An invalid fill-up is any fuel transaction that, for whatever reason, does not fit the logic of the vehicle or the moment. It may be human error, a mechanical problem, a measurement failure or an attempted fraud. What matters is not assuming bad intent: what matters is detecting it in time, reviewing it and learning from it.
In fleets without automatic control, invalid fill-ups go unnoticed and pile up as an invisible cost. In fleets with automatic control, every fill-up is validated against four rules before it is marked NORMAL and, if it fails any of them, it is flagged REVIEW so a person can audit it.
Invalid fill-ups are the first line of fuel control. If your system does not detect them, you are not controlling fuel: you are recording expenses.
A fill-up is considered invalid when it breaks at least one of four operating rules:
1. The number of liters exceeds tank capacity. If the model has an 80-liter tank and the fill-up recorded 95, there is a problem. It may be a typo, an incorrectly entered invoice, or an attempt to inflate the fill-up. It is the simplest rule and the one that surprises most when it is switched on: many fleets have this type of error without knowing it, simply because nobody cross-checks tank capacity against liters pumped.
2. Implied fuel efficiency is outside the expected range. If between the last fill-up and this one the vehicle covered 200 km and consumed 50 liters (4 km/liter against an expected 12), something does not add up. There may be an odometer error, a mechanical failure, a false fill-up or an operational use different from what was planned.
3. The price per liter is outside the historical range. If the fill-up’s price per liter is 30% above the previous month’s average in the same area, it is worth reviewing. It may be a real price increase from the provider, a station that charges more, or incorrectly entered data.
4. The fill-up location is inconsistent with the vehicle’s GPS. If the card was used at a station in zone X but GPS shows the vehicle spent the whole day in zone Y, there is a problem. It may be card misuse, a driver fueling a personal vehicle, or a disconnected GPS.
Every fill-up goes through all four validations. If they all pass, it is marked NORMAL. If even one fails, it is marked REVIEW.
Manual detection of invalid fill-ups does not work. A 50-vehicle fleet generates between 400 and 800 fill-ups a month. Reviewing each one by hand against four rules takes between 8 and 15 hours a month of review alone, not counting the investigation of flagged ones. No team does it consistently.
Automatic detection works as transactions come in: when a fill-up enters the system (via the fuel card integration), the four validations run in seconds. NORMAL fill-ups accumulate with no manual work. REVIEW fill-ups appear in a dedicated queue that the person in charge goes through once a day in 5-10 minutes.
The difference between manual and automatic detection is not marginal: it is the difference between having control and not having it.
When a fill-up is flagged REVIEW, the right workflow has 4 steps:
Step 1: Identify which rule failed. The system shows which of the four validations triggered the alert. Exceeding tank capacity is not the same as efficiency being out of range. Each one is investigated differently.
Step 2: Gather evidence. For rule 1 (capacity): check the provider’s invoice against the liters actually pumped. For rule 2 (efficiency): check the vehicle’s odometer and talk to the driver. For rule 3 (price): compare against the station receipt. For rule 4 (location): cross-check with GPS and the driver’s location at that moment.
Step 3: Classify the cause. Causes fall into four categories: data-entry error (the driver wrote down the wrong odometer reading, for example), mechanical problem (the vehicle really is consuming more), provider error (the station invoiced incorrectly), or attempted fraud. Classification matters because it defines the next action.
Step 4: Take action. Data-entry error: correct the data and train the driver. Mechanical problem: open a Corrective ticket. Provider error: file a claim with the provider and adjust the amount in the system. Fraud: apply the company’s internal policy (which may include a deduction, a sanction or dismissal).
The entire workflow is recorded in the system. The invalid fill-up moves from REVIEW to a final status (Corrected, Justified, Confirmed fraud) with the evidence attached.
A single invalid fill-up is an event. But when a pattern repeats, it usually points to a systemic problem that needs attention:
The monthly invalid fill-up report has to include these five patterns, not just the absolute number. They are the basis for detecting the hardest-to-see fraud warning signs.
To understand how the invalid fill-up queue is processed in a real operation, look at the 12 cases that came up in one month in a 55-vehicle fleet. The final breakdown by cause was: 5 correctable data-entry errors, 3 genuine mechanical problems, 2 provider errors and 2 confirmed fraud cases.
Data-entry error cases (5). Four were odometer readings written down incorrectly by the driver (187,450 km instead of 178,450 km, for example) and were corrected using the shop invoice that had the correct figure from the last service. One was a fill-up entered against the wrong vehicle in the provider’s system, corrected by reassignment.
Mechanical problem cases (3). Two vehicles showed implied fuel efficiency 22% and 28% below expected over 3 consecutive fill-ups. The mechanical inspection found clogged air filters in one (immediate replacement, back to normal range on the next fill-up) and an injector problem in the other that required a major Corrective ticket. The third case was a pickup with underinflated tires: an operational fix, with no need for a shop visit.
Provider error cases (2). One was a station that invoiced a price 35% above the historical average, detected by the system’s rule 3. Formal claim with the provider, adjusted afterwards. The other was a fill-up duplicated by a processing error at the card provider, identified because it appeared twice with the same timestamp.
Confirmed fraud cases (2). The first was a driver who pumped an extra 5-8 liters on every fill-up for 3 months and sold the surplus. The pattern was detected by the combination of fill-ups exceeding the model’s tank capacity (rule 1) plus low implied efficiency (rule 2). Traceability with timestamps and station video confirmed the case. The second was use of a corporate card on a personal vehicle during weekends, detected by a location inconsistency with GPS (rule 4) and off-shift timing (complementary sign).
Total time the person in charge spent managing the month: 4 hours and 20 minutes for the 12 cases. Documented monetary recovery from the 2 fraud cases and the 2 provider errors: USD 1,840. Without the automatic detection system, none of the 12 cases would have been visible.
VEC Fleet implements automatic detection of invalid fill-ups as a native part of the Fuel module:
How many invalid fill-ups does your fleet have this month?
With VEC Fleet, every fill-up goes through automatic validations, and those that fail are flagged for review, with full traceability of which control failed.
