Storage tank and terminal drone inspection: replacing scaffolding, ladders, and confined-space entry

Tank farms and terminals adopted drone inspection faster than almost any other downstream environment, for a simple reason: tank inspection stacks height, confined space, and hydrocarbon atmosphere into a single job. Every part of that work a drone can absorb removes a disproportionate amount of risk and cost.

External inspection: shells, roofs, and appurtenances#

External tank survey by drone captures systematic imagery of the shell (coating breakdown, corrosion, staining, settlement indicators), the roof (a particular win on floating-roof tanks, where seal and pontoon condition are critical and access is awkward), wind girders, stairways, and nozzles. Thermal imaging adds a further layer: insulation defects, wet insulation - a driver of corrosion under insulation - and product interface levels can all present as thermal signatures visible in a single scanning pass. Work that previously required scaffolding or mobile elevated platforms compresses into flights measured in minutes per tank.

Internal inspection: the confined-space case#

The stronger safety case sits inside the tank. Confined-space entry into a storage tank requires gas-freeing, forced ventilation, atmosphere monitoring, standby rescue cover, and still leaves entrants exposed to residual atmosphere and fall hazards from internal structure. Collision-tolerant indoor drones - caged airframes designed to touch surfaces and keep flying - now perform internal visual surveys of shell courses, roof support structure, and floors in emptied tanks, meaning the evidence is collected before any person enters, and in some cases instead of a person entering at all. Each entry eliminated is an entire category of exposure removed.

Where the standards sit#

Recognised tank inspection frameworks - API 653 being the most widely referenced for atmospheric storage tanks - define what must be assessed and by whom. Drone-collected evidence increasingly feeds these assessments, but it changes the collection method, not the accountability: a certified inspector still interprets the evidence and signs the assessment. This mirrors the pattern across every inspection domain we cover - the technology gathers and organises evidence; the qualified human makes and owns the judgement, a division of labour explored further in what AI actually means in inspection software.

The cross-cycle comparison problem#

A terminal campaign produces thousands of images across dozens of visually near-identical assets, and the question that matters for integrity planning is almost never what one image shows - it is whether the corrosion patch on tank 14, course 3, north-northwest orientation has grown since last year. Answering that reliably requires evidence that is organised, positionally consistent, and genuinely comparable across cycles - the same comparable-evidence problem that defines pipeline corridor inspection, concentrated into a tank farm. Programs that treat each campaign as a fresh pile of photographs re-derive their asset knowledge every cycle; programs that keep evidence comparable compound it.

Related reading#

For the sector-wide picture, see how drones are transforming oil and gas operations; for the corrosion mechanism that thermal tank scanning most often reveals, see corrosion under insulation explained.