Fuel is the largest single cost line in most heavy-fleet operations and the most weakly evidenced at the moment it is spent. The gap between what the bowser dispensed and what the engine consumed is real, and it hides for about three weeks. Structured capture at the dispense closes it at the event, where it is cheapest to close.
Fuel is the largest single cost line in most heavy-fleet operations, and the most weakly evidenced at the moment it is spent. The gap between what the bowser dispensed and what the engine consumed is real on almost every site. The problem is not that the gap exists. The problem is that it hides for about three weeks.
The reason it hides is structural, not careless. Fuel moves through three handovers: dispensed at the bowser, consumed by the machine, reconciled at head office. Each handover is a place the number can drift. The reconciliation that would catch the drift runs weeks behind the dispense, so by the time the variance is visible the pattern has already run for a full cycle.
Where the three weeks go
The bowser log lives on paper or in a system that does not talk to the fleet record. Shift consumption is estimated rather than measured. Head office reconciles at month-end. The variance that a structured record would surface on day one surfaces, in the conventional setup, around week three. By then the underperforming truck, the heavy-handed operator, or the leaking dispense has run unnamed for a month.
It gets worse when the operator is absent at the dispense event. On a contractor site where the mine stores and dispenses, the storage point's number becomes the default the moment your operator is on another machine. One of your largest cost lines is then recorded by the party with the least incentive to record it in your favour.
The five percent that is not a rounding error
Every fleet operator suspects a gap between the bowser and the tank. Most cannot prove it. Five percent recovery on a USD 100,000 monthly diesel spend is USD 5,000 back to margin every month, on one line item. (Illustrative model, your number from the audit.)
The variance is rarely one cause. Truck-to-truck on the same haul road. Operator-to-operator on the same machine. Bowser-to-engine on the dispense itself. Without per-event attribution, all three average into a single fleet number and none of them is named. The fleet number tells you there is a problem. It does not tell you where.
The shift
Structured capture turns the bowser from the weakest evidence line in your record into the strongest.
What closes the gap at the event
- Every dispense captured against operator, asset, quantity and time at the moment it happens, not reconstructed at month-end.
- Operator PIN at the dispense, even when it is the standby operator, so the event is attributed to a person rather than a tablet.
- A photo of the pump reading and GPS confirmation on the dispense, so the claim is evidenced rather than asserted.
- Variance over your threshold opens an issue automatically, the day it appears, not three weeks later in a spreadsheet.
The same gap, in every vertical
Mining contractors carry it on day-works fleets where fuel is supplied against deduction. Construction main contractors carry it across project sites where one bulk tank feeds dozens of machines. Plant hire firms carry it as a deduction dispute, because the hire agreement bills the machine and supplies the fuel, and the deduction at month-end is only as strong as the dispense record behind it. Civils contractors carry it on remote road sections where the bowser follows the works. The cost line is the same shape across all of them, and so is the fix.
Why this matters more in Ghana specifically
Bowser stations sit at remote pit edges and stripping benches where 4G drops. Fuel is a high-value target. On contractor sites the storage is controlled by the mine. The capture has to survive all of that. Every fuel event saves locally first and syncs when signal returns, so the operator never waits for connectivity to evidence a dispense, and a failed sync becomes a visible issue rather than a silent gap in the record.
What changes by the first week
Fuel variance becomes visible at source from the first week of structured capture, not the third. The first full month usually surfaces a pattern the general manager suspected but could not evidence: a particular operator, a particular shift, a particular machine running heavy on the same loop everyone else runs clean. The pattern is the value. The recovery follows from naming it.
What this is not
This is not a fraud-detection pitch and not an accusation. Fuel variance has many causes, and most of them are operational rather than dishonest: haul-road condition, idle minutes, throttle discipline, a worn injector. The point is evidence, not blame. A structured record names the pattern. What you do with the pattern is the operational decision.
What the audit produces
The free 30-minute Operational Audit maps where your fuel evidence breaks. It names the dispense points and the cycles where the variance is hiding, specific to your fleet and your contract structure. You keep the one-page map regardless of next steps.
Read the operating view for your role
