MPG variance between actual and computer?

i noticed the same thing and just threw the difference in there. according to my spreadsheet entering in number from Trip A in tenths, my total mileage was about 25,400 but my actual odometer reading was 25,480 or so. accounting for the mileage it had when i got it only explains 14 miles of it.

no explanation since that is about 25 fillups so the variance is more than 2 miles per tankfull.

 

I did my stats for the past 14 months.
MFD 51.4 mpg
Actual 50.1 mgp
Difference of 1.3 mpg

I have about 50 miles unaccounted for also, I thought it was my error. Interesting.

 

I stopped in the dealership last night and talked with the Service Manager.
He hadn't heard of the problem. But wanted to see my spreadsheet. So I will do that and see what he has to say.

 

We have a 2007 which we bought new, and have driven almost 32,000 miles in the 14 months we've had it. We kept meticulous track of fuel we put into it the first year, and averaged exactly (plus or minus odometer error and error in gas stations' fuel pumps) 52.0 mpg, which we consider excellent, attesting to the fact that the kind of driving we do (in the Florida Keys) is real good for getting optimum fuel economy.

I'm a techie/science guy, but I can't say I know everything about the Prius' computer. I THINK the main reason for discrepancy between actual fuel economy and what the computer shows is that (I'm pretty sure of this) the computer shows a time-averaged MPG number, rather than actually dividing miles by gallons. I seriously doubt that the Prius' computer knows how many gallons get put into it, as the fuel gauge is very non-linear (about the only gripe I have so far). If the computer were metering actual fuel usage, there'd be no excuse for such a non-exact fuel gauge.

By time-averaged MPG I mean this: I'm pretty sure the cumulative MPG the display shows (which if you don't hit "reset", keeps a longer-running average) a running average of the total of the 5-minute averages that it constantly displays. Think about it -- you could go, say, 10mph for 30 minutes, and rack up six 5-minute averages that are very high (99.9 MPG, which really is "infinity", but don't try figuring "infinity" into an average -- it'll blow your mind), but only go 5 miles. Then go 60mph for 30 minutes (more like 45 MPG). The display would average those six 99.9's plus six 45's, and display 72.5 MPG. But you would have travelled 35 miles and used 0.66 gallons, which is actually only 53 MPG. In actual practice, the gasoline engine would run part of the time you're going 10mph, so the display would show a little less than 72.5 MPG, but this example shows why time-averaging miles/gallon doesn't yield true MPG. Going 10mph for very long is not realistic, and our display (if we let it keep a longterm average by not hitting "reset" except rarely) typically shows a couple MPG higher than I know we actually get, and I'm pretty sure the above is the reason why. The computer on other model years may work differently.

 

Perhaps you're right, but it seems to be generally accepted by the knowledgeable members here at PriusChat that the MFD is actually dividing miles by gallons.

The lack of a precise specific indication of how many milliliters of gasoline are remaining in the tank is a common complaint here at PriusChat. I find that the gauge is sufficient for it's intended purpose (making sure I don't run out of gasoline or needlessly waste time filling when the tank is still relatively full), but I think most would agree that the Prius doesn't keep track of how much fuel has been added.

Usage tels you nothing about how much has been added. As has been explained in other discussions on this topic, the Prius measures fuel usage by keeping track of the pulse length of the feul injectors. The fuel injectors have no way of knowing how much fuel you added the last time you filled up, but they provide a pretty reliable method of determining how much has been used.

I'm pretty sure this is not what happens. I have personally had situations where after resetting the MFD, I've driven for 15 minutes with the first two 5 minute bars greater than 90 MPG and the third 5 minute bar at 48 MPG. The 90 MPG were P&G between 30 and 40 MPH (figure an average of 35 MPH) for a bit less than 6 miles (figure 5.8), and the 48 MPG was on the expressway accelerating to and maintaining 65 MPH for a bit more than 5 miles (figure 5.4).

Using your math the MFD should have displayed 76 MPG, but it actually displayed 64 MPG.

Keep in mind that 99.9 isn't necessarily "infinity". 99.9 is what the MFD shows anytime the actual value is greater than 99.85

Which kind of tosses a monkey wrench into your "average of the five-minute bars" theory. How would the average be computed when you drive for 5 minutes without using any fuel? It's easy if dividing total miles driven by total fuel used, but if computing an average of the five-minute bars, you end up with a mess.

 

Some pretty knowledeable Pri-ites here, so just a quick question from a 4-monther:

Why doesn't the odometer even display tenths of a mile? Is there a setting that I haven't yet located, for measuring in tenths or even 1/100's ?

In doing our various calculations, even just one decimal pt. in the odo-readings would allow some additional precision.

 

Danny, thanks for helping me understand the means (length of the injectors' pulse length) by which the 'puter quantifies/estimates fuel consumed, which I wondered about. By that measure, I then agree that the MFD does divide miles by that estimate of gallonage. (I didn't think it was as crude as the old method Studebaker used, intake manifold vacuum, which was kinda nifty - I'm showing my age there.)

