Drone Pipeline Inspection: Visual, Thermal, and Methane Workflows

On This Page

• Match the sensor to the inspection question
• Visual inspection: locate and describe visible conditions
• Thermal inspection: interpret surface patterns
• Methane inspection: separate detection from quantification
• Specify the job from route inputs to field follow-up
• Require deliverables the asset team can use
• Questions to settle before awarding the work

Match the sensor to the inspection question

The following comparison separates the measured signal from the conclusion a buyer might want. The suggested handoffs are procurement recommendations, not a universal inspection standard.

Source basis: PHMSA's in-line inspection explanation, FLIR's thermal-camera guidance, FLIR's optical gas imaging explanation, Pergam's laser-sensing description, and Morales and colleagues' controlled-release research. Conditions and limits are discussed below.

These services can complement each other, but buying all the sensors does not automatically answer all the questions. The workflow must connect each observation to the appropriate follow-up.

Visual inspection: locate and describe visible conditions

Split the visual scope into corridor patrol and component inspection. A corridor patrol searches for changes along the route. Close component work needs views that identify the particular valve, support, exposed pipe section, or other item under review.

SPH Engineering describes pipeline corridor applications that include encroachment, third-party activity, erosion, and exposed pipe sections. Those are useful categories for a patrol brief; its application description is not proof that a particular flight will detect every instance. SPH Engineering's inspection overview also emphasizes connecting mission planning to reportable data.

For procurement, define a located observation rather than asking for an unspecified photo count. An image should let the reviewer find the asset, understand the observation, and see enough surrounding context to interpret it. Where the underside of a support is required, a top-down corridor pass should not be accepted as completion of that view.

Ask the contractor to distinguish visible surface discoloration, suspected coating damage, and an engineering diagnosis. A photograph of a discolored area does not measure lost metal. PHMSA describes separate in-line tools for metal loss, wall thickness, geometry, and crack-like defects. That distinction explains why an aerial visual survey cannot replace the particular integrity assessment needed for those questions. PHMSA's smart-pig fact sheet describes those measurement mechanisms.

A practical visual finding should say what is visible and what remains unresolved. For example, a hypothetical entry might describe an exposed pipe section beside an eroded bank, link its photographs, and request an operator assessment. It should not infer burial depth elsewhere or declare the pipe structurally safe.

Thermal inspection: interpret surface patterns

A thermal camera measures infrared radiation from the viewed scene and can derive surface-temperature information. FLIR identifies distance, ambient temperature, and material emissivity among the factors affecting measurement accuracy. Emissivity describes how effectively a surface emits thermal radiation. The camera observes the visible outer surface, which may be cladding or ground rather than the pipe wall. FLIR's thermal-camera guidance establishes that surface-measurement boundary.

That boundary matters when interpreting an apparent hot or cold area. Require the analyst to identify the surface being measured and explain the comparison. A temperature difference may justify closer examination, but its cause needs supporting operating information or another inspection method.

Record the relevant operating state supplied by the asset owner, the viewing distance and angle, weather, and the surfaces used for comparison. For quantitative work, retain emissivity and reflected-temperature settings. FLIR's GF7x manual explains that reflected radiation from surrounding objects affects interpretation, especially at low emissivity. It also distinguishes radiometric files, which retain thermal measurement information, from ordinary viewing formats and modes with reduced information. These are file and measurement principles, not a statement that this particular camera is a drone payload. FLIR GF7x manual, measurement parameters and image files.

Specify original thermal files plus readable report images when temperature analysis is part of the contract. Ask the recipient to open a sample before mobilization. A screenshot may illustrate a finding without preserving the information needed to revisit the measurement.

Ordinary thermography is not methane identification. Optical gas imaging uses spectral sensitivity matched to gases that absorb infrared radiation in the camera's band. FLIR also explains that a gas cloud needs radiant contrast with its background to be visible. A general thermal image does not acquire that capability just because it uses a colored palette. FLIR's OGI explanation.

Methane inspection: separate detection from quantification

“Methane survey” can describe several different outputs. Establish which one is being purchased: screening for elevated signals, locating a probable source, confirming a component leak, or estimating an emission rate.

Remote laser sensing and air sampling measure different things

A remote laser instrument measures absorption along an optical path. Pergam's description of drone methane sensing reports the result in ppm·m, parts per million multiplied by metres. This is a concentration-path-length quantity, not a mass flow rate. Ask for the instrument's exact units, background treatment, valid-signal criteria, and relationship between the recorded position and the sampled path. The historical Pergam technical description supports that distinction; its old product availability statements are not current purchasing information.

An air-sampling instrument instead measures the air reaching its inlet, either directly or through a sampling system. The integration review should address inlet location, response time, and synchronization with aircraft position. Morales and colleagues examine sampling delays and clock alignment, which affect where a measured peak belongs in a survey.

Optical gas imaging needs suitable viewing conditions

An OGI scope should identify the gas sensitivity, viewing conditions, and record required for each observation. EPA's December 2023 Appendix K fact sheet describes an operating envelope using wind speed, gas-to-background temperature difference, and viewing distance. It also addresses camera checks, operator training, monitoring plans, and records. EPA's Appendix K technical fact sheet.

That document concerns specified OGI use at natural gas processing plants. It is not blanket approval of a drone method for every pipeline survey. If the work must fulfill a regulatory obligation, require the proposal to identify the applicable rule, facility scope, accepted method, and any approval conditions.

