Treat preparation as decision practice: classify every scenario by governing standard and work type first, compute corrosion rate and remaining life from paired thickness readings, and name the examination method that verifies each repair. Score yourself with a rubric each week so gaps surface in classification and math, not on exam day.
Which API Standard Governs the Decision: 653, 650, 651, or 652
Scope separation is the first decision in any tank scenario: API 650 covers new tank construction, API 653 covers in-service inspection, repair, alteration, and reconstruction, API 651 addresses tank bottom cathodic protection, and API 652 addresses tank bottom linings.
Memorizing scope statements in isolation leaves you unprepared for the actual task: deciding which document's requirements apply to a mixed situation. Build a labeling habit instead. Before working any scenario's details, tag it as construction, in-service, bottoms-environment, or lining work, then verify the tag against each standard's scope. The tag tells you which document's examination, testing, and repair provisions you should reach for, and working from the wrong scope produces answers that are internally consistent but governed by the wrong requirements.
Reconstruction is the boundary people miss. Relocating or substantially rebuilding an existing tank looks like construction, but API 653's reconstruction provisions govern that work, so requirements for foundations, examination, and return-to-service testing come from 653 rather than 650. Similarly, a question about protecting a tank bottom from soil-side corrosion sounds electrical but belongs to API 651, and internal lining evaluation belongs to API 652. Drill this by taking five mixed scenarios and tagging the governing standard for each, then compare your tags against the explanations in your question bank.
Scenario 1: Turning Thickness Readings into Corrosion Rate and Remaining Life
Interval logic starts with data, not assumptions: compute the corrosion rate from two inspections at the same location, then remaining life as current thickness minus tmin, divided by that rate, and set the interval inside code limits.
Worked example: a condition-monitoring location on the first shell course read 0.28 in in 2014 and 0.22 in in 2024, with a calculated tmin of 0.15 in. The corrosion rate is (0.28 − 0.22) ÷ 10 = 0.006 in/yr. Remaining life is (0.22 − 0.15) ÷ 0.006 ≈ 11.7 years. The two conceptual traps in this chain are substituting a nominal assumed rate for the measured one, and treating the full remaining life as the inspection interval rather than dividing it by the applicable factor. Keep the steps — rate, remaining life, interval — as separate named operations in your notes so you can check each one independently.
A plausible mistake: pairing readings from different locations, or averaging unrelated CMLs, which produces a rate that fits neither point. The better decision is to confirm the two readings come from the same location and comparable service, check that product and specific gravity have not changed, and recalculate tmin if they have. This matters because an interval built on an assumed or mismatched rate either forces unnecessary out-of-service inspections or allows real degradation to outrun the next examination.
Make the same computation a daily drill: generate pairs of readings and dates, and practice until the three steps — rate, remaining life, interval — run without notes.
Shell Minimum Thickness: Calculated tmin versus Measured Thickness
Shell tmin varies by course because hydrostatic head increases with depth. The inspector evaluates each course against its own tmin — from a calculation method for in-service tanks or from original construction records — and compares it with measured UT thickness.
The one-foot method illustrates the concept: shell thickness requirements are driven by the pressure head of stored liquid, so the bottom shell course demands the most thickness and upper courses demand less. When a stem gives you tank diameter, course heights, and specific gravity, your first task is identifying which course each CML belongs to, because a single tank-wide tmin does not exist. Match every thickness reading to its course before judging acceptability.
The classic error is applying the bottom course's tmin to every course, which makes sound upper courses look deficient and thin upper courses look acceptable. A subtler error is pulling tmin from original design records without checking whether the stored product's specific gravity changed — a switch to heavier product raises the head-driven requirement. The defensible approach is to recalculate tmin per course under current service conditions, then compare each CML against its own course requirement and flag any reading that falls below it for closer evaluation.
Practice check: sketch a four-course tank, assign a plausible tmin to each course, and confirm you can state why the values differ before moving to calculation-based practice.
Classifying the Work: Repair, Alteration, or Reconstruction
Classify every work scenario before evaluating it. Repairs restore material, alterations change physical or design characteristics, and reconstruction re-erects a tank. The classification drives which welding, examination, and testing requirements the correct answer reflects.
The definitions carry practical weight. Replacing a corroded shell area with an insert plate of equivalent material and dimensions is typically a repair in kind, while adding a new shell nozzle where none existed is an alteration, and dismantling a tank for re-erection elsewhere is reconstruction. The distinction matters because each category carries different welding, examination, and testing requirements: a work procedure that satisfies the requirements for a repair may not satisfy those for an alteration, so evaluating the work under the wrong category produces a sound-looking procedure attached to the wrong standard.
Use the stem's physical description, not its vocabulary. A stem that says "we repaired the tank by adding a new shell connection" has described an alteration regardless of the verb used. Read what physically changed — material replaced, dimension changed, or tank re-erected — and classify from that. The table below is a reusable decision aid; rebuild it from your own notes so you internalize the logic rather than the layout.
