Storage Tank Shell Inspection: Build a Course-by-Course Visual and UT Map

Why are convenient shell thickness spots not enough?

A shell survey becomes useful when another crew can return to the same piece of steel. A low value beside a nozzle needs a course, plate, elevation and circumferential position so the crew can confirm it and expand the examination. A long list of values also says nothing about the shell areas the team never reached.

The problem appears during comparison. One survey uses drawing north and measures clockwise. The next starts at the stair and records clock positions from the opposite view. Both tables look orderly, but the rows do not describe matched locations. A calculated trend built on that mismatch can be more misleading than no trend at all.

Treat the shell as a mapped asset. An observation records what is visible at a known location. A measurement records a procedure-accepted UT result and its acquisition context. A mechanism hypothesis joins service, material, geometry, history and wider evidence. A disposition applies the owner’s adopted program and engineering judgment. Do not collapse those four acts into one spreadsheet cell.

API 653 places applicable aboveground steel storage tanks within an inspection, repair, alteration and reconstruction program, while API RP 575 supplies complementary inspection practice. Use the owner-adopted editions and addenda. Neither a robotic route nor a large reading count replaces that program.

Start with the tank, service and decision

Identify the tank before choosing measurement locations. Record the controlled tank ID, drawing revision, construction type, shell-course arrangement, plate records when available, material information, nominal thickness basis, current and past service, operating temperature, coating and insulation state, repairs, replacements and previous inspection history. Mark uncertain records as uncertain; do not let field assumptions become design data.

State the decision in plain language. The scope may need to screen broad external coating breakdown, revisit a known low area, examine shell exposed after insulation removal, or build a repeatable baseline. Each question needs a different distribution of evidence. “Take UT readings on the shell” gives the crew no basis for choosing locations.

What should the visual pass record before UT?

Start wide. Photograph the tank elevation with a permanent orientation reference, then move to each course, plate and feature. Record coating breakdown, rust staining, wetting, corrosion product, deposits, leakage evidence, insulation damage, weatherproofing openings, bulging or other visible distortion, and every physical obstruction. Tie each view to the same datum used for thickness data.

For flammable-liquid tanks within its scope, UK Health and Safety Executive guidance calls for competent-person examination and intermediate external examination of aboveground tanks. Use visible cues to target that work; surface appearance does not diagnose a mechanism or establish remaining wall.

Escalate geometry observations through the right route. Bulging, peaking, banding, settlement effects and loss of verticality require appropriate dimensional or structural evaluation. Ordinary thickness UT does not decide those questions.

Build the course-and-plate coordinate system

Declare how courses are numbered and where the circumferential datum begins. Tie zero to plant north or a permanent physical feature, state the viewing direction, and use one circumferential convention throughout: degrees, clock position or developed distance. Record the elevation from one defined reference. Never mix drawing north, true north, clockwise and counter-clockwise without a documented transformation.

Use original plate IDs where controlled records support them. When they do not, create a temporary plate or segment grid that another crew can reconstruct. Map every vertical and horizontal seam, nozzle, manway, reinforcement pad, insert plate and repair before placing results. A point described only as “course 3” still leaves too much shell unresolved.

Give photographs, attempted readings, accepted readings and obstacles the same coordinate language. A useful identity might read tank, course, plate, datum convention, circumferential position, elevation and distance from the nearest seam. Add the feature reference only as context, not as a substitute for coordinates.

How should the damage hypothesis shape coverage?

Begin with credible mechanisms, not a preferred travel line. Atmospheric exposure, coating breakdown and water-trapping details may direct attention externally. Service history may make broad or localized internal loss credible. Insulation condition may identify areas that need controlled exposure or another validated approach. Repairs, attachments, drains and previous findings can change the priority within a course.

A thickness value does not identify which side of the wall lost metal. It also does not assign a mechanism. Join valid thickness with the as-found visual record, service and material history, inspection history and any complementary examination before proposing cause.

Set the density from the decision, anticipated morphology, tank geometry, technique footprint and location uncertainty. There is no universal shell grid that fits every tank. A convenient vertical strip describes only the valid locations it covers unless the owner’s inspection basis supports a wider conclusion.

Where can MagRover fit on a tank shell?

MagRover is a magnetic wall-climbing platform for visual inspection and UT spot readings or scans on suitable ferrous tank shells. The owner's inspection plan sets the locations and evidence needed; MagRover offers one possible route for collecting that evidence.

Screen the actual tank and configured system. Review material, surface profile, coating, temperature, course geometry, welds, nozzles, manways, wind girders, stairs, clips, pipework, insulation, power and couplant arrangements, tether path, exclusion zones and recovery. A clear line on an elevation drawing may be interrupted by permanent features that are hard to see from grade.

Agree the site's work controls before deployment, including permits, hazardous-area equipment suitability where relevant, tether management, exclusion below the route and a recovery method. A successful route trial does not replace authorization to work on the tank.

Prove the route on representative shell features before crediting coverage. Magnetic adhesion, physical travel, stable probe seating, coupling, location control, acceptable UT response and controlled recovery are separate gates. Passing one does not pass the others.

What makes a shell UT result usable?

A displayed number is not automatically a thickness result. The approved procedure must address the material and sound-velocity basis, instrument and probe setup, calibration and verification, surface condition, coating technique when used, temperature, couplant, geometry, stable seating and the acceptable back-wall response. Qualified personnel apply the owner’s requirements to the configured technique.

