Treat API 653 as a layer built on top of API 650 and API 575. For every practice question, identify which document governs, then apply the suitability-for-service logic: establish minimum required thickness, compute corrosion rates from thickness histories, evaluate bottoms and shells against those minima, and classify any change to the tank as repair, alteration, or reconstruction. Drill this sequence until document choice and calculation become two separate, deliberate steps.
Choosing the Governing Document Before You Calculate Anything
API 653 governs in-service inspection, repair, alteration, and reconstruction, but it borrows construction logic from API 650 and inspection practices from API 575. Practice tagging every question with the document that actually supplies its answer before computing anything.
The three documents occupy different layers. API 650 contains the design and construction rules used for new tanks; API 653 references those rules when evaluating, repairing, or rebuilding a tank already in service; API 575 describes how to inspect tanks and what records to keep. A question about evaluating a corroded shell course is answered from API 653, but its thickness logic may point back into API 650. Train yourself to tag every practice question with the document that actually supplies the answer, and write the tag at the top of your scratch work before touching any formula.
Build a simple decision map on one page. New tank construction goes to API 650. Evaluation of an existing tank, repair, alteration, relocation, and reconstruction go to API 653. How an inspection is performed, what techniques are used, and how results are recorded go to API 575. The edge cases deserve extra reps: a tank that was dismantled and re-erected, and a tank whose service conditions changed, both live in the boundary zone where picking the wrong document produces a confidently wrong answer.
| Situation | Primary document | What it supplies |
|---|---|---|
| New tank being designed and built | API 650 | Design thickness, materials, construction details |
| In-service evaluation, repair, alteration, reconstruction | API 653 | Minimum required thickness logic, inspection intervals, repair rules |
| How to inspect and record | API 575 | Inspection methods, out-of-service inspection scope, records |
| Corrosion and damage mechanisms background | API 571 | Mechanism descriptions that explain why loss occurs |
Two Thickness Numbers You Must Keep Separate: tmin and As-Found
Minimum required thickness (tmin) is the code-calculated floor for continued service under current conditions; the as-found UT reading is a measurement. Confusing the two, or substituting the nameplate design thickness for tmin, corrupts every downstream decision.
For a shell course, tmin depends on the tank's radius, the height of liquid that course actually sees, the design specific gravity of the product, and the joint efficiency of the welds. That means tmin for an in-service tank is not simply the thickness on the nameplate, which reflected original design conditions. Older riveted and lap-welded tanks introduce additional considerations because their joint configurations differ from modern butt-welded shells.
The practical drill: for each practice tank, write down three numbers in separate columns — original design thickness from the nameplate, calculated tmin for current service, and the as-found UT reading. The as-found value compared against tmin tells you about the tank's present condition. The design thickness compared against the as-found value tells you how much metal the tank has lost over its life. The exercise in Section 6 uses exactly this column layout, and your rubric should penalize any answer where the columns blur together.
Corrosion Rates: Short-Term Versus Long-Term and What Each Drives
The long-term rate uses original thickness spread over total service life; the short-term rate uses the change since the previous inspection. API 653 directs you toward the more conservative of the two when projecting the future.
Worked scenario A. A tank's nameplate shell thickness is 0.375 in; current UT reads 0.290 in after 20 years of service; a reading taken 5 years ago was 0.305 in. The long-term rate is (0.375 − 0.290) / 20 = 0.00425 in/yr. The short-term rate is (0.305 − 0.290) / 5 = 0.0030 in/yr. A plausible mistake is computing only the long-term rate and moving on. The better decision is computing both, recognizing they differ, and applying the direction the code gives toward the more conservative result — here, the long-term rate.
Why it matters: remaining life is projected as the difference between current thickness and tmin, divided by the governing rate, and the inspection interval logic in API 653 builds on that projected life. With tmin assumed at 0.200 in for this illustrative tank, the long-term rate gives about 21 years of remaining life versus 30 years under the short-term rate. Which rate you pick changes the answer materially, so verify the exact interval relationships in the edition you study and always confirm which rate the question's data supports.
Tank Bottoms: Separating General Metal Loss From Pitting
Bottom evaluation runs on two parallel tracks: general thinning is judged against the bottom minimum thickness, while pitting is assessed under separate pitting criteria. Both come from the same inspection data, so conflating them misclassifies an otherwise sound bottom.
