calculated vs. computer MPG - Please post your results

Wow... so the MFD can be widely inaccurate?

In my case, I'm hoping that it's underestimating... I'm showing about 41 MPG on my current tank (100+ miles), but on the other hand, the fuel gage still has 9 out of 10 pips showing, which seems to indicate better mileage than that...

 

with well over 5,000 miiles on our 2010 Prius the MPG guage is consisitently off by 2-3 MPG,. I see other owners are reporting the same.

Alfon

 

It is not so much widely inaccurate, but more consistently ~5% overestimating the MPG.

Kris

 

I filled up for the second time this week. I drove 636 miles on this tank, and the car took 10.060 gallons.

MFD: 65.9 MPG
Calculated: 63.2 MPG

The difference of 2.7 MPG is around what most people are reporting.

As an aside, this tank included a 300+ mile roundtrip to NY and back (mostly highway, with 2 passengers and luggage in trunk), which my V averaged 64.2 MPG on the MFD. Not bad.

I've been using Shell 87 so far, but it is ~$0.25 a gallon more than the other stations. Am I wasting money???

 

First fill-up: 46.6mpg displayed, 47.3 calculated.

 

Finally my first tank of gas...

Miles: 468 @ 9.2 gallons
Displayed MPG: 53.6
Calculated MPG: 50.8

Right in line with most of the other posts.

Peter

 

After five tanks and a bit over 1600 miles, my lifetime pump MPG is 54.6 mpg. Computing fuel consumed according to the MID mpg displays for each tank, it is overstating my mpg by 4.3%, or 2.3 mpg.

The last refill was sloped towards overfilling the tank, so I hope the error will fall with more time.

Does Prius compute fuel consumption from some actual volumetric measure, such as injector pressure and time? Or does it use indirect methods similar to ScanGauge? The later will tend to underestimate fuel use / overestimate MPG when alcohol or oxygenates are blended in.

 

I've now completed my first 10 fill-ups, and I've driven my new 2010 5,000 miles. It looks as though the car's calculated MPGs are consistently registering 5% higher than the fuel economy calculated by measuring each fill-up. This is the equivalent, for my mileage, of having the MID screen's FE at 2.6MPG higher than the number arrived at by calculating FE based on gasoline purchases.

Here is a chart showing the percentage by which the 2010's MID readout exceeds the calculated fuel economy. One line is for the percentage above for each individual fill-up, while the other line is the cumulative percentage by which the computer number exceeds the calculated one.

 

Raw Data July 26 fill-up:

A-trip: 475 miles, 55.0 mpg, 32 mph average speed
Fill-up: 475 miles, 9.8 gallons, 48.5 MPG

Driving profile:

50 miles -> 52 MPG
120 miles -> 48 MPG (wife driving, not reset)
160 miles -> ~52-54 MPG (my return at legal speeds)
140 miles -> around town, ~60 MPG

Analysis:

I'm going to suggest that there is a flaw in the 'carry forward' that may treat the previous mileage, the last power-on cycle, MPG, as a 'starting point' and the current trip is added to that point for each power-on drive cycle. Rather than keeping a cumulative fuel burn like it currently keeps the mileage, it uses the previous MPG as a starting point everytime the car is started. This means an earlier, longer, terrible mileage segment gets 'eaten' and the displayed value will be heavily weighted by the most recent mileage.

This means the 'tank' display, whether Trip A or Trip B, can easily be off by variable amounts based upon the driving profile and not a constant error rate. In fact, it may seem somewhat random.

TESTING:

To test this hypothesis, we'll need to record each trip in a log while keeping the other trip meter set for the tank. Then we'll have to sum up the individual trip records with a spreadsheet and compare with the calculated value. For good measure, include one high speed segment of say 100 miles in the middle of a bunch of urban trips.

Bob Wilson

 

Roughly speaking, my calculated MPG over the last 3 tank fills is about 2 or 3 MPG lower than indicated. What's worse is, with no changes in driving habits, my average MPG has dropped from near 50 to 45. Unless my 67 year old wife has taken to drag racing the Prius when I'm not with her, I can't imagine why. ?????

 

We have about 4500 miles on ours and on every tank the calculated is off 3 to 7 mpg. My wife gets about 54 mpg on the fcd and her tanks are usually 4 to 5 mpg less calculated. Last week I had two 500 + mile fills and The first was 5.5mpg less and the second 7.1 mpg. This was with a fcd display of 62 mpg which included a lot of around town in Madison for Hybridfest. The second read 64.4 and it was fully highway and was off by 7.1 mpg.

5 to 10% is out of the box for error. I fill up at the same station, same pump, same number of clicks, and same direction each time with the exception of the fill in Madison last week of which I estimated slope at the station I filled at.