When classifying feels uncertain, ask: did the work restore what was there, add something new, or move/rebuild the structure?
| Work described in the stem | Typical classification | What to check next |
|---|---|---|
| Replacing a corroded shell area with an equivalent insert plate | Repair | Welding details, required examination, and whether the code edition requires hydrostatic or alternative testing |
| Adding a new shell nozzle where none existed | Alteration | Design basis, reinforcement area, and examination and testing applicable to alterations |
| Dismantling and re-erecting a tank at a new site | Reconstruction | Reconstruction provisions, foundation condition, and testing before return to service |
| Weld buildup or laid plates over floor plate pits | Repair | Repair organization procedures and post-repair examination, such as vacuum-box testing |
Scenario 2: Floor Pit Depth, Underside Unknowns, and Repair Verification
For bottom corrosion, compare pit depth against the applicable floor minimum, weigh whether soil-side condition is actually known, and verify any repair by examination — never by assuming the patch restored integrity.
Worked example: an internal inspection finds clustered pits on a 0.25 in floor plate, deepest at 0.12 in, leaving about 0.13 in of metal. The tempting shortcut is to declare the floor acceptable because a remembered remainder "sounds fine." The better evaluation checks the governing floor criteria, then confronts the underside question: leak history, cathodic-protection status per API 651 concepts, and whether a lining per API 652 changes how surface corrosion should be read. Underside corrosion is invisible during internal inspection, so its condition must come from records and indicators, not optimism.
On the repair decision, a plausible mistake is specifying weld buildup across an area larger than that technique is suited for, when a laid plate is the appropriate remedy for broader thinning. The better decision is matching technique to the corrosion pattern, then verifying: vacuum-box examination of the completed welds and a dated entry in the inspection record. This matters because the repair's defensibility rests on examination evidence and documentation — an unverified patch is indistinguishable from no repair at all during the next audit or inspection.
Drill the pattern: evaluate depth, evaluate the unknown underside, choose the technique, name the verification. Four answers per scenario, every time.
Settlement: Edge, Bottom, Peaking, and Banding Are Different Problems
Edge settlement is vertical displacement at the shell-to-bottom junction; bottom settlement is differential settlement across the floor; peaking and banding are shell roundness deviations at welds. Each condition has its own measurement description and evaluation approach.
Stems describe the measurement setup, and the setup identifies the condition. A straightedge or lattice measurement run across the floor adjacent to the shell describes edge settlement; a broad dishing measured across the floor away from the shell describes bottom settlement; deviations measured along a shell weld — bulging at a vertical seam or hollowing at a girth seam — describe peaking and banding. Practice reading the setup first, then map it to the condition, then to the evaluation that applies to that condition.
A plausible mistake: treating any floor depression as edge settlement and applying the wrong acceptance framework to what is actually a bottom settlement profile, or missing that the measurements were taken along a weld and are a roundness issue, not a settlement issue at all. The distinction matters because the conditions drive different concerns — junction stress at the shell-to-bottom joint, plate strain and drainage across the floor, or shell roundness at seams — so the correct answer describes a different evaluation for each.
Self-test: write one-sentence measurement descriptions for all four conditions from memory, then check them against your study material's definitions.
A Scored Classification Drill and an Eight-Week Preparation Sequence
Build fluency with weekly mixed drills: for each scenario, record the governing standard, the interval math, the work classification, and the verification method. Score against a rubric and let the error pattern set next week's focus.
The exercise: take ten mixed scenarios and produce four labels per scenario — governing standard, work classification, computed interval (where thickness data is given), and verification method. When you score the drill, check specifically for two error sources: classification mistakes at repair-versus-alteration boundaries, and interval mistakes traced to mismatched reading pairs or a forgotten factor. Treat eight or more fully correct scenario chains as a milestone, not a passing prediction. When a chain breaks, reread the relevant scope or definition rather than only the answer explanation, because the explanation teaches one question while the scope teaches the family.
An adaptable eight-week sequence: weeks one and two, scope boundaries and definitions with daily scenario-tagging; weeks three and four, thickness math — tmin per course, corrosion rate, remaining life, interval — as timed numeric drills; week five, work classification using the decision table until it's reflexive; week six, settlement identification and floor evaluation; week seven, mixed timed sets plus one documentation exercise per scenario; week eight, error-log review and a final scored drill. Documentation practice matters: write each scenario's record entry — findings, calculations, repair, verification — as you would in the field.
- Readiness check 1: compute rate, remaining life, and interval from two paired readings without notes, units correct
- Readiness check 2: classify ten work descriptions correctly using physical change, not stem vocabulary
- Readiness check 3: name a verification method for each repair type you classified
- Readiness check 4: assign the governing standard to five mixed scenarios and justify each assignment in one sentence
- Readiness check 5: produce a complete inspection-record entry for one repair scenario, including examination results
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.