Study for CASTI by practicing decision chains: turn UT readings into averaged thickness, thickness into a corrosion rate, and a corrosion rate into remaining life and repair choices. Work numeric scenarios with labeled assumptions, compare API 650, API 653, and API 575 roles, and check yourself against a rubric instead of raw memorization.
Turning UT readings into a corrosion rate you can defend
A defensible corrosion rate comes from comparing thickness at the same location across two dates, using consistent units and a stated basis (short-term versus long-term). Practice writing the calculation chain, not just the final number, so every step can be justified.
Start every practice problem by naming three things: the measurement date and location, the reference thickness or minimum thickness you are comparing against, and which corrosion rate basis you are using. A short-term rate uses the most recent inspection interval; a long-term rate uses the full recorded history. Choosing a basis is itself a judgment step, because a rate that changed between intervals changes the conclusion.
Then practice the downstream steps in order: actual thickness minus minimum required thickness, divided by the rate, gives remaining life. Say out loud what each intermediate number means. If you can explain why remaining life is driven by the thinnest credible point rather than an average, and what data would change your rate choice, you are practicing the reasoning a real inspection decision requires, rather than rehearsing isolated arithmetic.
Which standard governs what: API 650, API 653, and API 575
API 650 covers new tank construction, API 653 covers inspection, repair, alteration, and reconstruction of existing tanks, and API 575 covers inspection methods and practices. Knowing which document answers which question prevents applying construction rules to in-service decisions.
The confusion is structural: the same physical feature, such as a shell course or a bottom, is treated differently by a construction standard and an in-service standard. API 650 tells you what a properly built tank looks like; API 653 tells you how to evaluate a tank that has corroded, settled, or been altered since it was built. Practice tagging each study topic to its home document before you study the content itself.
A useful drill is to take one scenario sentence, such as a tank being returned to service after a shell repair, and list which document governs the repair method, which governs the evaluation of the surrounding metal, and which describes the inspection technique used to find the condition. This mapping habit carries directly into scenario practice, where an early decision is always which framework applies to the facts you are given.
| Document | Primary role | Typical question it answers | Common misapplication |
|---|---|---|---|
| API 650 | New tank design and construction | What minimum shell thickness should a new course have? | Using construction thickness alone to judge an old, corroded shell |
| API 653 | In-service inspection, repair, alteration, reconstruction | Is this existing tank fit to stay in or return to service, and what repair applies? | Treating repair rules as if they were design rules for new builds |
| API 575 | Inspection methods and practices | How is this feature inspected and what is observed? | Confusing an inspection technique description with an acceptance criterion |
Shell course scenario: the one-reading trap
Shell corrosion decisions are sound only when based on an adequate grid of readings and a justified rate, not a single convenient number. The plausible mistake in practice is anchoring remaining life on one spot reading that later proves unrepresentative.
Scenario 1 (practice values, stated assumptions): a shell course shows a minimum required thickness of 0.180 in. A UT scan reports 0.260 in at the worst of several points taken in a small grid. The previous inspection at the same grid showed 0.290 in five years ago, so the short-term rate is (0.290 minus 0.260) divided by 5, or 0.006 in per year. Remaining life is (0.260 minus 0.180) divided by 0.006, about 13 years.
The plausible mistake is to take one additional reading of 0.310 in nearby and recalculate a lower rate, extending the interval, while ignoring that the grid shows a spreading pattern. The better decision is to use the worst credible point and the observed rate trend, note the assumption that the corrosion mechanism is uniform, and flag localized pitting for closer evaluation if the grid is sparse. Why it matters: the rate choice, not the arithmetic, controls the answer, so always write down the grid adequacy and the mechanism before computing.
Floor and annular ring scenario: repair versus continued monitoring
Floor decisions hinge on location: metal near the shell, the annular ring region, carries different requirements than the remainder of the bottom. The plausible mistake is treating a corner-adjacent finding like ordinary floor plate and choosing a simple patch.
