The meter records how much fuel left the tank. It has no idea which vessel received it — and that gap is where every unexplained litre hides. Closing it means binding volume to identity at the moment of transfer, then reconciling against what the machine actually burned.
A flow meter does one thing well: it counts volume leaving the tank. It has no idea which vehicle, drum or generator received that volume — and that single blind spot is where unexplained litres hide. Without attribution, your ‘loss’ is just the sum of everything the meter could not label.
The instinct is to call the variance theft and start watching people. Sometimes it is theft. Usually it is a measurement-and-attribution failure. The fix is not more suspicion; it is binding each dispensed volume to an identity at the moment it is drawn, then reconciling against what the machine actually consumed.
Unauthorised dispensing — fuel drawn into a vessel that should not have received it, counted by the meter as a perfectly normal transaction. Ghost transactions — a dispense recorded with no delivery, or a delivery with no record at all.
Siphoning and meter drift — tank level falling with the engine off, far from any dispenser; or a slow, consistent meter bias that reads as steady low-level loss and never trips a threshold. Temperature variance — volume genuinely appears and disappears with ambient temperature, real physics routinely mistaken for loss, and a convenient place for real loss to hide. Only one of the four is ‘someone stole fuel’.
The place to fix attribution is the moment of transfer, enforced in hardware. The operator authorises with a credential before the pump will enable, recorded against the transaction. An NFC tag on the receiving vessel is read — and held for the full duration of the draw — so the volume is bound to a specific machine, not just a timestamp.
Metering is positive-displacement or Coriolis, temperature-compensated to a reference volume, so the temperature column stops masquerading as loss. The transaction is then reconciled against consumption, stored, and ticketed where required. And because the fuel often travels to the equipment, this lives on a mobile dispenser — truck- or skid-mounted, with store-and-forward records from sites that have neither mains power nor reliable coverage.
The most important design choice is the least technical: establish a trustworthy baseline first — meter calibration, temperature compensation, bulk reconciliation — before anyone is accused of anything. Only against a clean baseline does a genuine loss stand out.
A system that starts from measurement rather than suspicion is one crews will actually use instead of bypass. Shrinkage is usually not a people problem; it is an attribution problem, and attribution is something you can engineer.
Not usually. Of the common loss modes — unauthorised dispensing, ghost transactions, siphoning, meter drift and temperature variance — only one is outright theft. Most ‘shrinkage’ is a measurement-and-attribution failure: the meter counts volume but cannot say which vessel received it.
An NFC tag on each receiving vessel is read at the nozzle and held for the duration of the draw, so the dispensed volume is bound to a specific machine rather than just a time. That turns an unlabelled volume into an attributable transaction you can reconcile against the vehicle's consumption.
Fuel volume changes with temperature, so the same mass reads as different volumes hot and cold. Without temperature compensation to a reference volume, that physics shows up as apparent loss or gain — and real loss can hide inside it. Compensated metering removes that noise.
A dispensing event recorded with no corresponding delivery, or a delivery that happened with no record at all. Both create a gap between what the system thinks happened and what physically moved, and both are invisible to a meter that only counts total volume.
Yes. Mobile fuel dispensers are built for mines, farms, construction sites and depots that often lack mains power and reliable connectivity, with store-and-forward records that sync when a link is available and a sunlight-readable, glove-operable interface.
Because meter calibration error, temperature effects and bulk-reconciliation gaps can all look like loss. Establishing a trustworthy baseline first means a genuine loss stands out clearly, and the system is one operators trust and use rather than dismiss and bypass.
NFC identity at the nozzle, temperature-compensated metering and reconciliation that works from off-grid sites — built to start from measurement, not suspicion. Tell us where the fuel goes and where it disappears.