Technicality of hypermiling

Ooh, I would be interested in knowing the answer to this as well. Right now I am getting 50 MPG, but that's the best I've done so far. (Got my 2009 Prius in Nov).

 

There is no simple answer to your questions. There are several techniques used by manny. The best thing to do is dive in to this site and see what works for the people putting up the best numbers.

Certainly search for threads on pulse and glide, and super highway mode as a start.

 

1)Pulse and Glide below 42 mph. Anticipate the road so that brake use is minimized. The most common examples are to coast to red lights, or before turns.
2)Technique does not matter much if at all. Slower is higher mpg
3)Moderate acceleration (gas pedal about 1/2 way depressed)

For the committed hypermiler (in increasing level of intensity):
Install and use an engine block heater in the winter
Avoid cabin heat use in the winter until the engine is warmed up
If you live in a dry climate, use outside air to keep the windows clear rather than the defroster.
Install a scanguage (to monitor engine temp) and block off part of the radiator vents in the winter.

The committed hypermiler can close a lot of the winter-summer gap in mpg, which is around 20%+ in snow country. E.g, I run 65 - 75 mpg in the summer, but fight to see 55 mpg in the winter without radiator blocking or engine block heater use.

 

who's manny?

 

I bet manny is many. Wee alll makee typingg mistakess.

 

Here is one of the more complete writeups on driving the Prius for maximum fuel economy. Study this and, as I think I suggested to you in another thread, practice and have patience. I'm still fine-tuning now after 3 years with the car.

 

yeah, just a bad joke (re: manny).

anyway, I'm no hypermiler, but I do my best to get good mileage despite bad circumstances. I'm driving back to my office and gliding to a light that I nearly always miss, and lo and behold, I've timed my glide perfectly, as the road turns slightly at the light and my glide is going to get me at a perfect turning speed as I see the light turn green.

but, alas, there is a problem, a car is waiting at the light... unfortunately, this car fails to see the light change, and stays stopped until I am right on their behind, causing me to come to a complete stop before the car finally moves. The worst part is there is a second light in this sequence, that again I usually miss, and it's green, but the few seconds we blew by not noticing the light turn causes us both to miss that light just barely.

They are turning left at that light and I am continuing straight, so I see that it's a young lady who is fooling with her cell phone. ugh.

and the worst part is, she was driving a prius.

sigh. tales of a frustrated hypermiler.

 

My take is that from a MPG point of view, gliding is the most efficient because you are not consuming any energy, whereas in battery operation you are consuming some of the energy converted from using gas and regenerative braking.
From an energy point of view, battery operation is the most efficient because you are using energy generated from the ICE, coasting, and regenerative braking. This energy moves you further down the road than gliding does - hence it is more efficient than the energy expelled burning gasoline.
Now, introduce one more energy source that few are using at this time.. Plug-in electricity and one can now establish a rank order of the different energy sources as they push you down the road:
In order from least efficient to most efficient: gasoline, Plug-in, regenerative braking, coasting, and <gliding>.

 

This is a common source of confusion. I suggest changing your perspective, from trying to manage the battery, to trying to manage the engine (ICE). After all, all car movement is from the engine and petrol use at origin -- all you can do is use less petrol/distance.

I like to view it in terms of engine power and efficiency: engine power output of less than 15 kW is inefficient -- don't do it. Either press down harder on the go pedal, or switch to battery power. Or switch back and forth between periods of more pedal, and periods of soft pedal (EV) -- the pulse and glide routine.

How much is 15 kW ? A rule of thumb is less than 1/4 depression of the full distance the pedal can move. It is not exact because the power contribution from the battery will vary with state of charge, but you do not have to be exact.

So, is it a good idea to use only battery from stop ? If you can use the ICE efficiently you should: you get up to speed quicker (less aggravation to people around you, nicer driving for you), and you save the battery for times when ICE use just cannot be efficient.

The other thing to talk about is why P&G is below 42 mph: at this speed and below, the ICE does not spin at all during coasting or gliding, thus saving that work. Notice we are not talking about fuel burn in the ICE, just whether it has to spin.

 

No doubt you can improve. My last tank was 68 MPG.

The others are correct in saying to avoid battery propulsion. Deliberately using and then replenishing a lot of battery power introduces efficiency losses from energy conversion: ICE energy converted to electrical energy converted to stored chemical energy, and then back again to use it. I deliberately use EV mode, as it's called, only in low power-demand situations when I think it might be inefficient to call upon the ICE -- low-speed drives around a parking lot or my driveway, for example. I also might use it to prolong a glide a bit as I approach a traffic slowdown, where a brief burst of ICE power might be less efficient.

