Does cruise control save energy, or not?

[Cottonwood]

I agree. I've been experimenting with cruise control on basically flat terrain for a 20+ mile stretch of highway to and from work, and have noticed that with just a moderate amount of attention I can keep the orange line much steadier than cruise control will, which jitters quite a bit. My guess is that the cruise control is so finely calibrated that it tries to keep exactly the same speed, to something like a tenth of a mile per hour, over every little bump, rise, etc. which wastes energy. Having said that, I'm not sure that the difference is worth the inconvenience of the attention and the effort of feathering the pedal. Cruise control at 55mph has been getting me about 270-280 wh/m while manually I can get about 260-270 depending on the conditions I've experienced so far.

I have to respectfully disagree.

Unless the downhill puts you into a regen situation, holding a constant speed will use the lowest energy per mile over a trip, if you use the constraint of a constant travel time to achieve the same average speed over the route. Taking longer to drive the route (lower average speed) can always use less energy.

This happens because wind resistance is proportional to the square of speed and is therefore a non-conservative force; the extra that you put in going faster is not made up for going slower.[SUP]*[/SUP]

Let's take a simple example with plausible numbers. Assume a route that is 2 miles long and at a speed of 60 mph, further assume that the energy to overcome wind resistance is 200 Wh/mi at 60 mph. If I drive the route at 60 mph, wind resistance consumes 2mi*200Wh/mi or 400 Wh. If I drive the first segment at 45 mph (3/4 of 60), that segment will take 1:20 and use (3/4)[SUP]2[/SUP] or 9/16 of 200 or 112.5 Wh. To drive the second mile in 0:40 (2:00 total), I need to drive 90 mph or 3/2 of 60 mph. That will consume (3/2)[SUP]2[/SUP]*200 or 450 Wh for a total of 562.5 Wh.

This is just one example, but in this case, I used 400 Wh for 2 miles or an average of 200 Wh/mi for aero losses driving a constant 60, to take the same amount of time, I could drive the first mile at 45 and second mile at 90, for an average speed of 60 and an average energy for aero losses of 281 Wh/mi. Try other cases, but you can't beat a constant speed for minimum energy use.

There is a counter argument that because of resistive losses in the battery, inverter, and motor, which are proportional to the square of current, it is less efficient to run at higher powers. This is true, but at a very small level. Resistive losses are in the single digit percents or lower at moderate hill climbing powers. That is tiny compared to aero losses that are well over half of the energy used at highway speeds.

Slowing down will use less energy, but the best gain is had by slowing down to a constant speed and not just slowing down on the uphills.

Enjoy your driving!



[SUP]*[/SUP]This often causes confusion. Wind resistance is proportional to the square of speed, but the power to overcome wind resistance is proportional to the cube of speed. This happens because power is force times speed and energy is force times distance.

 

[KurtR]

I have to respectfully disagree.

Unless the downhill puts you into a regen situation, holding a constant speed will use the lowest energy per mile over a trip, if you use the constraint of a constant travel time to achieve the same average speed over the route. Taking longer to drive the route (lower average speed) can always use less energy. This happens because wind resistance is proportional to the square of speed and is therefore a non-conservative force; the extra that you put in going faster is not made up for going slower.

I think you're changing the parameters of the original question in the thread, and in the process backing up my point. There is no "constraint of a constant travel time to achieve the same average speed." Not using cruise control lets you maintain more of a constant energy output while letting speed decrease up a hill, increase down it, and then temporarily increase energy output only to make up for the net loss in speed over the hill. Because, as you say, wind resistance is proportional to the square of speed, you save energy because your average speed is slightly lower than it would be on cruise control.

The original post asked whether it was wasteful of energy to maintain a specific speed, and thought it might be more efficient to let the car slow over a hill and accelerate back to speed on the other side. I think that's right. As jerry33 noted, "steady power beats steady speed" for efficiency, and my point was that even over what seems like a very flat surface, I've been able to beat cruise control efficiency at 55mph by holding the orange power line at a point that essentially drives at 55mph but gently accelerating if I see 54 and letting up slightly back at 55.

Granted, this is anecdotal and based on only about 400 miles of driving so far, but it makes sense and also tracks what I experienced in my last car where I could get approximately 75-80 mpg highway mileage by using the same technique in the original Honda Insight, but which did not have cruise control. (Loved that car, by the way - if Honda still made them I very well might not be driving a Tesla right now).