And you're right that the fuel gauge is good enough to keep you from running out... and of course the first segment that disappears from the gauge after a fill-up depends on how much more you squeeze in after the gas pump's auto shut-off. It is a minor gripe, but even the other segments disappear very non-linearly.

I am pretty sure the running average is time-averaged, but I'll do some more observing, and play with resetting the MFD at intervals to try and confirm. I do reserve the right to be wrong, and I don't mean to sound smarter than I am.

And yup, I realize 99.9 could be less than infinity, but the only time I see 99.9 as the "current" (i.e. instantaneous) MPG is when I know it's running on electric-only, which is infinity MPG. As a 5-minute average, you're right, it could be anywhere between 99.85 and infinity.

Thanks for helping me understand a bit more, and sorry I haven't yet figured out how to use the "quote" feature.

We love our Prius, and I thank you for this great website!

 

Because the law doesn't require it, and they decided to design it that way. They apparently didn't feel that tenths of a mile was important on the odometer. I'm not trying to be sarcastic here. There just isn't a better explanation for "why".

Nothing for hundredths, but I thought that the trip odometers displayed tenths. It's been awhile since I last looked at either of the trip odometers, so I could be mistaken on that.

 

Perhaps there is an element of time-averaging involved in the display of MPG on the MFD, but it isn't relying on or using the 5-minute intervals. Here are my two "reasons" why:

1. Using your example of 30 minutes at 10mph and then 30 minutes at 60mph, I offer the alternative of having the car sit still for 10 minutes right inbetween these two 30 minute cycles. Try this with two variations for the 10-minute period. For one 10 minute "stopped" period, the Prius is in "Ready" mode with no significant draw on the batteries (12V included) so that the ICE never turns on. In the second variation, have the A/C on with tunes blasting away, all while the lights are on (anything to ensure the ICE is running the whole time). In each variation the MFD will show 0 mpg for those 10 minutes. In reality, the gas consumed will be significantly different between these two variants, because we know the ICE is burning fuel. It won't use a gallon of gasoline, but then you haven't yet used a gallon of gas in your one hour experiment anyway! The MFD would give you different read-outs in these situations, because the MFD is actually measuring gasoline used by the ICE.
2. My second reason is simpler. The MPG on the MFD updates every ten seconds. It is not relying on 5-minute intervals. (If it is relying on 10 second intervals, then we need to talk some more, but I believe the MPG is calculated using the odometer in comparison to the fuel usage.)

You will see the 99.9 even when the ICE is running. You don't see this as frequently as when the ICE is off, but it does still happen. I see numbers in the 80s and 90s quite often, so I believe the screen would be showing 100s and 110s and 120s if only it had room for another significant digit. I believe people have reported from other instrumentation that they have measured instantaneous MPGs at better than 200 in certain circumstances.

What haven't you figured out about the "quote" feature? Some things require the secret knowledge, but others are available to the rest of us. Is there a specific question you have about "quote" that we could try to answer?

BTW: Welcome to PriusChat!

 

Thanks, a priori! What I haven't figured out quote-wise is how you quote just certain selections from a longer post to reply to (no great big deal). Hitting the "quote" button appears to duplicate the entire post I'm replying to. I tried highlighting selections, but that doesn't seem to do it. I have some vision problems and can look right at stuff and not see it - LOL.

And thanks for the further discussion on the MFD. It's more complex than I thought, and I'm glad I found some folks who know. Nobody at my dealer seems to know any cool stuff about this most amazing car. (I stand corrected, said Tom in his orthopedic shoes - a Tom Swifty, if you know what that is.)

 

Ah hah! I can fix that!

. . . All you need to do is wrap the phrase in the beginning and ending codes. The ending code (or whatever it is really called) is always "[/quote]", while the beginning code identifies the specific posting.

Of course there are other ways of doing this, but to "quote" this phrase:

Hitting the "quote" button appears to duplicate the entire post I'm replying to. I tried highlighting selections, but that doesn't seem to do it. I have some vision problems and can look right at stuff and not see it - LOL.

. . . just encase it in the quote= . . . and /quote phrasing like this (I've changed out "quote" for "quite" so you could see behind the curtain):

[quite=Fuzzy;618330] Hitting the "quote" button appears to duplicate the entire post I'm replying to. I tried highlighting selections, but that doesn't seem to do it. I have some vision problems and can look right at stuff and not see it - LOL.[/quite]â€

When I replace quite with quote, I’ve included only the desired portion of your post in my reply:

Hope this helps!

 

Over the last just under 10K miles, my pump average is 45, MFD average is 48.5. This is from October to May. Hopefully it will improve a bit over the summer, though I noticed that A/C usage eats-up about 3-5 MPG so it might not change the average due to the hit from A/C usage (and I seem to need to use A/C when temps are over 80 F and I can't open the windows).

I have almost never had an MFD lower than the pump reading.

'02 with 126K miles...

 

Aha... I think I got it, a priori (trying it now to see if I got it right). And by the way, those codes are called html tags. Thanks. And we caught another rat in a trap under our hood last night (I posted to the "Rats attack our Prius" thread).