Most efficient speed for best MPG?

As I already said above. I igore the 99. I go by my total trip mpg.

The 99 combined with the mpg when the engine is running is generally up near 80 mpg.
When my trip gets less than that it is due to the number of stop signs and lights on my route, or traffic.

Battery will alway get you more mpg total average than running the engine. That is why the car does it. It would even improve mpg at higher speeds too if it weren't for the rpm limits.

Wayne's chart is the steady state mpg with engine going.

 

Oooohhhh!!

OK, I get it now... It just read kind of odd to me...

 

The pulse and glide would be an excellent procedure to do if you were in a bind on gas and had to really milk it to get to a station.... just put your emerg flashers on!

 

Yea and no..... the conversion losses using the battery exist and are only worth it if the alternative is to waste the energy anyway like a conventional car would...

Seems the best use is to use of energy is to let the ICE keep the car above 30mph and take it easy as to not charge the battery and bring it up to 40 and then coast or dead band with no energy going anywhere....until you coast back down to 30 again.

So basically the ICE runs or not at all.
they actually try to not use the battery, because any energy used had to be created with gas!

Its better to let the engine take it easy as long as your at optimal rpms and get it up to 40mph, and then let it turn off as you coast back down to 30mph

At least this is how the guys did it that got 110 mph.


Now when in real use, you need the Ice to get up to serious mph or serious torque to get moving on the freeway, in which case the ICE will take assistance from the battery when needed and then put it back in.
This protocol they chose because it seemed to be the most efficient with all things considered since we need freeway speeds. If the speed limit was 40, then they could reprogram the computer to chance its protocol.

In other words... there are cool things you can do for special uses and trick the computer or sneak around its regular programming by use of an EV button or learning to feather the pedal in those deadband zones etc, but the computer is programmed for everyday use.

But when you have the chance, if you want to play and no traffic is around, you can pulse and glide to save gas big time!

 

according to stats i made, assuming being legal and not lagging. best street speed is about 35 mph (30 mph actually gets a bit worse nearly 50% of the time). on the freeway, 50 mph is probably better, but 60 mph seems to do the best because of the more constant speed. we do have 50 mph highways around here with minimal traffic control, but even one stop every several miles is enough to kill the high numbers.

besides, check out the hypermiler thread. i believe they stuck it at around 35 mph also.

 

The battery is your friend and with it you can get over 70 mpg every day if you wish to, where you can keep moving steady, and not drive other drivers crazy with pulse and glide speed variations.

With pulse and glide and no battery you may on some roads get maybe 90 mpg but I doubt you could get more. Don't forget the marathoners modified their car in ways normal drivers would not with the tire pressures and thin engine oil. Battery losses are not that major and are negligible compared to using the engine at low HP. This is very simple to confirm with the mpg readouts. The average mpg not the bar graphs!

Most people here seem to be trying to not use their battery and then complain their mpg is only 45 or something. This is why in a lot of cases. Only other reason is they could just be driving fast on highways, where max mpg is not possible anyways, or be in places like I am where you have to stop about two times per mile.

Whatever your conditions, you can do relatively way better than 80% of Prius drivers using max battery. If you are getting 70 mpg daily and want more and can put up with pulse and glide go ahead. I can't be bothered, and I don't see many reports from people doing it and getting 100 mpg!

 

I made a graph of Speed vs MPG using Wayne's Prius simulator and the best steady speed is in the 30's. The graph peaks at 32 mph. Avoid > 41 mph, mileage drops suddenly at 42 mpg due to spinning the ICE, and steadily above that due to drag.



This simulation is for 68F, sea level, 40% RH, 29.92", 225 payload, no AC, no wind. Click on the graph to download the excel spreadsheet.

 

<div class='quotetop'>QUOTE(kk6yb @ Feb 15 2006, 04:22 PM) [snapback]210092[/snapback]</div>

I disagree with this graph:

1) It does not reflect the lower power requirements at speeds below 41 mph.

As the vehicle rolls slower and slower, the aerodynamic componet goes to zero. This means the energy required to keep the vehicle rolling has gone down. Yet this graph shows not only a flattening but worse, an increase in power required. I do not know if the simulation is wrong but the graph is wrong.

