The myth of pulse and glide

Correct, I ran out of time to finish.

I do agree that it is important to account for the SOC change. Over a series of pulses assuming that some charging occurs, the error will become considerably smaller. For a single if one determined the SOC change only for the glide and determined that the particular pulse had no charging component (no SOC change) then one could do an efficiency based conversion and not need to subtract from the pulse.

The problem in accounting for SOC change is in figuring the MPG equivalent. One needs both the kwh and an efficiency conversion to gasoline. This makes the calculation considerably more complicated than the OP was indicating. (And the kwh/mile varies with speed as well.)

100 mpg was far too low of a value to begin with, because it puts upper limits well below what can be acheived. And that is a big part of the problem, the mpg figure for the glide is a WAG, but a critical one because the glide distance tends to be several times longer than the pulse distance.

This is a difficult problem because as soon as one starts to add gallons to a the glide a whole Pandora's box of required assumptions, correlations, other data such as SOC, etc. is needed to do the accounting both for the glide and the pulse.

 

Assuming the car is maintained well ---

Minimize energy conversions;
If ICE is on, shoot for 37% efficiency;
Car speed below 42 mph (in the Prius II; other speeds for different hybrids) saves ICE spin power consumption during glides

That is it. There is no other magic. Minimizing energy conversions is the least important of the three, in most circumstances. Cutting out free ICE spinning is huge -- notice that NO ONE has come close to 100 mpg at 43 mph or above in real world hypermiling of the Prius II.

As an aside, if the discussion could move to kwh/distance Shawn would be able to cut through OP's confusion a bit quicker.

 

I have found it useful to start from a 'best possible' number and then work backwards. Petrol has about 36 kwh/gallon energy, and 36 - 37% engine efficiency is about the best we can hope for in current hybrid petrol engines. This gives us 13 kwh/gallon of motive energy to play with.

A heavy car like the Prius travelling 30 mph uses 100 wh/mile as a practical minimum, only including requisite frictions. Dividing one into the other, and we see that 130 mpg is a theoretical maximum. ICE freewheeling costs about 30 wh/minute. So it's easy to see that 100 mpg is only obtainable in a 'glideable' car, given current engine efficiencies. This calc is assuming no partial power losses, which is clearly not the case when a car is driven at low speeds and constant low engine power.

I realize that OP went for the catchy title to drum up interest in the thread and the enclosed advert for his blog. Neither is going to gain much favor around here, even before we read the gross misunderstandings. I'm glad he is interested in his new FFH car, and is willing to spend time and effort to promote it. But a bit of modesty is needed: his car's tech is HSD, which has been around 12+ years. A lot has been figured out, by some very bright people. He should get up to speed first.

 

This is an interesting chart, Bill, but it has a basic flaw. I don't think there is any problem with the basic calculations -- I've done the same and I see the same. The problem is that very few people are going to see glides below 100MPG. Most are not going to see glides below the 10,000MPG level. Show the chart, again, with the lines approaching their theoretical maximum expressions.

For example, using the 20MPG pulse example, and measuring, as your chart shows, the "Total MPG" for various distances (noted a multiples of the pulse distance), the theoretical end point Total MPG is higher than your chart show. I have measured such numbers for the distances you have noted (being 1x, 2x, 3x, 4x and 5x).

The end point Total MPG for a 1x multiplier would be 40MPG. This fuel exonomy would be reached so long as the glide MPGs were to reach 1,600 (or higher!). When the glide is two times the distance of the entire course, the end point Total MPG would be 60MPG reached at a point were the glide MPGs would have to have equalled or exceeded 5,000 RPM. At a multiplier of 3, the end FE would be 80MPG at a glide of 10,000MPG. A 4 multiplier brings 100 total MPGs for any glide at 16,000MPGs or better. A mulitplier of 5 takes us to a total MPG of 120MPG at glide MPG 24,000MPG or better!

