Short measure at the pump?

[JockoT]

I filled up with petrol, this morning, and I believe I got short measure at the pump. I went into my local Morrisons, but was unable to use my usual pump. I filled up as usual, noted my mileage and the reading on the Fibometer and my ScanGauge, and thought nothing more of it.
When I returned home, and did the calculations on the spreadsheet, I was very surprised to see I had achieved 46 mpg. This is my worst ever figure, even worse than driving in snow during the "Beast from the East". I double checked all my numbers, and even went back to the car to check I had recorded the correct mileage.
As you all know, I am rather OCD when it comes to mpg figures. 46 mpg could be the real number, except I had easy journeys, the Fibometer was reading 55.8 and the ScanGauge 55.4 mpg. As I have reported elsewhere, both these are usually within 1 or 2%. I have never seen a 20% error.
This afternoon I have been for a gentle 85 mile trip, the Fibometer (reset at top up) says 68 mpg and a check on my return found all wheels cold and all tyre pressures showing 36 psig. If the petrol had been leaking out I would have surely smelt something, especially as the car is kept in a garage with carpeting under the car. For the same reason I do not suspect that petrol has been siphoned off.
The other pump on the island I used was out of order, not suspicious at the time but ringing alarm bells now. I have contacted the store, making my concerns known, and also contacted Trading Standards. I will keep you informed.

[trollbait]

If the pump had shorted you, your fuel economy would have been better than average(less fuel divided among the distance). With worse numbers, this pump overfilled your tank in comparison to your normal pump. Such variations are why single tank measurements can be suspect.

If the station was actually cheating, the pump would be displaying more than it dispensed, and you wouldn't notice anything until you saw shorter distances between fill ups.

[JockoT]

No. If the pump was dispensing half litres instead of litres it would put twice as many in to fill the tank, charged me twice the price and my miles would be divided by twice the amount of fuel I actually got, giving me half the mpg.
I understand that a single top up does not give an accurate figure, but not an error of that magnitude.
If the pump had cut out before the tank was full I would have paid less and got a better mpg than expected.

[trollbait]

Okay, I see what you meant by short now. Referring to the unit of volume dispensed being less, not a fill up short of full. If there is the case, the pump might have dispensed 85% of what it said it did.

But looking at your log book, while this current fill is your worse result, it isn't markedly so. Then the previous fill to this one spiked up from the the few preceeding it. When I see this pattern in my logs, I assume the previous fill up was short; tank not filled all the way, and the current low result fill correcting for that. Calculating fill #48 and 47 as one tank yields a MPG value slightly higher than what was reported for #44, 45, and 46.

[JockoT]

The weather here is warming up considerably we are experiencing a particular dry spell, and my journeys have been much more FE friendly, hence the increasing mpg figures. I actually expected this one to be as good as the last, and the ScanGauge and on-board readout indicated so. I use SPC (Western Electric Rules) and today's data point was only just in zone B.
You may very well be right about a spike last time but I have never seen such a discrepancy between the calculated figure and the two displays. The on-board display has no connection to the amount of fuel added at the fill up, other than that I reset my Trip at that time.

[JockoT]

What I didn't say was, a full top up for me is normally 7.5 gallons from the light coming on. I once got 7.9 gallons in after running for a couple of days with the light on. Yesterday the gauge had just passed the half way mark and I got 6.5 gallons in! I know petrol gauges are not accurate, but you learn the foibles of your own car, and I know my gauge doesn't drop from there to zero in 50 miles. Nearer 150 miles, in fact.

[trollbait]

" The on-board display has no connection to the amount of fuel added at the fill up, other than that I reset my Trip at that time. "

In the case of per tank values, there could be a connection. Generally, the tank level sensor reads full before the station pump shuts off. That's why it can seem like the gauge stays of F forever before it starts going down. Since the volume above that fill mark isn't seen by the car's computer, variations in that amount could throw off its calculations. Same with the Scangauge; each time you tell it a fill up occurred, you are recalibrating it.

There can be a substantial amount of volume above that full point. Wayne Gerdes at CleanMPG once got more than 10 gallons of gas over a pick ups rated capacity. Of course, his fill up procedure is as extreme as his hypermiling.

Never owned a Honda, but the gauges on Fords seemed to measure two thirds of the tank volume over the top half of the gauge. Assuming that the tank was properly filled. The Chevy gauge increments seemed closer to fill point of the tank. The Prius gauge used digital pips, and the tank had an inner rubber bladder. How much gas you had could be a mystery at times.

The pump you usually use was down for repairs. Did you use it for the previous fill? Perhaps it 'broke' before you used it.

If you were cheated, I hope justice is served.

[JockoT]

My normal pump was not being repaired. There was a car using it so I went to a vacant pump. The vacant pump was on an island with a pump out of order.
The on board display does not read the fuel gauge. It reads the odometer and the pulses from the fuel injection system and calculates from those.
The Jazz fuel gauge is pretty good in that the mid point is reasonably mid placed, mileage wise, between full and empty. I get roughly 400 miles on a tankful and hit 200 miles, or there about, at half way on the gauge. Above and below that are not linear, but aren't too bad.
I have anecdotal evidence of other drivers in the UK, using pay at the pump units, having the pump start to run at zero litres and zero price, but charging up before the trigger is squeezed to dispense into the tank. I have no idea if that happened to me or not. I believe it can happen if the non return mechanism in the pump is defective and the pump loses its prime.
I am not too worried about the short measure or the few pound it cost me. It is just that it appears to have thrown out two years meticulous (some say OCD) fuel calculations and record keeping.

[JockoT]

Just received an email from Trading Standards. They are going to test the pump I suspect was faulty.

[JockoT]

Filled up at my usual pump this morning and 63.7 mpg. I knew that the numbers were increasing, as the weather and the traffic improves.
I still feel that the last top up was not a genuine figure.