A more accurate graph, for the NHW11 03 Prius is:




The simple act of benchmarking the various speeds has confirmed the above chart is accurate:
http://hiwaay.net/~bzwilson/prius/pri_test.html

Bob Wilson

 

<div class='quotetop'>QUOTE(bwilson4web @ Mar 11 2007, 01:07 PM) [snapback]403612[/snapback]</div>

Hi Bob,

I don't think we can perform 120 mpg at 5 mph.
We need 24 hours to drive 120 miles by 5 mph.
Prius requires about 300W to maintain itself, "basal metabolism". About 7.2 kWh for that 24 hours is a lot of energy compare to just against driving resistance.
I would like to support Wayne's simulator.

Ken@Japan

 

<div class='quotetop'>QUOTE(ken1784 @ Mar 11 2007, 12:49 AM) [snapback]403689[/snapback]</div>

We can repeat 15 mph / 24 kph in a single hour in an empty parking lot. I've already performed a 40 minute test and 'pegged' the MFD:

DISTANCE: 10.3 miles
MODIFICATION: thermistor hack (critical in cool weather!)
METHODOLOGY:
(1) fully warm up car and charge battery (~20 minutes above 42 mph)
(2) do one full lap in a flat, empty parking lot at 15 mph (*)
(3) reset MFD display
(4) hold 15 mph until all bars bars have updated, ~40 minutes
(5) take photo and have a happy

* - the goal is to normalize the battery SOC. It may make sense to modify the procedure to have "n", 5 minute display updates before resetting and taking a measurement. Regardless of how long #2 is, the key is to start the test with a normalized, battery SOC.

I would note that the simulator graph does not exceed 87 MPG. Also, the North American temperature problem that causes the ICE to run at coolant temperatures below 60C could easily explain the poor, low-speed performance in the simulator graph.

My energy based model does not reach 87 MPG until a speed of 30 mph / 48 kph. I did not include my measured NHW11 overhead ~511 W. because it is not a significant load until the lowest speeds. But between 0 and 30 mph, 15 mph is an excellent mid-point to determine what is going on. Even 20 mph should give an important data point.

Bob Wilson

 

Hi Bob,

Thank you for the data.
It looks the majority(for 25 minutes vs 40 minutes total) period is about 75 mpg which meets Wayne's chart.

However, your total average shows 99.9 mpg.
My math tells me you have to record more than 200 mpg on the rest of 15 minutes to achieve the 99.9 mpg average.
What happened on another 15 minutes?

Ken@Japan

 

Hi Ken,
<div class='quotetop'>QUOTE(ken1784 @ Mar 11 2007, 11:06 AM) [snapback]403791[/snapback]</div>

During this test, I didn't have the Graham scanner recording the data. But during the whole of the test, the MPG numbers were never less than 95 MPG at the end of the engine runs. I suspect but do not have the supporting data, yet, that the graphical display follows a different summary approach from the numerical value. At the end of this note, I'll describe some of my limited testing.

I am sorry for not posting both the spreadsheet, calculated_MPG_Rev_A.xls, as well as the graph:



In this revision, I added a second MPG graph based upon my measured 511 W. of fixed overhead power. Regardless, folks can pickup my spreadsheet and easily do 'what if' analysis by changing engine efficiency, fixed power overhead and easily make fully metric graphs. Certainly, analysis of my formulas is fair-game to see if I may have made a mistake. But what is most important, critical to any model, is getting benchmarks since it is testing and testing alone that lets us validate a model.

I have a couple of suggestions for folks who plan to add benchmark data. The NHW20 and NHW11 models have different rolling and coefficients of drag. Also, vehicle maintenance, tires and lubricants, have significant impacts. Finally, the North American NHW11 has an ICE coolant, lower limit of 60C and without a thermistor hack, should be used only for higher speed benchmarks where, sad to say, the differences between the two curves is moot. However, all other NHW11s and all NHW20s should provide valuable, low-speed data. In particular, one hour tests at a constant speed of 15-20 mph on a flat, empty parking lot would be very helpful.