A leak rate requires an additional measurement method

In a controlled-release study, Morales and colleagues estimated methane emissions by sampling a vertical plane downwind of a source and combining methane measurements with wind information. They examined clock alignment, sampling delays, plume coverage, and alternative processing approaches. Missing part of a plume could undermine the estimate. Their results concern that experimental method and conditions, not a universal drone accuracy specification. The 2022 research paper.

For a rate reported in kg/h, request the method, measurement interval, background estimate, wind data, uncertainty, and grounds for accepting or rejecting the result. A map of elevated readings is useful screening information even when it cannot support a rate. Label it accordingly.

Keep non-detection equally precise. Report the surveyed extent, time interval, method limits, and any gaps alongside “no methane detected.” Otherwise, an unsampled or invalid interval can be mistaken for a satisfactory result.

Specify the job from route inputs to field follow-up

The following sequence is a recommended commercial handoff built around the sensor limits above.

1. Define the assets and decisions. Supply the operator's route geometry, segment identifiers, stations and components, transported product, buried or exposed status, previous findings, and the reason for the inspection. Name the person who will accept the report and assign follow-up work.
2. Agree on coverage and outputs. Separate corridor imagery, component views, thermal assessment, methane screening, localization, and quantification into explicit work items. Define what qualifies as usable coverage for each one and how exclusions will be reported.
3. Resolve access and flight feasibility. Identify launch locations, site restrictions, obstacles, crew access, and the operational authority for the route. For U.S. Part 107 operations, FAA lists flight beyond visual line of sight under the provisions requiring a waiver. A long-range aircraft specification is not that authorization. FAA Part 107 waiver guidance.
4. Check a representative sample. Before full collection, have the recipient review sample photographs, a thermal original if relevant, and a sample gas-data export. Confirm that asset identifiers, units, timestamps, coordinates, and files survive transfer into the recipient's tools.
5. Collect and record exceptions. Log the actual routes and capture conditions. Keep obstructed views, invalid gas readings, and deferred segments separate from usable observations. Agree beforehand who can authorize a return visit and how recollection is charged.
6. Review and hand off findings. Assign each finding a stable identifier, its supporting records, the analyst's interpretation, and the proposed next action. Agree on the immediate notification route for urgent observations so they do not wait for the final report.
7. Close the follow-up loop. Link field confirmation, repair, or a decision to investigate further back to the original finding. Where repeat measurement is required, preserve its date, method, and conditions rather than treating a repair entry as measurement evidence.

This separation lets a buyer compare bids on the same task. A corridor screening proposal and a component-level leak investigation may both be useful, but they should have different completion criteria.

Require deliverables the asset team can use

Ask for a sample handoff organized around asset and finding identifiers. A folder of images can support the record; it should not be the only way to locate a reported problem.

This table proposes contract deliverables; it does not prescribe a regulatory reporting format.

Measure completion against the original agreed scope. A useful proposed metric is:

Usable corridor coverage (%) = accepted unique corridor length ÷ originally scoped corridor length × 100.

The units must match, overlapping passes count only once, and “accepted” must refer to the agreed sensor task. Report component-view completion separately. Neither measure is a defect-detection probability or proof of pipeline integrity.

For example, a segment can be visually complete while its methane survey remains deferred because the agreed measurement conditions were not met. Keeping those two states visible tells the asset manager exactly what still needs work. The distinction between observation, measurement, and diagnosis is explored further in what drone-inspection evidence can prove.

Questions to settle before awarding the work

Ask the provider to answer these questions against the actual route and deliverables:

• Which findings can the proposed method support, and which require another inspection?
• What are the exact methane output units, and does the quote include screening, localization, confirmation, or quantification?
• What conditions make a measurement unusable, and how will those gaps appear in the report?
• Which original files and metadata will the asset owner receive, and can its software read them?
• What crew arrangement and flight authority support the proposed corridor coverage?
• Who reviews findings, receives urgent notifications, performs confirmation, and records closure?
• What work triggers another mobilization, and who pays for it?

Award the scope that answers the operational question with usable records and a clear follow-up owner. Use visual capture for visible condition and context, thermal work for a defined temperature question, and methane sensing for a defined gas question. Commission emission-rate measurement explicitly when a rate is needed.

Source notes

• PHMSA: In-Line Inspections (Smart Pig). Government technical explanation of tools for pipe-wall and geometry assessment.
• FLIR: Using a thermal camera. Manufacturer guidance on surface-temperature measurement and influencing factors.
• FLIR GF7x series manual. Primary documentation for measurement parameters and radiometric file treatment; no drone compatibility inferred.
• FLIR: The Science Behind Optical Gas Imaging. Manufacturer explanation of spectral sensitivity and gas-to-background contrast.
• Pergam: Methane detection from drones. Historical manufacturer explanation of laser methane readings and their units.
• Morales et al., 2022: Controlled-release experiment to investigate uncertainties in UAV-based emission quantification for methane point sources. Primary research on sampling and quantification uncertainty in a specific experimental system.
• EPA: Appendix K optical gas imaging technical fact sheet, December 2023. Government explanation of OGI operating conditions and the specified processing-plant scope.
• SPH Engineering: Drone Inspections. Provider account of corridor inspection applications and planning; no performance guarantees adopted.
• FAA: Part 107 Waivers. Current U.S. agency guidance on operations requiring a Part 107 waiver.

Last checked: September 6, 2026.