

 

[AmpedRealtor]

If you allow the car to slow down gradually as you go uphill by limiting your power input, and reserve your acceleration events for when the car is going downhill, you should see a significant improvement in efficiency. You are using less energy on the uphill and gravity is assisting your acceleration on the downgrade.

 

[David99]

Almost every hypermiling technique can be reduced to the same thing: drive slow. Not accelerating fast but slowly means you are going slower overall, getting off the accelerator earlier and letting the car coast is overall going slower. The range vs speed curve on Tesla's blog is show that your most efficient speed is around 25 mph. If you want to be the perfect hypermiler, just got 25 all the time

 

[billarnett]

Almost every hypermiling technique can be reduced to the same thing: drive slow. Not accelerating fast but slowly means you are going slower overall, getting off the accelerator earlier and letting the car coast is overall going slower. The range vs speed curve on Tesla's blog is show that your most efficient speed is around 25 mph. If you want to be the perfect hypermiler, just got 25 all the time

That's all well and good but I did not buy a car with 691 horsepower in order to drive 25 mph !

 

[Panu]

That's all well and good but I did not buy a car with 691 horsepower in order to drive 25 mph !

Using those horse powers is not hypermiling, we all know that. But it's very important to know that when you need to extend your range as much as possible the optimal speed is 25 mph.

 

[hiroshiy]

Nothing scientific, but from my experience, these are the points I need to follow to reduce power usage, in order of priority :
1. Slow down.
2. In long distances, try to draft a truck or a bus to reduce wind resistance. Same effect as 1
3. Do not use regen when you can avoid using it. For example try keep more space between cars and delay starting of regen, and try to keep the throttle on the 0kW line as long as possible. To reduce speed, use regen not brakes.

Especially in crowded places like Tokyo, technique #3 works very well I believe.

 

[RDoc]

No, the X is the distance traveled--not the time it takes.

If X is the distance traveled, then Y is the Force. Both fast and slow acceleration are boxes, neither is sloped as fast acceleration means a horizontal line with greater Y than slower acceleration, but both are flat. The formula for distance is s=V^2/(2a), so to get to a particular velocity, the distance to get there will be less if the acceleration is faster so the box is shorter in X for faster acceleration to get to a given speed and the areas are the same. If you substitute that into E=F*s you get the original formula with the a term cancelled out.

This got me thinking about how to accelerate in terms of distance traveled since we generally travel for a distance, not for a time. Assuming that there's a need to get to a higher speed, such as an entrance ramp onto a highway, there's a maximum distance available for the acceleration, in this case the merge lane. Ignoring for the moment the higher heat losses the the drive train, I think the most energy efficient way to acceleration would be to go at the minimum safe speed until you're just at the minimum distance from the merge lane end you'll require to get up to highway speed and then floor it. That way you travel at the slowest safe speed the longest distance but still are at highway speed when you finally do the merge.

If you accelerate at a constant rate to just get to highway speed at the end of the merge lane, you'll spend more time at higher speed. OTOH if you accelerate hard at the start then go at highway speed the remaining distance to the end of the merge lane you'll travel longer at higher speed in the merge lane where lower speed is safe.

The actual best rate of acceleration at the end of the merge lane is very likely less than maximum because of high current heating losses, but we'd need a lot more data to know what the trade off between heat loss and drag loss is.

 

[stevezzzz]

That's all well and good but I did not buy a car with 691 horsepower in order to drive 25 mph !

Nor did I, but it's remarkable that those extra horses don't exact a penalty in range unless you actually let them out of the barn, so to speak. My lifetime average Wh/mi in the P85D is remarkably close to my lifetime average in the Sig S85; for the miles put on this winter, only, the D is actually doing slightly better than the S. And that's with the D handicapped by having torque sleep enabled for only about half of those miles.

 

[RDoc]

Nor did I, but it's remarkable that those extra horses don't exact a penalty in range unless you actually let them out of the barn, so to speak. My lifetime average Wh/mi in the P85D is remarkably close to my lifetime average in the Sig S85; for the miles put on this winter, only, the D is actually doing slightly better than the S. And that's with the D handicapped by having torque sleep enabled for only about half of those miles.

I wonder if the P85 has lower high current heating losses. It's designed to handle higher maximum current and power output, so perhaps for a given acceleration rate less than that maximum, the P85D would actually be more efficient than the S85D. Of course there are likely other factors such as wheels and tires and gearing.