How to improve fuel efficiency?

Wait a sec, are you saying the dash display of mpg regularly underestimates? I think I'm just confused; it's practically part of the Toyota Mission Statement...

 

EVERY vehicle I've ever owned that had an onboard fuel mileage display has always been around 6% on the high side.
Maybe 10 of them.

 

I measured mine once at 10% high

 

Yes, those numbers are estimates, which don't well match most people's real world driving.

But because those numbers are the measured output of a very well defined test set, you can also legitimately call them "ratings". Even the EPA itself does:

It has a better city rating (or estimate, or score) than highway rating. But as I've described before, that is an artifact of the artificial test scale that derates (discounts with a somewhat arbitrary fudge factor) the highway test results more than the city results. Real world results still vary, and a fair number of drivers won't see that city inversion.

 

Our previous 06 Civic Hybrid was the first car for us to have mpg display. It was either spot-on, or very slightly on the pessimistic side. As an example, if the display said 5.0 liters per 100 km, my calc would be 5.0 or sometimes 4.9 (lower number better).

 

Yeah my calc's dance around, highighted one a bit of an anomaly:



Average is 7.0~7.5%

 

I'm not sure how you can calculate mpg more accurately than the computer unless you know PRECISELY the level you're filling your tank to.

 

Well, the car's computer does calculate quite accurately I think, and then, before displaying it, adds a fudge factor, likely around 7.5% for third gen.

It's quite easy to get accurate calculated fuel economy. First off, always fill the same way. I would suggest to fill till it shuts off, and stop. This does a couple of things: it's relatively uniform, and prevents the various bad things that can happen from filling to the neck.

Then consider: if you calculate one tank, sure there could be a little discrepancy, due to when the pump shut off. If it shut off a little early, your mpg goes up, and the converse if it shut off a little later.

Say it shut off a little early. Next tank, it'll take a little more gas to fill, and it will balance out.

So with even just two fillups, the fill error more-or-less cancels out.

A long of string of calculated fill ups. It's pretty much DEADLY. And if you keep a running total, all the miles and all the gas, it's bullet-proof.

 

Long term, just normal filling can give you a very accurate number.
The onboard computer is really not that sophisticated either.
That is, it doesn't have a very accurate way to precisely measure the fuel used.

 

Don't blame TOYOTA only - they're all the same. KIA was about 7% optimistic, both FORDS about 5 or 6%, PRIUS about 5%.

It's never going to be accurate anyway. It's reading from a plethora of sensors which are really there for other purposes, and providing you an approximate calculation. PLUS - it displays it in about 9 different formats!!

The Owner's Manual clearly says (about DTE) "... is computed based on your average fuel consumption. As a result, the actual distance that can be driven may differ from that displayed." SUBARU says almost the same. FORD says "... approximate distance the vehicle will travel on the fuel remaining in the tank. Changes in driving pattern may cause the value to vary"

 

My current Subaru, a '14 Forester, is also nearly spot on after adjusting for the built-in odometer error.

Unadjusted, it appears to read a couple percent high. But the odometer itself (not the speedometer) is biased 1.8% low, apparently a reaction to that class-action warranty fraud case alleging defective (over-reading) odometers. Several manufacturers were hit with similar suits.

Applying that adjustment factor leaves a residual MPG error well under 1%, closer than I've cared to calibrate.

 

Most of us ignore DTE. There is a reason @jerrymildred calls it the guess-o-meter. We have been referring to the dashboard economy readings. For you it would be l/100km.

 

I haven't found that. Other cars I've owned (I've got records of every fill since 1975) used 3-6% more with E10. PRIUS is the only car I've owned which has negligible - ie impossible to categorically calculate. When I first got her, I filled her with 91 for 3 fills, then 98/95, then E10 - and almost no difference, not enough to swear that one was better - maybe about 1% better with 98.

I put it down to several factors, the most likely being that it drives 30-40% in EV mode which doesn't care if it's got E10 or 99 in the tank.

 

My dash display mpg is always off by appx 6 - 8%. That’s why I use fuelly to track my actual mileage.

 

Wait. You're telling me the computer that allocates fuel in millisecond pulses with a constant gallon/minute fuel pump is less accurate than going to pumps that shut off with pressure, which could vary with temperature and use? I find that hard to believe.

 

@Gas Mizer, regarding your skepticism of the accuracy of calculated:

I get the impression you picture multiple tank calc's like a novice potter, working with clay on a spinning wheel. The clay is prone to get off-center, develop a wobble, and fly completely off.

Think of it more like a pendulum, that someone's perhaps given a little nudge. Over time it settles down, it's inherently stable.

If you fill by the stop-at-first-shut-off method, record your gas pumped, note the odometer reading, you'll be very close, and any "zig" on an individual tank is more-or-less cancelled out by a "zag" on the next.

Accumulated calculations, for multiple tanks, are virtually 100% accurate.

The only fly in the ointment: someone borrows the car, puts $10 in, stuff like that, lol. Even that you can bounce back from, as long as they kept the receipt, know how much was pumped in. You can lump it in with the next tank.

A good tactic: use the Odometer to calculate the distance travelled, not the trip meter. Or use the trip meter too, but just for feedback, and as a cross-check.

 

Retail fuel pump meter accuracy is regulated by most states' consumer protection laws. Car MPG display accuracies are not regulated by any law, so most of the manufacturers have not chosen to make them accurate, even if they could. The result is many MPG displays being wrong by much more than is legally allowed for fuel pump meters, and car manufacturers not wanting to be taking on the competitive disadvantage of displays being accurate instead of biased high.

Meter shutoff-point is not regulated, but we can work around that by keeping a running fuel log of many consecutive tanks, or even the full life of the vehicle. The error ratio term shrinks as total fuel purchased and total miles driven get ever larger over many tanks.

 

That does not, however, somehow make it not true.

AND.....the shutoff point for the pump has no bearing on the accuracy of the readout of how much was delivered.