ASTM E797/E797M-21 describes one-sided contact pulse-echo thickness measurement for suitable conditions and is primarily directed to flat, parallel surfaces. Its applicability is limited on non-parallel or concentric geometry. On a curved shell, demonstrate probe seating and a resolvable back-wall response with the approved procedure and configured probe.

Separate travel, attempt and acceptance. The route log can show where the platform moved. The attempt log can show where contact was made. Only a response that meets the procedure becomes valid UT. Keep unstable, unresolvable or inconsistent responses as rejected or no-response records at their planned coordinates.

Recheck a credible low or abrupt change before interpretation. Confirm the location, surface, coupling, setup and response, then repeat or expand as the approved plan directs. Never replace an invalid point with a nearby value while retaining the original coordinate.

How should coating, insulation and welds be handled?

Record the as-found surface before preparation. Coating type, thickness variation, roughness, adhesion, corrosion product and contamination can affect probe seating or the approved technique. A through-coating reading belongs in the map only when the project procedure supports that coating and configured system. Validate any coating correction for the actual setup.

Insulation hides the shell. Damaged weatherproofing, wet insulation or staining can support a corrosion-under-insulation concern, but an unexposed shell remains unexamined by ordinary external contact UT. The owner controls insulation removal, reinstatement and any complementary method. Do not interpolate through an insulated band.

Treat horizontal and vertical weld profiles as both map features and technique constraints. Weld reinforcement or nearby geometry can interrupt travel, seating or response. Test each crossing with the configured platform and UT setup. If travel or response fails, mark the gap and assign another access or examination method.

Map every obstacle and coverage gap

Draw the obstacles before the route begins. Show nozzles, manways, reinforcing pads, stairs, clips, wind girders, connected pipes, grounding attachments, repairs, insulation limits and inaccessible upper or lower bands. Give each one a location and a likely effect on visual access, platform travel or acoustic contact.

Use explicit states. Planned means the location belongs to the scope. Visual only means a usable view exists but no valid thickness was obtained. Valid UT means the response met the procedure. Rejected or no response preserves an attempted location. Inaccessible records a physical boundary. Complementary follow-up assigns the unresolved question to another method or access route.

Calculate achieved coverage only from the evidence state defined in the inspection basis. Platform travel, camera view and valid UT answer different questions and should not share one undifferentiated percentage.

How should credible low readings be expanded and confirmed?

First confirm that the low reading belongs to the stated metal. Recheck the coordinate, instrument and probe setup, verification status, surface, coating technique, couplant, seating and back-wall response. Repeat the acquisition as the procedure directs. A single displayed minimum without response quality or location control is not a defensible map.

Expand around a confirmed result according to the approved inspection plan. The pattern may need circumferential and vertical extension, closer spacing or a different method where localized loss is credible. The objective is a reviewable boundary for the owner’s decision, not a hunt for the lowest pixel on an unlabeled image.

Keep interpretation separate from disposition. The expanded map may support a mechanism review and an engineering calculation, but it does not itself decide minimum allowable thickness, remaining life, repair or fitness for service.

When can shell readings be compared with history?

Compare identity before arithmetic. Confirm the same tank, course, plate, circumferential datum, elevation and local reference. Then compare material assumptions, technique, instrument and probe configuration, velocity basis, coating treatment, surface preparation, temperature and retained response evidence.

Review what changed between surveys. A replaced plate creates a new baseline. Grinding, blasting or recoating changes the surface. A changed service, repair or insulation arrangement changes the damage context. A migrated datum or an approximate old location may prevent point-to-point comparison even when the numbers look compatible.

When the match is defensible, keep the trace from each value back to its source record. When it is not, state the break and start a controlled baseline. Do not manufacture a corrosion rate from unlike points to avoid an empty field.

What belongs in the final shell record?

Build the deliverable so an independent reviewer can reconstruct the work without calling the field crew. Include the controlled asset and drawing identity, coordinate convention, shell-course and plate map, planned and achieved routes, as-found visual register, valid UT results, rejected responses, inaccessible areas, preparation history, setup and verification records, exceptions and follow-up ownership.

Keep the four claim classes in separate fields. Observation describes visible condition. Measurement records accepted data and context. Mechanism hypothesis states the technical explanation under review. Disposition records what the authorized owner roles decide under the adopted inspection, design and assessment basis.

For U.S. processes covered by OSHA’s process-safety-management rule, the mechanical-integrity provisions address written procedures, inspection and testing, documentation and correction of deficiencies. That rule applies only to covered processes; the owner’s tank program still controls the shell decision.

Close the work pack with a clear exception list and the next decision. If a wind girder blocked a band, name the band. If the response failed near a weld, retain the coordinate and reason. If another method is needed, identify the unresolved question rather than prescribing an unsupported answer.

Frequently asked questions

Technical references

• API — API 653 Aboveground Storage Tank Inspector program
• ASTM E797/E797M-21 — Manual ultrasonic pulse-echo contact thickness measurement
• PetroBot — MagRover
• API — 2026 API 653 Publications Effectivity Sheet
• API — Standards Plan listing for API RP 575
• EEMUA — Publication 159, sixth edition
• HSE — HSG176 Storage of Flammable Liquids in Tanks
• HSE — Corrosion and Materials guidance
• ISO 16809:2025 — Non-destructive testing: ultrasonic thickness determination
• ISO 9712:2021 — Qualification and certification of NDT personnel
• OSHA — 29 CFR 1910.119(j), Mechanical integrity
• PetroBot — Tank Inspection Services