# Measuring gas flow

When I was working in the chemical industry, one of the hardest things we had to do was to accurately measure small liquid flows. And the instruments cost over \$3000 each. Today, almost every car has a device that measure miles per gallon. How do they do that?

Operative word here is “accurately”, which MPG gauges generally don’t do. I’m not quite sure exactly how they all work, but I believe the computer more or less deduces a probable fuel consumption rate based on the vehicle speed sensor output and the fuel injector pulse rate. I think some also do it with the fuel level sensor which makes it even less reliable.

Well they are not as accurate as you might think. A 10% error would be a small error for them, but in the chemical industry there are many applications where 10% would be totally unacceptable.

They don’t measure actual fuel flow. They measure the electric impulses of the fuel injectors and convert that to flow. Very cheap and easy to install. Factory calibrated.

The computer assumes that there is a constant flow rate of fuel through the injectors while they’re open.
Real time flow rate is calculated by multiplying injector pulse dwell, injector pulse frequency, the number of injectors and a correction factor.
MPG is calculated by dividing vehicle speed by flow rate and multiplying in another fudge factor.
Voila!

Just speaking for the one in my Lincoln, it displays figures based on a number of inputs; the PCM, speed sensor, RPMs, fuel tank sender, fuel temp sensor, etc. It even has an anti-slosh module inside the display module to account for gasoline expansion and sloshing. A computer within a computer if you will.

The display in my Lincoln is amazingly accurate although it took me a year or so of filling and figuring before I learned to fully trust it. If there is a variation from the actual mileage figures it’s so slight that it can’t be noticed.
About once a year it seems the unit will go into a “stupid phase” for a day, give some goofy readings, and then straighten back up.

Is there a particular condition when it goes into the “stupid mode” I mean tank half full, completely full, a lot of stop and go or straight highway, weather etc… any thing that might be causing it.
My boss has an 5 series BM and he trusts the Miles to go on it like any thing. I personally have never had the courage to go below 1/4 tank.

In the old Fox 'Birds, Ford used to use a fuel flow sensor in the fuel line to measure flow. Wonder how long they kept that up.

It’s just a vacuum gauge that reads zero to 70 MPG. They don’t measure the fuel flow. It’s useful but it’s not high science.

It used to be just a vacuum gauge, back in the '70s when they made it an option on some cars, but they do make a more earnest attempt to measure fuel flow than simply measuring vacuum these days.

Well, on a carburated car, the fuel flow is more-or-less a function of the engine vacuum.

No, there’s no pattern to it going stupid. It can happen anytime and under any situation but it’s something I don’t worry much about since it’s a rare thing to happen. It’s been a long time since it went stupid on me but I think it was trying to do that today. It was showing about a 2-3 MPG drop in mileage and an erratic “Miles to Empty” figure. Hopefully that’s the once a year hiccup. Seems to be fine now.

In deference to a post below, my car does not have or use a vacuum gauge, but does use a number of other things in doing those computations. Intake manifold runner sensors, fuel flow sensor as mentioned by mr. josh, and a handful of others.
No wonder that car weighs over 4000 pounds; it’s absolutely stuffed with wiring and sensors.

When cruising, sure, but when you step on the gas the accelerator pump “pumps” gas into the intake manifold, which is not dependent on manifold vacuum.

Well, it gives it one squirt to keep it from dying out when the manifold vacuum drops suddenly, but once the vacuum spike is over and vacuum increases because of RPM, fuel metering is vacuum-dependent again.

I rented a new Ford Explorer V6 for 4 months earlier this year. The computer consistently gave me 16.5-17 mpg city and about 18 mpg highway. The car was realtively new, so it might improve later. It probably uses the same input simulation factors as the Lincoln. Since it does not measure gas flow directly, it might be accurate one day and not the next. The fuel gage and “low fuel” signal are useful, but I think the money could be better spent on other things.