I stopped at the dealer and had them check it out to get a paper trail started. Nothing was found so I followed up with a call to start a claim with customer care. I have claim number and expect a call back this week.

 

Our is off 2-3 MPG on a regular basis with over 5,500 miles.

I am curious to see if the computer could be adjusted to reflect this discrepancy.

alfon

 

Bob,

I think I've been doing this. From the very beginning, I've been recording the two Trip Odometers separately, one for City vs. Hwy driving and the other for each tank. Clearly, the two never meet, exactly, but if I record the Trip A info at the same time I fill up (Trip B), and I use that partial information to develop "lifetime" numbers for Trip A fuel economy, then I get the same overall fuel economy there as I do for the Trip B tanks.

 

I hadn't noticed a problem until the recent trip to Nashville when my wife drove way too fast. But the protocol I'm suggesting is:

• Fill-up tank and reset Trip A - zero both
• For every "power on"
  • Record in a log the previous B and zero just B
• Fill-up tank and record Trip A and gas reciept

Then we take the gas receipt, Trip A data, and all of the Trip B data and calculate the fuel burned for each Trip B segment. Sum up the Trip B items for the fuel burn and compare to the gas receipt for the accuracy. To make sure it is a valid test, drive one or more Trip B segments 'very badly' (aka., very high-speed, hill-climb, into a head wind) and end on some very nice segments (aka., speed under 30 mph for at least half an hour or longer.)

It is the sum of the Trip B segments manually recorded each time the car is "power on," that I suspect will agree with the pump. However, I suspect a mixed mode, especially with an early "pedal to the metal" followed by the rest being 'nice' segments that we'll the biggest error between Trip A and the pump.

In math, this is a problem of weighted averages where the weights are not proportional to the actual mileage. For example, trying to average the averages without calculating the actual fuel consumed. In such cases, the results resemble badly sampled data and will often be misleading.

It might be easier if I show my work using my wife's car, which I filled upon Saturday. But I'm using Trip B to calibrate the ECO power-bar. So it may take several weeks before I have the data to show for this problem.

Bob Wilson

 

My wife's 4Runnner (I'm trying to downsize her car) is off by about 5% also, is this Toyota just trying to look good?

 

Bob,

I would post my ridiculous spreadsheet here, but then it might just scare people due to excessive ugliness and silly detail. Please PM if you'd like to look at my raw data to see if it has any useful info for you. I do have several shorter Trip A data points all within a Trip B (tank fill, for me), and it may give you something to go on. I could start recording the progress point of my Trip A data at each fill up, also, as this could bring me closer to what you are now seeking.

 

The trips that have the highest error for me are the ones that show the highest mpg on the fcd. The 2 tanks that I spoke of were straight highway driving in one day, running a range of 55 to 65 mpg and being passed by everyone except the cars broke down on the side of the road.

My wife drives it like a normal car around town and back and forth to work. She doesn't speed, but she does sit in it and talk on the phone with it running with the ac set on 65 until she gets cold.

 

Two comments about the A/C: First: I haven't yet noticed that A/C is a negative on FE for the 2010. Perhaps it is on short trips, but I don't often use it that way. Once I get up to 45mph, I close the windows and turn on the air. I use the A/C on Auto all the time, and I NEVER did that on the 2007, because it was very clear that mileage was negatively impacted by such a move.

Second: I just don't understand why anyone would ever set the Auto Temp at either 65 or 85. My wife does this all the time, and I can say ('til I'm blue in the face) that the system will do whatever it takes to get it to your desired temp, so why keep adjusting it all the time. After all, that's the benefit of the Auto feature, isn't it? This is particularly big in the winter, when I'd rather have the fan on low or off until it is able to blow WARM air and not COLD air!

 

Let me suggest we start with something like:

	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6	Column 7	Column 8	Column 9
0	Trip	mile	MPG	Avg	Trip	mile	MPG	Avg
1	A	. . .	. . .	. . .	B	. . .	. . .	. . .
2	A	. . .	. . .	. . .	B	. . .	. . .	. . .
3	A	. . .	. . .	. . .	B	. . .	. . .	. . .
4	A	. . .	. . .	. . .	B	. . .	. . .	. . .
5	A	. . .	. . .	. . .	B	. . .	. . .	. . .
6	A	. . .	. . .	. . .	B	. . .	. . .	. . .

It doesn't matter which trip meter is used for the whole tank (or until fill-up) as long as the other one is reset frequently, preferably on every "power on." The habit should be to record when you stop and either reset on then or when you "power on" next.

What we can do is load this into a spreadsheet and compare the running fuel consumption of the meter on the tank versus the sum of the individual trip meter measurements. Then in the end, we compare it to the pump.

To replicate the Nashville trip, I'll do some high-speed stuff on one of the earlier, short section. The rest will be urban stuff.

Bob Wilson