Fuel Consumption Conundrums

Fuel economy is one of those topics that everyone seems to have an opinion on, but which very few can give accurate facts about. Unless you're a hypermiler or a Prius owner, you're probably about as adept at calculating your fuel usage as I am at calculating income tax returns (and no, I don't know how to do that).

But even for those of us who do know something about calculating fuel use, sometimes, the numbers just don't add up. We fudge some numbers, try to make concessions to test methodology variations, and... well... sometimes we just have to guess.

Terrible, isn't it?

This has been on my mind for the past few days due to a very unusual car... the Toyota Corolla.

It's not an unusual car, per se, but it's unusual in the fact that it has a very big engine in it... a two liter lump from the Toyota Camry. Now, I've always championed the fuel economy of Toyotas... though their cars of late don't impress me otherwise... as a sign of engineering expertise used correctly. I wasn't expecting Yaris or Prius-like numbers from the combination of small car, big engine, auto-box, but I was hoping for the Corolla to post numbers significantly better than some of the two liter porkers that have gone through my hands, lately, like the Dodge Caliber or two-liter Ford Focus.

But did it? No. Maybe. I don't know.

Confused? So am I.

My co-tester, Carlo, drove the car for two days. Claimed brilliant economy... doing 9 km/l on one day, and an amazing 16 km/l the next. All in-city driving, though the 16 km/l was not stop-go traffic. I was, frankly, skeptical, but he had the gas receipts to back it up.

My turn at the wheel got me two days (with two separate fill-ups) of 6.5 km/l in pure city weekend driving. Mostly low speed stuff between stoplights. Just about what you'd expect from a slushbox equipped two point oh on Manila streets... so... maybe the car is better on the highway, right? Automatics love highway driving, with their long gearing and smooth shifting.

On the highway, the trip computer, previously pessimistically under-rating economy by 0.2 km/l, was enthusiastically proclaiming a mixed 12.5 km/l, with highs of 16-18 km/l at maintained speeds of 80 km/h. It only sunk down to 12.5 km/l due to the stops we made along the way.

Imagine my disappointment when a trip to the pumps and a quick set of calculations on the back of the gas receipt netted me 10 km/l. A quick drive around, to do acceleration and handling testing, (with the trip computer reset) gave me 7 km/l by the clock and 8 km/l on the receipt from the same pump.

What the hell was going on?

Granted, you expect onboard fuel consumption computers to be inaccurate, but you don't expect this kind of schizophrenia. The odometer was accurate by our GPS meter, but that doesn't really matter if you're comparing onboard calculated economy to trip economy based on fuel consumption and odometer readings.

The first clue I had when trying to puzzle this all out is the "Range" readout on the Corolla's trip computer. We make it a practice to fill up the tank to the automatic shut-off on whatever pump we're using. Of course, there are pump variations, but with our schedule and drive locations, you can't always return to the same pump. But looking at the "Range" readouts showed how ridiculously large the variance in fill-up levels was on the Corolla. On most pumps, the "Range" went up to 299 kms at the "full" mark. On the last pump I used, it went up to 360 kms. That's a difference of potentially 3-6 liters! And the "Range" calculations don't reset with the trip computer... they're calculated internally, and don't change whether the last hour was driven like a hypermiling loon or a gas-guzzling hoon.

This indicated that, much like the Prius Toyota had lent us two years ago, the Corolla's filler neck is a literal pain-in-the-neck for accurate fuel consumption. Most cars will have a variance of one or two liters, but this was off the scale.

Given some quick corrections based on assumptions of how far off each pump was, we ended up with the 16 km/l trip being around 12 km/l... one 6.5 km/l trip being 7.5 km/l, and the 10 km/l trip being 12-13 km/l. The trips where we were able to fill up at same pump (6.5 km/l in the city... 8 km/l in hard driving) remain unchanged. Maybe. Maybe there are bubbles in the system that make these inaccurate, too. We're not really sure.

But short of jumping up and down on the rear bumper at the gas station and topping off till gasoline comes out of the overflow pipe (which would render all calculations moot, anyhow), I don't know of any other way we can assure reasonable accuracy, except to drive far enough to deplete the tank and lower the percentage of inaccuracy.

If you're doing this kind of testing at home, that's what you've got to do to make sure you're getting real numbers. Drive as far as possible on one tank then fill her up (at the same pump) and measure from there. It's not 100% accurate, but it's the best you can do without taping a fuel bladder to the hood of your car.

For our next Toyota test... that's what we'll probably be doing, maybe with an IV bottle stolen from the hospital dispensary.