Scenario 2 (practice values): vacuum-box testing of a tank bottom finds through-leakage pitting under a plate whose edge sits close to the shell, and UT spot readings suggest thinning in the annular region. A tempting decision is to place a lay-on patch over the pitting and close out. The better decision first asks whether the thinning region falls within the annular ring area governed by stricter thickness expectations, and whether the corrosion is contact-side or soil-side, because that determines the repair type and the extent of further examination.
The better plan distinguishes the failure mode: isolated through-pits on non-annular floor may be candidates for pads, while thinning near the shell-to-floor junction may require replacement plate rather than a patch, plus verification that the repaired area meets the applicable minimum. Why it matters: a repair that is adequate for general floor plate can be inadequate where the shell and bottom interact, so the location test must come before the repair-method choice in your reasoning sequence.
Reading settlement data without overreacting
Settlement evaluation separates edge settlement, bottom settlement, and out-of-plumb of the shell, each with its own measurement approach and acceptance logic. Practice describing what you would measure and what would make a case borderline, rather than memorizing limits in isolation.
Edge settlement is a depression or bulge near the shell measured over a defined reference length; bottom settlement is a general dish or hump of the floor; out-of-plumb is a lean of the shell relative to vertical. These differ in consequence: edge settlement can load the shell-to-floor junction, while general floor settlement mostly affects drainage and support. Keep the three named concepts distinct in your notes and in any scenario write-up.
A practice habit that pays off: for any settlement figure you encounter, sketch it, label the measurement you would need, and list two findings that would move it from acceptable to review, such as associated cracking, distortion of the shell, or changed drainage. Then state what further evaluation you would request. This converts a limit-table topic into a reasoning topic and gives you sentences you can reuse in scenario answers.
In-service versus out-of-service decisions and the documentation that supports them
Inspection planning links the corrosion-rate chain to timing: in-service techniques can carry some evaluation between internal inspections, while certain evaluations and repairs require the tank out of service. Documentation must show the chain so a reviewer can retrace it.
Practice classifying evaluation steps: which observations can be made while the tank remains in service, and which require internal access or emptying. Then practice writing the record that connects them: readings with dates and locations, the rate basis chosen, the minimum thickness applied, the remaining-life figure, and the resulting plan. A record with conclusions but no chain is the classic weak document.
A second habit is to separate what the standard requires from what conditions on site add. Jurisdictional rules, product hazard, and the owner's operating history can all make a conservative choice the right one even where a calculated life is long. In scenario practice, state the code-based step first, then the condition-based adjustment, and keep the two clearly labeled so a reviewer can see which part is calculation and which part is judgment.
A four-week practice sequence with a self-check rubric
Week one, map topics to documents; week two, drill thickness and corrosion-rate chains; week three, add floor, annular, and settlement decisions; week four, run full scenarios under time and score them with the rubric below.
In week one, build a one-page map assigning every syllabus topic to API 650, API 653, or API 575, plus a note on whether the topic is a measurement, an evaluation, or a decision. Weeks two and three, do two numeric chains and one decision scenario per session, always writing the assumption line before computing. Reserve week four for mixed scenarios completed in one sitting, then scored against the rubric rather than against a feeling.
Self-check rubric (learning milestones, not pass predictions): for each scenario, award one point each for naming the governing document, stating the rate basis and assumptions, computing correctly with units, choosing the decision step consistent with location and mechanism, and writing a record a reviewer could retrace. A useful milestone is scoring four of five consistently before exam week; consistently missing the same point tells you which chain link to drill, and the issuer's program page remains the reference for current exam administration details.
Readiness checks before you sit the exam: you can reproduce a corrosion-rate chain with a stated basis in under a minute of setup; you can explain, in two sentences each, why shell, floor-annular, and settlement evaluations differ; and you can write a complete inspection record for any scenario you have practiced without looking at notes.
- Map first: assign each topic to its governing document before studying content.
- Compute with assumptions: state rate basis, grid adequacy, and mechanism before any number.
- Decide by location: run the location test before selecting a repair method.
- Score with the rubric, not with confidence, and drill whichever chain link you miss twice.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.