 

[wycolo]

@FlasherZ:

If one is hypermiling one would never just hit RESUME but rather bring it ever so slowly up to ~original speed and then hit RESUME.
--

 

[MorrisonHiker]

@FlasherZ:

If one is hypermiling one would never just hit RESUME but rather bring it ever so slowly up to ~original speed and then hit RESUME.
--

I don't have a Tesla yet but that's the way I've always used cruise on my old Honda CR-V. If I leave the cruise engaged, it will often slow down by 5 MPH or more on steep hills. It then revs downshifts and revs really high to bring the speed back up. To avoid that, I typically disengage cruise on very steep hills or if there's a lot of traffic. I then manually bring the speed back up. It really keeps the revs down and helps with the gas mileage. I've averaged 26.9 mpg over the past 190,000 miles and it was only rated at 23 highway.

Hopefully Tesla will introduce an ECO mode for those of us who would appreciate it.

 

[Panoz]

Whoa - how about having an ENERGY cruise control? That is, set the kW per mile you want to use and the car maintains that energy usage within a certain speed range? That way it might slow down a bit on an incline, and regen on a decline. Set the energy you want to use, not the speed. How about it?

 

[Cottonwood]

I think you're changing the parameters of the original question in the thread, and in the process backing up my point. There is no "constraint of a constant travel time to achieve the same average speed." Not using cruise control lets you maintain more of a constant energy output while letting speed decrease up a hill, increase down it, and then temporarily increase energy output only to make up for the net loss in speed over the hill. Because, as you say, wind resistance is proportional to the square of speed, you save energy because your average speed is slightly lower than it would be on cruise control.

The original post asked whether it was wasteful of energy to maintain a specific speed, and thought it might be more efficient to let the car slow over a hill and accelerate back to speed on the other side. I think that's right. As jerry33 noted, "steady power beats steady speed" for efficiency, and my point was that even over what seems like a very flat surface, I've been able to beat cruise control efficiency at 55mph by holding the orange power line at a point that essentially drives at 55mph but gently accelerating if I see 54 and letting up slightly back at 55.

You are not quite understand my argument. If you take a route where you slow down for up hills and speed up on down hills, I can find an average speed that covers the route in the same total time; travel that average speed continuously over the route, using more power on the up hills, less on the downhills; get to the destination in the same amount of time; and use less total energy for the trip.

Up hill and down, as long as you don't go into the regen mode on the down hills, constant speed will always win over constant power, as long as you put in the constraint of the same average speed over the route. In short, steep, rolling hills, avoiding regen and coasting on the downhills can be more energy efficient than constant speed, but on long hills (my Colorado world), constant speed wins again, because of the severe loss in efficiency speeding down hills.

 

[Khatsalano]

The TACC definitely applies the brakes and regenerative braking.

The TACC is less efficient than I am. It it speeds up and slows down more abruptly than I would. It also doesn't seem to like coasting and I naturally just do because of previous ICE influences. I find that with TACC, I'm averaging 325 Wh/mi on my average 50-mile round-trip commute during the week in my S85D. If I don't use TACC and exclusively manually control the car while paying attention, I can get to work with about 307 Wh/mi easily ... sometimes even less.

I have a loaner P85D right now for 5 days total because my car is in for some warranty service. Same commute from San Mateo to SF, I'm getting 344 Wh/mi using the TACC, but 328Wh/mi using manual control.

- K

 

[jerry33]

Up hill and down, as long as you don't go into the regen mode on the down hills, constant speed will always win over constant power, as long as you put in the constraint of the same average speed over the route. In short, steep, rolling hills, avoiding regen and coasting on the downhills can be more energy efficient than constant speed, but on long hills (my Colorado world), constant speed wins again, because of the severe loss in efficiency speeding down hills.

If you say so, but that doesn't match my experience. Hands up if you use CC or TACC and get a low Wh/mi.

 

[thegruf]

This is an age old discussion really, ICE or EVs, the difference really is that CC cannot anticipate.

A driver normally will anticipate what is ahead, and slow down rather than braking for a slow vehicle or curve ahead, maybe allow the speed to increase a little downhill on an undulating road, then allow the speed to bleed off on the next incline.

So a driver does not typcially maintain an exact speed, and can improve economy based on anticipation, CC is (mostly) dumb and just sets an exact speed.
So called "eco" cruise typically just permits a few more mph variance from set speed, but can increase economy based on this, but can be annoying for following vehicles.

 

[bob_p]

The cruise control software could do more to improve energy efficiency while the cruise control is on.

Allowing a larger margin of speed differential around the set speed could help - such as eliminating unnecessary slowing down or speeding up when going over most small hills and overpasses.

The navigation map may also have relevant information that could help - such as upcoming curves and hills - that could help the CC software anticipate what's ahead - and even the real-time traffic data might provide some help with anticipating what's ahead.

With the 6.2 navigation software, there could be more assistance in adjusting speeds when trying to stretch energy to reach the next charger.

 

[MorrisonHiker]

This is an age old discussion really, ICE or EVs, the difference really is that CC cannot anticipate.

