Diesel and fuel theft: how it actually happens, and what stops it

Why fuel is the theft that keeps happening

Copper gets the attention and equipment gets the insurance claim, but on most active sites the loss that repeats month after month is fuel. It repeats because it has the best risk profile of any theft available on a jobsite: the target is outdoors and unsecured by design, it is worth taking in quantity, it is instantly resaleable or immediately usable, and unlike a skid steer it carries no serial number, no telematics and nothing for anyone to look for afterwards.

It also repeats because it rarely gets reported. A missing excavator produces a police report, an insurance claim and a conversation with the owner. Four hundred gallons missing across six machines produces a superintendent who assumes the fuel card was misused, a fuel vendor who gets asked to check their delivery tickets, and a schedule that slips by half a day while someone drives a service truck around. The event is absorbed into operating friction rather than recorded as a crime, which is exactly why it is worth doing repeatedly.

The practical consequence is that the same site often gets hit three or four times before anyone treats it as a pattern. By then the crew doing it knows the approach, knows which machines sit furthest from the gate, and knows roughly what night of the week the yard is quietest. That accumulated familiarity is the real reason losses accelerate rather than stay flat.

Three different thefts that all get called fuel theft

Machine tank draining is the common one. Equipment parks overnight with a working day of fuel aboard, and a crew with a portable transfer pump and a few drums works down a row of machines. It is fast and it is quiet, because a twelve-volt transfer pump makes less noise than the traffic on the nearest road. The tell the next morning is not an empty tank but a fuel gauge that does not match the hour meter, which is why it is so often mistaken for a gauge fault or a fuel-card problem.

Bulk tank theft is the larger single event. A skid tank or a bowser on site is a concentrated target, and a crew that arrives with a vehicle-mounted tank is not taking a few gallons for a truck, they are emptying it. This is the version most likely to be a repeat visit from the same people, because a site with a bulk tank has effectively advertised that it holds fuel in quantity.

Then there is internal loss, which is a different problem wearing the same costume. Fuel cards used for personal vehicles, cans filled at the end of a shift, deliveries that get signed for short. It is genuinely common, it is usually smaller per event, and it matters here because it changes what you should be looking at. If the pattern is weekday and daytime, surveillance of the fuel point is answering an accountability question rather than an intrusion question, and the evidence you need is a record of who was at the tank and when rather than an alert at two in the morning.

• Machine tanks — many small takes, noticed late, usually mistaken for a gauge or card issue
• Bulk or skid tanks — one large take, high repeat risk once the site is known to hold fuel
• Internal and card loss — weekday pattern, solved by a record rather than an alarm
• Mixed patterns are common on larger sites, and they need different responses

Why the usual defences do not hold

Locking fuel caps are worth fitting and will not solve this. They defeat an opportunist with a hose and a jerry can, which is a real category of thief, and they are irrelevant to anyone carrying a pump and a pair of bolt cutters. The same is true of anti-siphon screens: they are designed to stop a tube being fed down the filler neck, and a crew that has already decided to open the tank is not attempting a siphon.

Fencing performs the same way. A fence establishes that entry was deliberate, which matters legally and matters to a carrier, and it delays a crew by the time it takes to cut or drive through it. What it cannot do is tell anybody that the cut has happened. On a fuel theft specifically that gap is decisive, because the whole event fits comfortably inside the window between the fence being breached and the first person arriving the following morning.

Lighting is the defence most often over-credited. Bright light genuinely reduces opportunistic crime, and it does so by increasing the chance of being seen by a person. On an empty jobsite at three in the morning there is no person, so the lighting is illuminating the work rather than deterring it. Several of the fuel thefts we have reviewed happened on well-lit sites, and the lighting improved the quality of the footage more than it changed the behaviour.

Telematics deserve a specific mention because they are often assumed to cover this. Machine telematics will show you a fuel level drop, which is valuable, but they show it as a data point after the fact and typically on a reporting interval rather than in real time. Telematics tell you that you were robbed. They do not interrupt it.

What detection has to do to catch a fuel theft

Fuel theft has an unusually short and unusually specific signature, and that shapes what a camera system has to do. The event is short — a machine tank is a matter of minutes, a bulk tank a matter of tens of minutes — so detection has to be immediate rather than reviewed. The approach is on foot from a vehicle parked short of the target, so the vehicle and the person often arrive as separate events. And the activity happens at ground level beside a machine, which is a harder camera problem than someone walking across open ground.

That argues for zone placement that most sites get wrong. The instinct is to point cameras at the perimeter, because that is where entry happens. For fuel specifically the higher-value zone is the fuel point itself and the row where equipment parks, because that is where the person has to stand, has to stay for several minutes, and cannot do the job from cover. A detection zone drawn tightly around the bulk tank and the equipment row produces far fewer false alerts than a perimeter zone facing a public road, and it fires on the part of the event you can still interrupt.

Timing matters as much as geometry. Fuel theft concentrates in the hours when a site is genuinely empty, which on most DFW jobsites is between roughly ten at night and four in the morning. A schedule that arms the fuel zone outside working hours and stands it down during the day removes the single biggest source of nuisance alerts, which is your own crew fuelling machines, and it is the difference between a system people keep armed and a system somebody turns off in week three.

The last requirement is that something happens on site at the moment of detection. A notification tells a superintendent at home that fuel is being taken. A siren, a strobe and a recorded talk-down naming what the person is doing interrupts the take while it is still in progress, and on a theft measured in minutes that distinction is the whole outcome. This is the specific case where active deterrence earns its place rather than sounding like a feature list.

The setup we recommend, and where it is not enough

On a typical DFW construction site or contractor yard, a single unit placed to see both the equipment row and the fuel point covers this properly. The SiteFortress 360 at $995 per month is the common answer because its eighteen to twenty foot mast sees over parked machinery that would block a lower camera, and because four fixed cameras can hold the fuel point, the equipment row and the approach simultaneously rather than panning between them. Smaller yards where everything is close together are well served by the Mini at $695 per month.

Large yards are the case where the PTZ Pro at $1,695 per month is genuinely the right call rather than the expensive one. On several acres of gravel, fixed cameras will reliably tell you that a person is at the far tank; a zoom camera can establish whether it is a crew with drums or your own night operator, and that distinction determines whether anybody gets called. If your rows run a few hundred feet, optical reach is not a luxury.

Where this approach is weaker is internal loss. A camera on the fuel point creates an excellent record of who was there and when, and a record is the correct tool for a card or shrinkage problem, but it resolves the question after the fact rather than preventing it. Sites with a serious internal fuel problem usually need the record plus a process change — assigned fuelling, reconciliation against hour meters — and we would rather say that plainly than sell a trailer as the answer to a payroll problem.

It is also worth being clear that fuel storage on site carries its own safety and code obligations independent of theft. Bulk storage, spill containment and the separation of fuel from ignition sources are governed by fire code and site safety requirements, and a security decision should never quietly become the reason a tank ends up somewhere convenient for a camera rather than somewhere compliant.

Frequently asked questions