Optimum Cruise Speed with Cruise Control engaged

The Prius MPG Simulator predicts the best fuel economy at a steady speed to be at about 20 MPH. From there it decreases more or less linearly with increases in speed. That's below the threshold at which CC will engage, so the best CC-maintained speed for fuel economy will be the lowest speed at which it engages: 25 MPH or so, IIRC.

EDIT: Here is a chart with Simulator data from the Simulator developer himself, along with other variables used and information on his development methodology in the linked post and subsequent ones.

 

On cruise control, about 24 mph. Although cruise control can hold 23 mph, if you hit any sort of upgrade, it will drop. However, 24 mph gives barely enough margin to stay in cruise control.

I've done my work with a 2001-03 Prius and this is what I've found:


What we find is any speed below 42 mph, the ICE will cycle on-off at an efficient power mode, nearly the same efficiency, all day long. So in this speed range, the trick is to approach 18-20 mph, that seems to best match the load to efficient engine power production.

Above 42 mph, 65 mph seems to provide an excellent 'knee' for optimium highway performance and minimum time. Going faster brings the mileage down without a significant time savings. Faster than 75 mph is really rough on mileage.

Now 42 mph is a special case and my data suggests that staying either above or below is the way to go. I use a 4 mph guard band so my fastest low speed is 38 mph and 46 is my slowest high speed.

Your right about running the engine at peak efficiency ranges but the CVT makes that dang near impossible in normal driving. We don't really have fixed ratio gears. The electricity conversion is a little bit of a red-herring.

Bob Wilson

 

1700 - 2300 RPM at highway speeds. RPM can go a little lower at
about 35 mph as long as you stay torque-loaded in a pulse. That's
hard to tell without instruments, but as some other folks have
noted it results in an MPG reading between 1.0 and 0.5 times
the MPH reading, roughly. If traffic allows the leeway to
run your MPG just under your speed, yielding fairly slow
acceleration, it's a very efficient state. It's all weird in
a prius because various control curves are nonlinear. The
cruise control won't have any concept of optimizing that,
however.
.
_H*

 

It's optimum engine speed, and we are talking about the optimum vehicle speed.
Or, am I missing something?

Ken@Japan

 

Within the optimum *engine* speed and loading conditions, vehicle
speed is going to vary all over the place depending on terrain.
Simply because of the way the Prius works I don't think you can
base any of this thought on vehicle speed, or to try and do so
gives a set of very limited and distorted results with
relatively little practical value.
.
_H*

 

The nomograph that comes bundled with the aforementioned Simulator suggests that MG2 will be at 4549 RPM at 75 MPH. However, an ICE RPM of 3500 will provide rather brisk acceleration, not a steady state. From experience, I'd say a steady speed of 75 MPH on level terrain will put you in the neighborhood of 2000 RPM or so. Assuming that, MG1 RPM is -4627 RPM.

At 75 MPH, the Simulator predicts 41 MPG, assuming 50F ambient temp, level terrain, no wind, smooth dry pavement, load of 300# (driver & cargo), and Goodyear Integritys inflated to 44/42.

 

My data suggests the "fixed" load is about 511 W. used for daylight running lights and control computers. The graph without the 'fixed' load is based upon just the drag formula. Both assume an engine running at 31% efficiency but that varies based upon fuel quality, engine rpm, temperature and loading.

Bob Wilson

 

With my 2003, NHW11 Prius, it will be ~39 MPG on a standard day. I've got two measurements of this value. My current thinking is at 75 mph, my NHW11 loses the ability to 'idle' the ICE but has to run faster than idle to avoid letting MG1 spin faster than 6,500 rpm (the NHW20 has a 10,000 rpm limit.) The data point offered by an NHW20 driver indicates they get better mileage:

I like to have a good model but data is the gold standard. The only problem with data is variability when the test conditions change.

Bob Wilson

 

I have an idea to run a more or less level course several times alternating Pulse and Glide, followed by another run with the CC set at the average speed of the P&G run, and to record the data to determine which is most fuel efficient.

Would this be worth doing? Has anybody done a real-life test like this?

I could write it up and present at HybridFest.

 

Please do!

Here are my results:
http://hiwaay.net/~bzwilson/prius/pri_PnG_010.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_020.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_030.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_040.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_050.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_060.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_100.jpg
http://hiwaay.net/~bzwilson/prius/pri_PnG_110.jpg

This data suggest that 25-43 PnG, 33 mph average, the savings is ~11%. However, at 15-25 PnG, 19 mph average, the loss is ~14%. Unfortunately I haven't seen a good P&G model published.

It appears the built-in hybrid cycle of our Prius is more efficient at the maximum distance speed, 18-20 mph, than manual pulse and glide, 15-25 PnG. But at higher speeds with a significant dV, there appears to be some improvement.

What is needed is a set of speed range tests like:

5-43 PnG (? average)
15-43 PnG (? average)
25-43 PnG (33 mph average)
30-43 PnG (? average)
35-43 PnG (? average)

From this data we might be able to work up a Prius PnG model that has some credibility. But remember, we need both the PnG and the equivalent constant speed MPG and enough measurements to ensure they are reproducible.

I suspect that there is a minimum dV that has to be achieved in PnG before there is any savings over letting the normal Prius control laws work. Note that PnG must be on level ground.

There is a different kind of efficient driving call 'terrain driving' that exploits altitude changes. Take care not to confuse or mix the two. I would recommend multiple runs in both directions as well as checking a topographic chart and/or using GPS to read out the altitude. Google Earth is a good reference since it interpolates altitude at any point.

GOOD LUCK!

Bob Wilson

 

Bob and I have had some dialog about this; he obviously has made a lot more headway than I. I started the testing process and ran into some practical problems like having to repeatedly abort test runs (measurement mistakes on my part and traffic, as two examples of causes). Then I unexpectedly had to replace my tires after I awoke to an unrepairable road hazard one morning, rendering any subsequent comparisons to previously-recorded data invalid. I thought about publishing those data anyway, but they're really too incomplete to draw any conclusions from. For a variety of reasons I have had to postpone resumption of testing, but I hope to begin anew when the weather warms.

Here's an earlier thread with more details.

 

This sounds like a great idea!

Drop me a PM so we can talk off-line about when to present this at Hybridfest. Perhaps at one of the member-only sessions on Sunday morning at the hotel?

 

Can someone just give us the bottom line please when it comes to the best speed on a level highway?

All this RPM stuff, while interesting, might be confusing to our less technically inclined friends.

 

Bottom line: as slow as you can safely go. As I mentioned previously in this thread, fuel economy drops pretty much linearly with increases in speed.

Cruise control is probably your best option until you start hitting significant hills. I did a controlled test comparing CC to manual control of RPM on a highway with gently rolling hills, and the results were nearly identical.

 

Let me comment...
Above 42mph or on highway, as slow as you can.
Below 42mph, 20mph will be the best.

Ken@japan

 

He was asking about the highway, so that was what I responded to.

 

I agree.

Ken@Japan