A driver normally will anticipate what is ahead, and slow down rather than braking for a slow vehicle or curve ahead, maybe allow the speed to increase a little downhill on an undulating road, then allow the speed to bleed off on the next incline.

Looks like Ford is already working on GPS assisted, predictive cruise control: http://www.greencarreports.com/news...-logic-can-boost-cruise-control-mpg-5-percent

Surely Tesla could add such features without the need for any additional hardware.

 

[Cottonwood]

If you say so, but that doesn't match my experience. Hands up if you use CC or TACC and get a low Wh/mi.

I did not say that TACC or CC was more efficient than manual driving. What I am asserting and attempted to explain via a simplified aerodynamic drag example, is that constant speed is more efficient than constant power in most situations.

I get low Wh/mi using CC and TACC selectively. I use CC and TACC on the open road with little traffic so I can drive a constant speed. If I am in hypermile mode, I always accelerate, decelerate, and maneuver in and around traffic with a gentle foot, using brakes and regen as little as possible.

For the open-road, steady-state, no-traffic situations, I have found that CC and TACC do better than I for maintaining constant speed and obtaining the best Wh/mi. This applies to flat ground and big mountains. The only place where I have had gains with manual driving in open road situations, are when driving steep, short, rolling hills that trigger regen on the downhill segments.

 

[KurtR]

You are not quite understand my argument. If you take a route where you slow down for up hills and speed up on down hills, I can find an average speed that covers the route in the same total time; travel that average speed continuously over the route, using more power on the up hills, less on the downhills; get to the destination in the same amount of time; and use less total energy for the trip.

No one could possibly drive using that technique. How, for example, would you set cruise control to say an idealized 54.32 mph to achieve the same "average time" that a driver would take on a trip of a specific distance if they feathered the pedal between 55 and 54 on a flat stretch of highway? A driver has two options - set cruise control on an integer speed value, or drive without cruise control. From experience, when I set cruise control at 55, for example, I see the orange power needle jump all over the place and it is usually jumping above the range I can keep the power within when I feather the pedal - more than it jumps below that range. I think this is because the cruise control is instantaneously responding to differences in rolling resistance as the road surface varies so as to maintain a precise speed, but perhaps there is another reason. I also seem to be getting lower energy usage by feathering the pedal manually, and even though one can theorize that there might some hypothetical cruise control speed that would save me energy and get me to my destination at precisely the same time I will ultimately arrive there at some point in the future, i don't have a time machine or an oracle to figure out when that time is, and set my cruise control at an appropriate value. I can only choose between cruise control or no cruise control, and the "no cruise control" seems to save energy because of the slightly lower average speed.

 

[3mp_kwh]

I did not say that TACC or CC was more efficient than manual driving. What I am asserting and attempted to explain via a simplified aerodynamic drag example, is that constant speed is more efficient than constant power in most situations.

Reading the thread, I think the missing statistic is Tesla's regeneration efficiency? Are KERs systems in passenger cars even 50% efficient?

If I had to hazard a guess, I'd go with constant power. Wind resistance rises exponentially, yes. So, holding 75MPH, rather than 65-85 may make sense, but swinging from 50-70mph, vs. 60, may also make more sense. It depends on whether you lose more to regen inefficiency, than you do by increasing speed.

This is clearly an opportunity for Tesla. When they get bored - after they get all of the Autopilot software working - they can give us a new firmware that has an Eco cruise option.

I don't know much about their implementation, but Ford told us they were doing something similar on the Energi twins...
Walter

Walter, I've followed what you say, on GM-Volt. Do you remember someone saying the Volt was only ~35% efficient, when regenerating energy back to the battery? Very old thread.

I was in a focus group, with an Energy PHEV owner who I think was saying his car used location-based engine firing. Many in the post-2012 Volt crowd are selective about manually using the engine, with 'Hold Mode'. The idea would be if Tesla is already using location-based tech, for air suspension etc, they can come up with a location-based mode that adapts cruise for greater efficiency. I'd use this. The algorithm would know the Tesla's regen efficiency, altitude, frontal area, Cd and speed . The user would simply set their tolerance for speed range. It is another way of confronting range anxiety.

 

[bhuwan]

I've driven the same route in similar conditions - my range is ALWAYS better when I am using cruise control (I do not have autopilot).

 

[jerry33]

I've driven the same route in similar conditions - my range is ALWAYS better when I am using cruise control (I do not have autopilot).

It does take a bit of practice to get it better without CC.

 

[AmpedRealtor]

Driving the freeways here in Arizona, I can get as little as 270 Wh/mi with my P85. Once I reach a cruising speed of about 70 MPH, I try to keep my power input around 20 kW. That causes my car to slow to 60-65 MPH on a mild incline and then speed up to 70-75 MPH on the other side, all without altering the 20 kW power level. Prior to this experiment, my energy consumption was closer to 295 Wh/mi when using CC.