As you suggest, part of the battery's value is to recapture energy otherwise lost and put it to use. Another major advantage is in allowing a smaller and more efficient ICE to be used. The ICE is 1.5L, the same as in the much lighter Toyota Echo. It also uses a more efficient Atkinson cycle (don't ask me to explain that ... do a good search for a technical explanation that I couldn't begin to give you), rather than the more common Otto cycle. The size and cycle make for a very efficient yet underpowered ICE. If your driving conditions required nothing but steady-speed cruising on level terrain, the ICE would be more than adequate, and your fuel economy still would be better than virtually any car on the road. But those conditions aren't real world; we need to get up to that steady speed and most of us drive hills. The electric motor provides an immediate and substantial boost for high-demand situations like fast acceleration on a highway ramp or climbing a steep hill.

I like SageBrush's "ICE management" concept, though I wouldn't exclude the importance of battery management. I manage both. Since I monitor battery state of charge (SOC) with added instrumentation, I can see exactly what SOC is at any given time, and I know whether I have the luxury of being able to use some of it.

Meanwhile, the subject of how much go-pedal to give during ICE-on conditions is a matter of substantial discussion. First, I don't care much for subjective terms like "brisk" or "moderate" acceleration. What is moderate to you may be either brisk or gentle to others. So, some time ago I proposed an objective measure that anyone can apply without added instrumentation: During ICE-on conditions, keep the instantaneous MPG (iMPG) at least half the vehicle speed. See this for a detailed discussion.

As SageBrush suggests, low-RPM conditions are thought to be inefficient. The theoretical power curves of the Prius seem to confirm this. But anecdotally, many hypermilers (me included) report good results with keeping ICE RPM low, sometimes as low as 1200 or 1300. How does one know RPM, you might wonder, without a tachometer? First, that would be gentle acceleration by anyone's definition, despite the subjectivity of the word "gentle." Second, it would make the iMPG well above the vehicle speed, perhaps up to 1.5 times. It is critical, though, not to relax the pedal so much that you're driving on battery power alone. As the pedal is adjusted, there is a fine line between EV mode and low RPM ICE power.

I have done some preliminary testing comparing various acceleration rates during pulse and glide, and the results suggest rates in the range of 1300 are at least as efficient as those in higher ranges. The key word is preliminary, however; the test environment was less than ideal (traffic disrupted my glides on the two-lane rural road I was using), and I'd like to repeat the tests in a different environment.

 

JimboK: sure, go ahead and boast. I'll see you in July

Courtesy of wherever Hobbit got them from, a power vs efficiency curve:


ICE Efficiency drops off after ~ 45 kW, equal to about 4000 rpm or about 75% pedal depression.
So there you have it in a nutshell: feather the go-pedal to coast or glide, but if the ICE is on try to keep the pedal in the 25 - 75% range of depression. JimboK's approach is probably good too, since I think Hobbit blessed it some time ago. I find it too much work, and worry about taking my eyes off the road. My personal approach is based on knowing from experience how much engine noise corresponds to power output. But that is not something I can teach, since each person's ears are different. Nowadays I find my use of the MFD is to avoid the situation where I cannot hear the engine, but it is on -- I was trying to feather the go-pedal to coast, but am pressing too hard. In that case I release the gas pedal completely for a couple seconds, and then gently reapply.

Anyway, the graph should convince you that a wide range of power output is efficient. Avoid the extremes of just a liitle bit of ICE, or flooring the gas pedal, and the wonderful HSD will take care of the rest. Spend your unused attention on the road staying safe and anticipating slow downs to avoid brake use. That is all that is required for excellent fuel economy. In the summer anyway; cold winters require steps to preheat and keep the ICE warm -- a discussion for another day.

 

Let's agree on some definitions:

Coasting: no arrows on the energy screen. ICE is off.
Gliding: ICE off, arrow from the battery to the wheels

Both are equally efficient, in the sense that efficiency is only relevant when ICE use is being considered and here ICE is off in both cases. Picking one or the other is just a matter of what speed you want to travel at. Usually, if you have bled off speed and want to gain it back, use the ICE in the 15 - 45 kW power range to get back. An exception would be if the battery is fully green This is a situation I see fairly often, but most drivers do not. It is terrain dependent. The other practical consideration is that in general it is easier to coast than glide without inadertently slipping into low ICE power use.