General metal loss is a broad thinning of bottom plate area and is compared against the bottom tmin value; API 653 also treats pitting through its own criteria, which can permit localized pitting deeper than what general loss would allow. A bottom reading that fails one track may pass the other, so the evaluator's first job is classifying each indication. Small, isolated pits distributed across a plate behave very differently from a widespread thinned region, which is why the code separates them.
Method selection reflects the same split. A magnetic flux leakage (MFL) sweep of the bottom screens for areas of metal loss across large areas; UT spot readings then quantify the loss found. Vacuum box testing of bottom welds addresses leak-path integrity, which is a different question from remaining metal. When practicing, describe for each paper tank which method answers which question, and note that out-of-service bottom evaluation combines these techniques rather than treating any single one as complete.
Repair, Alteration, or Reconstruction: Picking the Correct Path
API 653 classifies work on a tank as repair, alteration, or reconstruction, and the classification determines which design rules, welding requirements, and examination apply. Misclassifying the work applies the wrong acceptance framework.
Worked scenario B. A tank needs a nozzle added to the second shell course and, during the same outage, a degraded patch of bottom plate replaced. A plausible mistake is filing both under generic repair and applying only repair-section rules. The better decision is classifying them separately: adding a nozzle to an existing shell is an alteration because it changes the tank's original design configuration, while replacing bottom plate within the repair framework follows the repair rules. Reconstruction, by contrast, involves re-erecting the tank or relocating it, pulling in design-construction considerations much more heavily.
Why it matters: the classification drives which weld procedures, examination requirements, and testing the finished work needs, so the wrong label produces work executed to the wrong standard even if the welds themselves are sound. In practice sessions, write the definition-based test for each category before choosing: does the work restore to original condition, change the design, or rebuild the tank? Then trace which set of requirements your choice activates and check that your answer states the classification explicitly.
Practice Exercise: Build a Decision Map and Stress-Test It With Paper Tanks
Construct a one-page flow map from question type to governing document to calculation, then run three paper tank cases through it. Score yourself against the rubric below; the map should catch your classification errors before the exam does.
Draw the map in this order: a first fork separating new construction from in-service work, a second fork separating evaluation from physical work (repair, alteration, reconstruction), and terminal boxes for the calculations — tmin, corrosion rate, remaining life, interval. Then write three paper cases: a corroded shell with two thickness histories, a bottom with mixed general loss and pitting, and a tank receiving both a nozzle addition and a plate repair. For each, annotate the map path taken and the final decision, including which rate or classification governed.
Expected observations: your first run-through will expose places where the map forces you to ask a question you had skipped, most commonly which thickness history supports which rate, and whether current service conditions match the nameplate design. Those friction points are exactly where deliberate practice pays off. Run the same cases again two days later without the notes; the second run measures whether the sequence stuck rather than the paperwork.
- Rubric item 1: identified the governing document before any calculation, and it matches the situation.
- Rubric item 2: kept design thickness, tmin, and as-found thickness in distinct columns with correct sources.
- Rubric item 3: computed both corrosion rates where data permitted and justified which one governed.
- Rubric item 4: classified general loss and pitting separately and named the inspection method for each question asked.
- Rubric item 5: stated the work classification (repair, alteration, reconstruction) explicitly with the definitional reason.
- Milestone: on a clean second run, all five items correct on all three cases signals the map is working; this is a learning checkpoint, not a passing prediction.
A Preparation Sequence and Concrete Readiness Checks
Sequence the work in three passes: map the documents, drill the calculations, then run full scenario cases. Readiness is demonstrated by correct document selection and self-consistent calculations under time, not by page counts.
Suggested adaptable sequence: spend the first phase reading API 653 alongside API 650 and API 575 only to build the decision map from Section 1, recording every cross-reference you encounter. The second phase drills calculations on hand-made thickness histories until tmin, both corrosion rates, and remaining life are routine. The third phase runs multi-part paper tanks that force classification, evaluation, and interval decisions in one sitting, mirroring how exam scenarios combine concepts. Adjust the phase lengths to your starting familiarity rather than copying any fixed schedule.
Readiness checks you can actually observe: you can classify any work description into repair, alteration, or reconstruction within seconds and cite the definitional reason; you can compute both corrosion rates and justify the governing one from raw data alone; you can explain why a pitting result and a general-loss result on the same plate can yield different conclusions; and you can state which document supplies each step of an answer. If any check fails, return to that section's exercise rather than rereading passively. For scheduling windows, application steps, and administrative requirements, use API's Individual Certification Programs page directly rather than secondary summaries.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.