Why testing litres change the answer
A dispenser's totaliser counts every litre that passes through the nozzle, and it cannot tell a sale from a calibration check. When five litres go into a test measure and straight back into the tank, the totaliser has moved but nothing was sold.
Counted as a sale, those five litres inflate expected revenue, and the shift then appears to be short by their value at the pump price. Ignored entirely, they turn up later as an unexplained stock gain. Deducted from sales but kept in dispensed — which is what this calculator does — both the cash and the stock reconcile.
A closing reading below its opening is always an error
Totalisers only count up. A closing figure lower than its opening is therefore never a real measurement: it is a transposed digit, a reading taken from the wrong nozzle, or an opening figure copied from a different shift.
The calculator flags it rather than returning a negative, because a negative litre figure quietly cancels a real sale elsewhere in the total. The one legitimate case is a dispenser change — a replacement unit starts its own count from a new number, and the two nozzles must be recorded as two reading pairs rather than as one that appears to run backwards.
Take readings per shift, not per day
One reading pair a day tells you what the outlet dispensed. Three tell you who dispensed it. On the days everything ties out that distinction costs nothing — and on the days it does not, it is the entire difference between a question someone can answer and a number nobody can explain.
It is also the only way to catch a nozzle that was never read. A shift that closes with a nozzle missing its closing figure is visibly incomplete; a day that rolls into the next one absorbs the gap into tomorrow's opening reading and leaves no trace at all.
Frequently asked questions
How do I record readings when a dispenser is replaced?
As two separate reading pairs. Close out the old nozzle at its final totaliser figure and open the new one at whatever number it starts from. Treating the changeover as a single pair produces either a phantom sale or a phantom loss the size of the difference between the two meters.
What about a power cut during a shift?
Electronic totalisers retain their reading through a power failure — the count is the meter's, not the display's. Note the interruption so an unusually low shift is explainable, but the readings either side of it remain valid.
Should I use the electronic totaliser or the mechanical one?
Use one consistently, and prefer the mechanical totaliser where both exist — it is the one that cannot be reset. Mixing the two between shifts produces a difference that looks exactly like a stock discrepancy.