As I mentioned earlier, I'm suspicious of the MFD display 'graphical' bars. I found that if I keep the speed below 6 mph, the graphical display, 5 minute interval appears to use some combination of the instant graphical display bar on the right for each 5 minute summary. This means you can use the battery to reach 6 mph to trigger a 100 MPG bar and back off. This fools the graphical display into displaying a solid, set of 100 MPG bars:

DISTANCE: 0.3 miles
MODIFICATION: thermistor hack (critical in cool weather!)
METHODOLOGY:
(1) fully warm up car and charge battery (~20 minutes above 42 mph)
(2) regen slow down to enter a flat, empty parking lot
(3) put car in "N" and coast to a stop with minimum electrical load and reset MFD
(4) after 'mileage bar' updates, put car in "D" and gently accelerate to 6 mph to get 99.9 'current' bar and then put in "N" and coast to a stop
(6) repeat step #4 until all bars show 99.9 MPG
(7) take photos and have a happy

My plan is to refine the Graham scanner analysis and see if that data can be used to cross-check the graphical and numerical values from the MFD. Then perform as series of tests to see if we can reverse-engineer the algorithms.

Bob Wilson

 

Hi all,

Find below a graph depicting what the latest Prius Energy Use Simulator outputs. Over the last 3 years I have gone out and performed several low speed tests to get better low speed data which was quite lacking before. A surprising things to note is that 20 MPH is the absolute best speed for high mileage in a Prius. This actually surprised me so, I tested for it several times. All tests were done without implementing pulse & glide techniques. I simply kept the speed on a level test track on non-windy days & let the HSD cycle through several full SOC patterns before taking the average. The graph below is based upon the following input variables:

Temp: 88 F -- About as hot as one can do without A/C & windows up for several hours.
Elevation: 400 feet above sea level
Humidity: 59%
Barometer: 30.03 in/hg
Wind: NONE
Auto A/C: OFF
Fuel: Summer Blend 114,500 BTU per Gallon US
Road Rolling Resistance: Smooth Dry Asphalt
Tire Rolling Resistance: 0.0077 -- OEM Integrity @ OEM pressures






Best Regards,
Wayne -- http://privatenrg.com

 

Thanks for your hard work Wayne with all these testing. Oh wow! 110 MPG at 20 MPH.

 

Hi Wayne,
<div class='quotetop'>QUOTE(mwbueno @ Mar 21 2007, 05:52 PM) [snapback]409821[/snapback]</div>

I've done one benchmark at 15 mph in my NHW11 that pegged the MFD mileage number at 99.9. I was wondering at what speeds under 30 mph you've run benchmarks?

The reason I ask is I was planning to run benchmarks at:
<blockquote>10 mph
15 mph
20 mph</blockquote>

Thanks,
Bob Wilson

 

I would like to raise a question along the same lines as was proposed here:

If you put a Prius up on rollers and made a chart of the performance for this car and perhaps using the cruise control to maintain each given speed above 25 mph, at what point would it peak? I am talking about the ICE and not trying to run off the battery or using pulse and glide. I think there must be charts or graphs of such things. It would be interesting to make such a chart at 1 mph increments from 25 to say 70mph. Anyone got the rollers you put one on? The instantaneous mpg will still change somewhat from moment to moment, I predict, but the changes could be recorded to find the mean for equal intervals of time, say 10 minutes. Keeping the engine at a constant temperature might be tricky, but a well-warmed up car could be tested by the lay-person, I would think.
Anyone?

The Lone Ranger

 

<div class='quotetop'>QUOTE(mwbueno @ Mar 22 2007, 07:52 AM) [snapback]409821[/snapback]</div>

Hi Wayne,

Nice to see the updated graph.

Thank you,
Ken@Japan

 

<div class='quotetop'>QUOTE(The Lone Ranger @ Mar 22 2007, 12:04 AM) [snapback]409990[/snapback]</div>

Let me offer some data and an explanation:

http://hiwaay.net/~bzwilson/prius/pri_test.html

The critical data is the cycling between ICE-ON and ICE-OFF modes:


The car constantly cycles from ICE-ON to move the car and recharge the battery, and ICE-OFF when the battery alone provide motive power. During the ICE-ON time, it runs between 10-20 hp, a very efficient power range. As the speed increases, the duration of the ICE-ON times increase and the durations of the ICE-OFF times decrease. But the ICE power range remains between 10-20 hp. There is one exception, at 15 mph, it stays between 10-13 hp. But the power cycling remains the same.

One other aspect is the Power Split Device tansaxle also keeps the ICE torque gently sloping from 1,650 to 3,050 rpm and flat thereafter. This is important because it gives exceptional specific fuel consumption:




This is magical-mystery technology compared to any ordinary car:





Questions?

Bob Wilson