If by support/assist you mean keeping the ICE out of the 1 - 14 and > 45 kW (approximate) power output ranges, then you understand exactly. Parenthetically, the so called 'mild-hybrid' designs solve the high power output inefficiency (and ability to reach it), but don't do nearly as well in the low ICE power problem because the ICE is almost always idling or a bit more when the car is moving. Both Honda and now Toyota have come up with innovative ways to keep pumping losses at low power ICE use down (and thereby increase efficiency), but they will never be as good as simply not using the ICE is those situations.

Addendum: I'm going to add the following mostly to avoid having a guru point out an inaccuracy in what I said above. Ignore it if you find it confusing: The HSD does not use programming to keep the ICE off between 1 - 14 kW. We the drivers are 'gaming' the system to avoid that inefficient power band, and luckily for us the Prius system design makes it easy to do. What actually happens in HSD is that the first 'x' kW of power (from 0 - 20 kW) requested by the driver is the battery, depending on battery state of charge. So long as we request power below that amount, the ICE stays off. This is why it is easier to coast/glide with a fuller battery; it's also why in general you do not want to use battery power if the ICE can be used efficiently. You might be wondering if the HSD system isn't actively managing 1 - 14 kW power demand conditions where the benefit for non hypermilers comes from. It is two-fold: smaller engines have less pumping losses, and the ICE turns off for everybody in a Prius who take their foot off the go-pedal completely.

 

Consider these three conditions of your foot on the go-pedal on level ground going 35 mph. Lets say that friction requires 5 kW to stay at 35 mph:

No pressure: ICE off, regen on and car slows down, since regen is this case is somewhere between 5 - 10 kW and friction are at play.
Tiny pressure (Coast): ICE off, regen off, car speed slowly goes down due to friction.
Small presssure(glide): ICE off, regen off, car speed may go up, stay the same, or decrease depending on how much power is coming from the battery.

Choose by the speed you want the car to go, not by any considerations of efficiency so long as the battery state of charge (SOC) is not low. If the SOC is low, use ICE in the efficient power band above 15 kW.

 

Actually, Hobbit recently has experimented using what some might call an "extreme" of a little bit of ICE, at low speeds anyway. It's in the neighborhood of the 1200-1300 range I referred to previously. He is using fuel injector timing as a surrogate indicator of engine load. The ICE really is inefficient when it's not running under load, and his view seems to be that as long as injector timing has reached a certain operational plateau, the ICE is loaded. [Hobbit, if you're reading, feel free to correct my paraphrasing as needed.]

Quoting him:

"Even a fairly gentle startoff from the line yields injector times of high sixes or 7.x at ridiculously low RPMs like 1250, as high torque is produced to help takeoff."
​

My own recent experience applying this and using injector timing as my load criterion suggests that in ICE-on conditions at low speeds it's actually difficult for the ICE not to be under load. Especially from a dead start, after "goosing" the go-pedal just enough for the ICE to light, I show injector timing levels well above the "loaded" threshold even with these low RPM.

In addition, many of the Prius hypermilers who travel the open road report excellent results with similarly low RPM using what has been dubbed Super Highway Mode (SHM.) Most are using ignition timing instead of injector timing as their load indicator. Hobbit addresses why he uses the latter in the previous link I provided, but either method gets results at -- to the point of this discussion -- very low RPM.

Well, re. those definitions, there is disagreement. Gliding, as most define it, is a no-arrows condition, triggered by completely backing off the go-pedal and then feathering it back down a tad, just enough for the arrows to disappear. Coasting as I define it is having the foot completely off the pedal, with the result being green arrows flowing to the battery.

True the ICE is off in either case, but coasting is regarded as less efficient when considering the overall efficiency of the car, not just the ICE. It allows as much kinetic energy as possible to be immediately used by the car, rather than being stored and suffering conversion losses in the process.

I like your description of how we "game" the system.

So Dominic, you are correct that a glide (again, as most of us define it) is a no-arrows condition.

I don't recommend battery propulsion from a dead stop unless, as I've suggested, it's a short, low-speed, low-demand situation like a parking lot. Again, I "goose" the pedal just enough to get the ICE to light, then almost immediately back off to just prior to the ICE cut-off threshold. (It takes practice to have a good sense of where that is.) How low of course is guided, as you suggest, by traffic conditions. Safety first, and from there I try to balance fuel economy and courtesy.

I really recommend you read the CleanMPG writeup I linked earlier. It gives a nice detailed description, with pics of the MFD.