Major disparity in displayed mpg

How did you estimate the .375 of the tank? If you are using the fuel gage, then you are using the wrong instrument. Best to fill the tank and do your calculations upon refuelling a follow-on full tank. Reset the "A" trip meter and then you will have better values to calculate. If you reset the "A" trip meter it will calculate the car's interpretation of MPGs which in my extensive experience will be about 5% overestimated.

 

This^

There is no way to accurately "estimate" how much fuel you used other than starting with a full tank and filling the tank at the end of the trip. The longer the trip, the less error introduced by inconsistent gas pump handles. Trying to figure it out from how far the gas gauge moved is pure guesswork.

And yes, the car's computer seems to always calculate optimistically. Our last tank was actually 51.29 mpg while the car claimed it was 55.8. The previous tank was an actual 52.7 and the car said 55.4. It's really pretty amazing that the trip computer is even that close. As I understand it, the car has a pretty good idea how much fuel it's injecting each time it fires a pulse of fuel. And it knows how many times it does that for each mile. The number of injector pulses per mile is astronomical, so for the mpg calculation to be reasonably accurate, the car's guess on each tiny pulse of fuel has to be incredibly accurate over a 400-500 mile tank.

 

You've tried the impossible.

You can only do it by 1) filling the tank;
2) driving till it's at least 3/4 used;
3) REPEATING 1) and 2);
4) calculating over 2 fills.

A lot of people (including the others I notice who responded) use TRIP A to keep track of every fill. Reset it when you fill, read the figure and reset next fill.

 

If your Gen4 fuel gauge is anything like prior generations, and even very many other car models, then it doesn't show the full tank range. A couple gallons are set aside as safety margin, below the bottom of the gauge. Also, the top of the gauge might not be at the actual top of the tank. This is on top of the uncertainties and nonlinearities of the gauge. Put all these together, and guestimates of fuel consumed based on gauge movement are commonly way off the mark.

Keep a fuel log book, writing down odometer readings and gas pump readings. These numbers are the only accurate way for ordinary car owners to figure MPGs.

 

Just gonna throw this out there as a shower thought about the 5%-7% "better" fuel economy that the car reads versus actual tank measurements. Feel free to chime in as I may have this all wrong, but I think it's even worse than this...

Fuel expands and contracts quite a lot with temperature. In fact, the gasoline/petrol has a temperature expansion co-efficient of 0.00095. To ballpark this, it means that the cold fuel (from being stored underground) going into your tank will expand by about 1-2% purely by reaching ambient temperature. So your 1 gallon or litre of fuel just became 1.01 to 1.02 gallons or litres.

By the time the fuel reaches the engine, it'll have increased in temperature more. For an extreme example, if the volumetric sensor is located within the hot engine area (let's say 80C / 176F) this would mean the volume of the fuel will have expanded by around 10% compared to what it was when your purchased it.

For example...

Let's say you pay for 1 gallon of fuel which goes into an empty tank and drive until it runs out... The fuel will expand to 1.1 gallons by the time it reaches the engine and volume sensor. So when you run out of fuel the car will have measured that you've used 1.1 gallons of (expanded) fuel. Let's say you travelled 100 miles before the fuel ran out. So the car thinks you're getting 100 miles / 1.1 gallons = 90MPG. But your own measurements read as 100 miles / 1.0 gallons = 100MPG.

At the very least, even just from the fuel expansion going from the ground to ambient temperature, you'd expect the car to read around 1% to 2% worse MPG than your own calculations, as its measured you using 1% to 2% more fuel to go a given distance than what you actually put in the tank.

Therefore, the MPG the car reads, which is already claiming around 5% over, is actually more like 7%-15% out.

 

I'm a rebel - I use Trip B.

 

Ahh - yep. I use "B" if I'm on a longer trip - like for a day or 6 - just to track that trip.

 

I use the odometer. A trip meter too, but just for feedback during the tank, and for a check.

I've got about 150 fill ups logged. The Toyota displayed mpg is invariably optimistic, around 7% on average. Now why would that be?

 

I measured once on my Gen 4 and found about 10% high, I am lazy but old school. If I want to know actual I can do the math in my head to subtract 10%.

 

I keep a spreadsheet showing "actual" versus "displayed" mpg, along with other data. I use Trip A for my daily commutes, Trip B for tank-to-tank.

And, as I've mentioned elsewhere on PC, there is a United Nations regulation (No. 39 - see attached) that requires the speedometer to read high by applying a formula. Section 5.4 of the attached regulation spells out this requirement. IMO, this is to protect manufacturers from lawsuits. Think about it, if the speedometer reads below the car's actual speed, and you're doing the posted speed limit based on the speedometer's readout, you're over the speed limit.

The variance between actual vs. displayed mpg will vary between 2% and 10%, depending on trip and weather conditions.

 

I don't understand what the (by regulation) speedo error has to do with displayed mpg error.

 

This has to do with the magical fairy dust in the car's ECUs. A tale for another time...

 

You can NOT start your hand calculations "with half a tank".
All the rest of this discussion is basically just a waste of time.

 

Variance between ACTUAL and DISPLAY l/100km - just accept that it won't be right. It is much the same with all brands. Pick a brand, GOOGLE the same query - and each will have comments about it. And it seems between 5 and 10% - some even more.
VW:

Jaguar: The Jaguar's display read 46.8 MPG, a 6.5% difference.

PORSCHE: "Boxster is closer typically off from actual by 10-15% also on the optimistic side"

BMW: BMW's trip computer reported mileage 1.4 mpg higher—nearly 6 percent

 

Our 06 Honda Civic Hybrid's mpg display was either spot-on or slightly under.

 

For my Gen3 and two PP, Toyota always had an mpg display more optimistic than actual. I also keep the spreadsheet record of a full-tank hand calculated mpg and displayed mpg for the tank. BTW, I also use Trip B for this as well as Odometer to verify the trip distance. The 2015 Gen3 mpg display discrepancy was average +2.13% over 77 fill-ups. The 2017 Prius Prime mpg was +5.99% over 53 fill-ups, and the 2020 Prius Prime is 7.00% over 4 fill-ups.

 

What has your avg been in the 2020?

Posted via the PriusChat mobile app.

 

Overall 73.42mpg, but that number includes EV miles. By extrapolation, the current HV only mpg is 61.72mpg.

 

Agree you have to analyze using full tank and by resetting A or B to track your miles and then at refill. My theory regarding the fuel gauge is that approx 1.5 gallons of fuel are used before the meter starts to tick down from full, and that there is probably 1.5 gallons of fuel after the gauge hits empty, and that is why you can get so many more miles out of the 'first half' of the tank then the 'last half' of the tank, assuming you refuel when it hits empty.

My best mileage was on my original tires. When I switched my tires at 70,000 miles, my overall mileage dropped slightly. My current overall average is about 58 MPG, but it depends. Long road trips (I just did a round trip from FL to OK) will be low with all the fast highway miles, and cold weather really drains the MPG when you have the heat running.

I see you are in NYC - first 3 years I owned my car was in NYC and I had some of my best MPG there especially the highways drives - with the the consistent slower speed and frequent spurts of traffic.