Two alternative methods for measuring MPG in an old diesel.
 

Dear fellow experimenters,

There are two possible alternative methods to determine fuel consumption in a non OBDII diesel vehicle.

I would call this methods as

1) the VSS - MAF method.

2) The GPS - aerodynamics method.


1) The VSS - MAF method.

It works as follows:

MPG = VSS / MAF * 7.718 for gasoline engines.

The constant is the result of units conversions (14.7 kg of air per kg of fuel).

See his complete paper here:

http://www.windmill.co.uk/obdii.pdf

Why cannot we use the same idea to estimate fuel consumption in a diesel?

He also presents an electronic circuit which is relatively simple and cheap.


2) The GPS - Aerodynamics method:

This method uses no connection to car sensors. It just uses a GPS to determine horizontal speed and vertical speed, as well as real km.

For determine the fuel consumption, it uses the aerodynamics and weight od the car, as well as the power of the engine. This method is used in an Android application called "Digifuel". This application is already available for gasoline engines. Equations are needed for diesel engines.

The idea is to input the characteristics of each particular car and the rest is done by the GPS.

Both methods need a display to show results in real time, as well as a processor circuit for calculations.

Any comments?

OldBeaver

The biggest challenge is measuring fuel flow. With a diesel you cannot infer fuel flow by measuring air flow, at least not accurately, because fuel mixture is varied unlike a gasser which needs closely held mixtures to prevent knocking and overfueling. That is why method one will give unreliable results, as it simply does what my Scangauge on my OBDII diesel does, infers fuel consumption based on airflow.

Method 2 may be your better option, but that will be vehicle specific based on turbo, injection method and injection pump type.

Honestly, I think the most direct solution may be a pair of high resolution flow meters, one on the supply, one on the return, and fuel consumption is the calculated difference.

In my opinion, neither option is worth a hill of beans, particulalry if you're hoping to use the FE info as feedback for driving modification.

As UFO has correctly pointed out, the first option assumes that air-fuel ratio is constant--a major error for a diesel. A constant air fuel ratio of 14.7 is a reasonable assumption on most gasoline cars, but AFR on a diesel can vary from neary 100 at idle to down near 15 or so (depending on speed, load, and engine). On transient cycles, the engine can vary from 100 to 15, back to 100 in a matter of a few seconds. Without any way of knowing whether you're at 15 or 100 AFR, there's no good way to get fuel flow from air flow. Gassers are a differnent story, because they are controlled to a nearly-constant AFR.

The second method is little more than an educated guess. It basically assumes any vehicle of a certain size with a certain weight and a certain engine will burn a certain amount of fuel. They're obviously huge problems with that. Lots of ecomodders have doubled they're fuel economy with driving mods--and they're size, weight, and engine are pretty much unchanged.

Unfortunately, I'm not aware of any effective, affordable ways to get diesel FE on pre-OBDII engines. If you'd like the details, I do know of a way to do it with 1 flowmeter as opposed to 2 as UFO suggested, but it's by no means cheap or simple. There's also a way to do it with air flow info and a wideband O2 sensor, but that's not cheap or simple either.

Comments to comments on alternative methods for diesel mpg
 

Dear UFO and Diesel_Dave,

Thank you guys for yr comments.

Well, it is clear to me that variability in air/fuel rate is too much for having a useful estimate of diesel mileage.

I have a Skoda Octavia diesel 2009 car with a onboard computer from Factory, and mpg variates a lot even in a flat surface with cruise control. A range from 10 to 100 is useless if we want to test mods and driving habits.

For making a measurement I use mean consumption with cruise control in a flat surface, starting recording at 70 km/h, not from stand still.

Any method will have, at least, the same variability of my Skoda.

The second method, using just GPS speed and aerodynamics, weight, and so for, must be calibrated several times comparing it with a full tank fuel consumption, to get any result close to reality. Even though, I expect a lot of variability o the road.

I tried to make a two flow sensors system, which "worked" when installed in the protoboard. I say "worked" but I cannot say worked well, because it didn´t. While soldering in a case for final versión it stopped to work and both flow meters begun to spill fuel and finally stop working. One main problem of very small fuel meters is dirty in fuel, that ends stucking the meters.

Oldbeaver

What type of flowmeters do you have?

I have acquired a clamp on sensor that used to indicate if there is a fuel pulse in the injector line. I'm going to see if I can use it with my MPGuino to get an estimate for fuel consumption. The clamps themselves are not cheap( around $200 American new) but might be found on Ebay in the test kits that I got mine in. It was in a Mac Tools tester that indicates pulse and one version also generates an RPM reading. I'm wondering if I can also use this for the rpm gauge signal. Might not work at all, but as I have all the parts, why not.

Flow meters
 

If you ask me, none, I discarded them.

They stop working (I think due to the debris in fuel) because they are too delicate.

Besides, the connections begun to spill fuel, and as they where plastic, I couldn´t tight them too much. They also don´t recommend to use any sealing paste.

Oldbeaver

What kind of flow meters do I have
 

No kind. I have been searching in the net and the best (for me) are dual chamber differential flow meters. Problem is they are expensive, and installation will never be as good as it should, with proper tools or robots in a Factory. Result: you need to cut the fuel pipes and insert the fuel flow meters.

The majority of these fuel meters have a small (and very delicate) turbine inside that generates a pulse every volumen of fuel. These turbines may probably stuck with the dirt of the fuel, despite filters.

Ultrasonic non invasive or such meters are the ideal.

My Skoda car has a Factory meter which has never fail, as well as many Peugeot and such cars. Sure these are not turbines and are Factory installed, fully sealed.

Oldbeaver