 

Can P&G be equated to a household thermostat? You set the temp to 72 and the heat kicks on and takes the temp up to 74 and cuts out and when the house cools back down to 70 the heat kicks back on.
Since this is the basic premise in millions of homes, I assume that it is more efficient than keeping the furnace running continously.

OK, it is a gross oversimplification, and I am sure I there are significant differences, but I need to dumb this down for me.

 

It depends on the furnace, but yes, the idea is similar. The problem is having a device that is over-sized for the job. The furnace in your house is too big for ordinary operation, and the engine in your Prius is too large for normal cruising. They are required to big larger to handle special cases, such as really cold weather and climbing hills. Operating each at its most efficient power setting in and on/off manner is more efficient than try to throttle the output. Note that some devices actually throttle quite well, in which case proportional control is better.

Tom

 

The house furnace is doing a sort of P&G, not because it is more fuel efficient, but because the furnace has only two operating power levels: Full Power and Off. This allows for a cheaper physical plant. The on/off duty cycle produces the correct average heating power, but the temperature oscillates.

A typical automobile ICE can operate at any desired power level between Full and Off, but efficiency various drastically with power level and RPM. For a traditional prehybrid ICE, highway cruise is not anywhere near the best efficiency. So P&G alternates between a high efficiency area, which is higher power than needed thereby causing significant acceleration, and Off (or idle), which causes a slow deceleration. Non-hybrids have seen some huge gains using P&G.

A typical hybrid ICE should be operating closer to peak efficiency during highway cruise, leaving less to be gained from P&G. But depending on the details of the specific hybrid power plant, it may be falling out of peak efficiency at lower speeds, leaving more benefit to be harvested. I haven't explored my new Prius very well yet, but at city street speeds my ScanGauge is showing signs of less than ideal ICE efficiency. And my first attempts at P&G at those speeds are encouraging.

 

Perhaps the Author's retraction will be titled:

"The myth of pulse & glide myth"

The article kind of reminded me of how scientists once struggled with bumble bee flight ... considering it, 'unexplainable', simply because of the models they had, and the mathematical assumptions that they made. The assumptions were there's too much mass, and wings are too small. The observers simply didn't have enough data ... here, and with P&G. High speed photography of wing travel eventually enlightened 'em. Similarly, when the fuel tank takes you from 500 miles without P&G, to 700 miles with ... you sorta have get a new set of assumptions.

 

So, to sum up: (Stop me when I am wrong)
1. fuel consumption evidence shows that P&G works, depsite graphs and theories to the contrary.
2. P&G does not buy much at highway speeds since the ICE is more efficient at those speeds.
3. P&G is more effective on flat roads or downhill grades than going uphill.
4. Neutral allows for longer glides, but does not regen the battery.
5. P&G can annoy other drivers behind you.
6. the trick on the glide is to hold the accelerator down just enough to see no power, no regen on the HSI.
7. you don't need a hybrid to P&G, but hybrid results are more impressive.

Am I making any mistakes or omissions?

 

The theory of P&G is well grounded, and well understood. By most, anyway. What is so hard to grasp: it is better to use the ICE in it's most efficient manner *all the time*.

 

Has anyone had a passenger take a video of P&G in action? There seems to be some points of finesse that might not translate into text. How hard to accelerate? What speed range is optimal for P&G? etc.

An instructional video would be helpful. Or maybe they exist and I don't know about them.

Just wondering.
Thanks

 

Argyle, I can give you 97% of the truth. The other 3% are exceptions, or very little return on mpg for the hassle ...

Don't bother on the highway, or on inclines
Accelerate more than a feather, less than pedal to the floor. "Moderate"
In the city try to glide at speeds less than 42 mph on the G2 Prius, 47 in the G3 Prius. During hte glide the engine will not spin, saving about 2 kW (a lot).
Time the glide duration by the traffic, red lights, and stop signs.

Be flexible in the speed you start to glide. Remember that other than 'pulsing' the ICE at an efficient power band, really good MPG comes from anticipating traffic and minimizing brake use. Hybrids and P&G turn a smart driver of a conventional car into a 65+ mpg city driver (Boston, NYc etc respectfully excluded).

