OT ICE engined cars to be banned in Netherlands?

[Nightjar]

On 27-Aug-16 12:19 PM, Dave Plowman (News) wrote: ...

The instantaneous display is a moving bar above a graduated scale. Unfortunately, when changing between mile and kilometres, the scale remains the same and it is the response of the bar that changes. As the highest graduation is 40, the bar remains filled most of the time and I have no idea how much better than that it is doing. It does give a readable display on continental motorways, but I've never been that worried about consumption that I've ever noted it down.

[NY]

Oh yes, there's a correlation but there's a lot more scatter than I'd have expected.

The figures are:

Honda

slope 0.3780 intercept 27.7317 r 0.5428

Peugeot

slope 0.6627 intercept 22.2030 r 0.7748

In the ideal case, I'd expect the slope to be 1 (ie the same mpg figures from computer and from my measurements, across the whole range), intercept to be 0 (the best-fit line doesn't show any offset error) and correlation coefficient 1 (where 0 means no correlation and +/-1 means perfect positive- or negative-slope correlation with no scatter).

The fact that slope is a long way from 1 (especially for the Honda) and there's a large intercept is intriguing: it's saying that for a theoretical

0 mpg as measured by me, the gauge would read 22 or 27 mpg.

[NY]

The display on the Honda is like that. The maximum mpg that it can display is 60 mpg, so you can't work out whether the car is more efficient descending a long hill in neutral (or with clutch pressed) or remaining in gear (linking to the thread about engine braking and coasting), whereas my Peugeot has a digital display that shows clearly that slowing down in gear with your foot off the throttle gives "999 mpg" (ie effectively the ECU has cut off the fuel) whereas when coasting it's around 200 mpg (ie very much less fuel used that normal, but a finite amount is still being used to keep the engine running).

On both cars, the instantaneous mpg fluctuates wildly even when you think you're driving at a constant speed on a level road.

[The Natural Philosopher]

which is why a 'trip' mpg indicator is so much better.

[Dave Plowman (News)]

That's what you do - but you do need to start and end with the same amount of fuel in the tank. If you do it over several 'fills', any error become less. But ideally, you'd first brim it and then finally brim it using the same pump.

But even on the old Rover, the OBC gives a figure so close to accurate as to make no difference. So I'd expect this with all more modern ones. It's more likely to make an error when calculating the old fashioned way.

[Dave Plowman (News)]

Does it have an overall one too? If you reset that, it will give near enough an instantaneous one at a steady speed - for a short time.

[NY]

Agreed. The trip mpg (ie average from one filling to the next, typically for most of a tank of fuel) is the only one that is useful. The instantaneous one needs to be averaged over a longer time (eg several seconds) to weed out the jitter that hides the real info about how much the consumption is increased during acceleration or going up a hill compared with the steady-speed-on-the-flat figure.

The "range left" is useful as a cross-check against the fuel gauge. I know that a full tank gives me about 700 miles range, so at any time, miles done since last filling (I reset trip meter when I fill up) plus estimate miles left with remaining fuel should always add up to about 700. If I have an estimated 100 miles left and the fuel gauge reads half-full, I'd be suspicious that the fuel gauge was mis-reading and I'd fill up just in case!

[The Natural Philosopher]

When I had one of those I used to run mine to -25 miles range regularly ;-)

[Nightjar]

I have readings since reset, which I chose not to change as it tells me that, since new, the car has averaged 37.3mpg and 27mph. If I were sufficiently interested, I could tell it to use kilometres, which would give me km/l on the instantaneous display, but that would also set my speed limiter, which I use constantly, to kph and I would need to convert km/l to mpg to mean anything to me.

[Vir Campestris]

The point is that the different cutoffs fill the tank to slightly different levels.

If you've used 50 litres, and brimming on one pump leaves a litre gap, but the other fills to the brim, that's a 2% error.

Andy