The Economics of Fuel Theft in Commercial Fleets
A breakdown of where fuel theft costs accumulate, why it is underreported in fleet P&L statements, and how operators can start measuring the real number.
The standard way to think about fuel theft is as an incident: a specific vehicle, a specific event, a specific volume extracted. This framing makes the problem seem discrete and manageable. The incident happened, or it did not. You caught it, or you did not.
The more accurate economic framing is that fuel theft in a commercial fleet without detection capability is a tax. It accrues continuously, across multiple vehicles and routes, and it hides inside the normal variance of fuel spend rather than appearing as a separable line item. The reason it hides there is structural: without per-vehicle telemetry data, there is no method to distinguish consumption that went through the engine from consumption that went through a siphon hose. Both show up as fuel cost.
Where the cost actually accumulates
Fleet operators often underestimate the per-month exposure because they think about single events. A 20-liter siphoning event on a single truck is perhaps 400 to 600 Egyptian pounds at current diesel prices. That number sounds manageable. The economic case looks different when you consider the frequency and distribution rather than the per-event volume.
In commercial fleets operating without monitoring, industry estimates for fleet fuel loss to theft and pilferage run in the range of 3 to 5 percent of total fuel spend annually. That is a widely-cited range drawn from commercial fleet insurer data and telematics deployment observations across multiple markets, not a precise measurement. The actual figure for any specific fleet depends on its routes, depot security, workforce characteristics, and vehicle mix. But it is a useful order-of-magnitude reference.
For a fleet of 25 trucks consuming an average of 300 liters per vehicle per week, total fleet consumption runs roughly 390,000 liters per year. At 3 percent loss, that is approximately 11,700 liters. At 5 percent, it is 19,500 liters. At current Egyptian diesel retail pricing, the annual exposure range for a fleet this size sits in the range of 230,000 to 390,000 Egyptian pounds. That is before accounting for any overhead costs of the investigation, reconciliation, and administrative work that theft events generate.
We are not claiming these numbers apply to any specific fleet, and we are not fabricating precise statistics. They are illustrative of the order of magnitude that the 3 to 5 percent loss rate implies for a fleet of typical size. Any specific fleet's actual loss rate requires measurement, which requires telemetry data.
Why it is underreported in P&L statements
The fundamental reason fuel theft does not appear as a distinct cost in most fleet P&L statements is that the accounting category for it does not exist. Fleet operators track total fuel spend. They do not track fuel consumed by the engine separately from fuel extracted from the tank. Without that separation, theft sits inside the "fuel cost" bucket alongside legitimate consumption.
When fuel spend increases year-over-year, fleet managers typically attribute it to one or more of: higher fuel prices, increased utilisation, vehicle aging or declining efficiency, or route changes. All of those can be true and are reasonable explanations. But they are also convenient explanations for a cost increase that includes a theft component that no one has the data to separate out.
This creates an organizational problem beyond the financial one. If theft is being attributed to aging vehicles or route efficiency, the responses are wrong. A maintenance program or route optimization exercise does not address theft. The theft continues while the organization focuses on the wrong lever.
Telemetry data creates the separation that makes fuel theft visible as its own cost category. When you can compare fuel consumed by the engine (derivable from OBD load and speed data) against fuel removed from the tank (measured by the level sensor), the unexplained difference becomes a distinct measurement rather than an invisible component of total cost. That measurement may not be precise, because both OBD consumption estimation and fuel level sensing have error margins, but it is directionally useful and substantially better than nothing.
The compounding effect across vehicles and time
Fuel theft is not randomly distributed across a fleet. It tends to concentrate on specific vehicles, routes, and shift windows. A vehicle that is accessible during a predictable overnight window at a low-supervision depot, or a route that passes through an isolated location with a regular driver pattern, creates repeated opportunity for the same extraction method to be applied multiple times.
Without detection, this pattern can persist for months or years. The cost compounds not because the per-event volume is large but because the recurrence is high. Two or three events per month on three vehicles over 12 months is 72 to 108 events, each removing 15 to 30 liters. The cumulative exposure from a single exploited pattern over a year can approach or exceed the total cost of deploying a monitoring system across the entire fleet.
Detection interrupts this compounding. Once a pattern is identified and the vehicle, route, or shift implicated is under investigation, the pattern typically stops. The value of detection is therefore not only in recovering losses from identified events, but in terminating the forward projection of recurring losses that would otherwise continue.
The documentation and insurance dimension
Some fleet operators carry commercial insurance products that cover fuel loss as part of broader goods-in-transit or fleet asset coverage. Filing a claim against such a policy requires documentation: evidence of the event, time and vehicle identification, and an estimated volume. A telemetry-derived alert log provides machine-generated evidence with timestamps, vehicle IDs, and volume estimates derived from sensor data. This is substantially more defensible documentation than a driver's verbal account or a reconciliation note in the monthly spreadsheet.
Even for fleets without specific fuel theft coverage, the documentation value applies internally. An audit trail of anomalous fuel events, categorized by vehicle and route, provides the basis for an internal investigation that goes beyond suspicion. The data does not replace judgment. But it gives management something to work from that is more specific than "we think there might be a problem on the night shift."
Starting to measure the real number
The first step toward accurate measurement is per-vehicle fuel level data. Without it, the only available number is the aggregate monthly spend, which combines everything. With it, you can begin to construct a per-vehicle consumption history that separates normal operating patterns from anomalous events.
The second step is time. The consumption baseline takes several weeks to establish. In the first two to three weeks of telemetry, the model is still learning normal behavior for each vehicle. By week four or five, the anomaly detection starts producing alerts with meaningful confidence scores. By month two or three, the historical pattern is clear enough to start making a rough estimate of what the ongoing loss rate actually is for that specific fleet.
That number, measured rather than estimated from industry averages, is the starting point for any rational decision about fleet monitoring investment, route security, and driver policy. Without it, you are managing a cost you cannot see. With it, you are managing a cost you understand.