 

Following is a copy of [URL="http://priuschat.com/forums/gen-iii-2010-prius-main-forum/64638-glide-question.html#post893892]my post[/URL].
======
In Japan, we have been discussing how to pulse&glide on the 3G Prius.

Our conclusion is we don't have to be so sensitive for the zero amp gliding and a weak battery use is acceptable, therefore 0-25% zone on the HSI screen is acceptable.
Also, a slow acceleration is not good for the ICE efficiency, therefore 75-100% zone is recommended for pulsing.

Ken@Japan

 

ok... completely uninformed as to options in the home heating world, but would it not make sense to use a variable speed fan on with variable heating to

1) keep a consistent heat level
2) use less energy

the full on, full off system has to reheat the ducts every time it comes on if it does not stay on which i would think would be wasting energy....

 

Yes, but it requires a more expensive furnace and thermostat. Some of the new systems work this way, or at least have a small number of out output levels.

Tom

 

I really don't understand why you persist so on this issue.

You've discovered a neat mathematical statement showing that for a pulse of x distance at y MPG that the maximum MPG is 2y for a combined pulse+glide of distance = 2x.

You've also discovered the reality of diminishing returns and lowered marginal return when approaching maximum values.

Neither discovery is something that can "debunk" the "myth" of P&G.

The reality is that whether in full pulse and glide or simply pulse and lowered acceleration, the driver who utilizes P&G techniques is not running glides at anything approaching distances only equally the length of the pulse. Further, the initial energy to raise the car from a standing operation is not dissipated to another zero state in the typical P&G.

Moreover, those of us who utilize some of these techniques have discovered we've been able to increase fuel economy. We wouldn't do it if it didn't have value.

 

It might be curiosity. Often discussing a subject can lead to insights but any such discussions need to be matched by experiment.

It is part of understanding what is going on. Now the proposed model needs experimental confirmation. It also helps to compare the model to other published reports such as SAE 2009-01-1322 "Vehicle Inertia Impact on Fuel Consumption of Conventional and Hybrid Electric Using Acceleration and Coast Driving Strategy."

It is part of forming a model that can then be tested in the field. However, I agree that 'myth' was a poor choice of words picked up from the article.

The problem has always been the lack of a math model or experimental data showing "how much" is gained by "what values." It is also complicated by mixing terrain driving with pulse and glide. Sad to say, the SAE paper used a dynamometer instead of field tests and their reported values are too good compared to my tests.

Now I agree that this forum seldom encourages the type of rigorous study, an empirical approach to the problem. This leads to calls to 'stop it' but I have no problem with asking questions and presenting a testable hypothesis. The next step is to "go to the field" and see if the measured values match the theory. I remember someone once posted their results comparing P&G versus the equivalent constant speed. Regardless, I don't think we have the last word on P&G and there are better models.

Recently, Hobbit posted something I've been thinking more about, operating the engine at peak BSFC. Measuring BSFC is not trivial:

• measure ICE torque - use MG1 current
• measure injector timing - provide fuel flow and ICE rpm

We already know how to measure the vehicle drag, rolling and aerodynamic. So the key is to operate the engine in peak BSFC region, an area some of us have already been looking at:

and:

and:

Also, SAE papers such as 2009-01-1048, 2009-01-1061 and others are giving a clue, tantalizing tips but raising more questions.

My point is if someone posts something you disagree with, post a better model, correct the errors or if all else fails, take it to the field and do the experiment. It isn't that hard and a heck of a lot more fun than questioning motives. After all, no one is expecting a Spanish Inquisition and it is perfectly OK to 'post through' our differences. We all may gain insights and understanding.

Bob Wilson

 

Bob, you are one of the hardest working folks on this forum. But I keep thinking that a shrink could show your graphs to his patients and find out that they have deep seated resentment toward their mothers and commitment issues.

Thanks for keeping us educated.