Prepare for API 570 by practicing the full reasoning chain: classify the piping system, place TMLs where deterioration concentrates, select the correct long-term or short-term corrosion rate, compute remaining life, and match the result to the right action — monitor, repair, re-rate, or replace. Drill this chain with worked numbers and circuit sketches until each step is a decision you can make quickly.
Why API 570 treats a piping circuit, not a single weld, as the inspection unit
API 570 organizes in-service inspection around piping systems: connected lines sharing a service and its deterioration mechanisms. TMLs, classification, and intervals all attach to the circuit, so practice scenarios describe systems, not isolated components.
A circuit runs between defined boundaries — commonly vessel nozzle to vessel nozzle or to battery limits — and carries one process service. Deterioration concentrates where the service changes behavior: downstream of injection points, in deadlegs, at mix points, in bends and low spots. When a practice scenario hands you a line description, the first move is identifying those concentration points, because decisions about TML placement and mechanism identification follow from them.
Build this habit while studying: sketch each circuit you read about, draw its boundaries, and mark every injection point, deadleg, and low-point drain. Then attach a deterioration mechanism to each marked location. This converts scattered code clauses into one mental map, so a scenario about a caustic line with a stagnant branch becomes a navigation problem rather than a memory test.
Four thickness values at every TML: nominal, measured, required, and retirement
Every reading sits among four distinct values: nominal or initial thickness, current measured thickness, the minimum required thickness for that component and pressure, and the retirement criterion that accounts for future service. Mixing them is the core calculation error.
Minimum required thickness comes from the design code formula for the specific component. Straight pipe, elbows, and miter bends do not share one formula, so applying a straight-pipe result to a fitting is a plausible mistake in practice problems. Another recurring trap is the corrosion allowance: the thickness installed included metal intended to corrode, so the floor for today's decision is the required thickness, not the as-installed number.
Drill this by writing all four values for every reading set in your practice questions: nominal or initial, current measured, required at that component, and the retirement criterion including future corrosion allowance consumption. Then state the action implied by where the measured value falls relative to the others. If you cannot name all four values for a scenario in under a minute, that is the gap to fix before touching interval problems.
Long-term versus short-term corrosion rate: choosing the right data pair
The long-term rate divides thickness loss since installation by service time; the short-term rate divides loss between the two most recent readings by the interval between them. The exam expects you to pick the pair that matches the scenario.
Worked practice example: a circuit installed twelve years ago measured 0.280 in initially and 0.220 in today, so the long-term rate is 0.005 in/yr. Four years ago it measured 0.236 in, giving a short-term rate of (0.236 − 0.220) / 4 = 0.004 in/yr. With t_required at 0.180 in, remaining life is 8 years on the long-term rate but 10 years on the short-term rate. A common mistake is feeding the initial thickness into the short-term window, or dividing the recent loss by total service life instead of the interval between readings.
Scenario cues tell you which rate the question wants. An unchanged service over a long life points to the long-term rate. A changed process, a new injection point, or a sharply accelerating reading pair points to the short-term rate, because history no longer represents current conditions. When you review any practice item, underline the cue first, then compute — never compute both and guess which one sounds better.
| Feature | Long-term rate | Short-term rate |
|---|---|---|
| Thickness readings used | Initial (or as-installed) and current | Two most recent readings |
| Time divisor | Total time in service | Interval between the two readings |
| Best suited when | Service has been steady over the life of the line | Conditions changed recently or loss is accelerating |
| Tendency if misapplied | Can smooth away recent acceleration | Can overstate or understate a stable, slow-eroding line |
From rate to interval: remaining life, half-life, and the controlling limit
Remaining life is the difference between current and required thickness divided by the corrosion rate. The code then caps inspection intervals by fractions tied to remaining life and to the half-life of the corrosion allowance; the smaller limit governs.
Continuing the worked example: with 0.220 in measured, 0.180 in required, and a 0.005 in/yr rate, remaining life is 8 years. Suppose the corrosion allowance concept implies a half-life figure shorter than that; the interval must satisfy both constraints, so the tighter one controls. A plausible mistake is treating a classification-based base interval as an entitlement and ignoring the remaining-life cap entirely — the caps are maximums, and deterioration data can shorten an interval below them.
Convert every rate you calculate into three outputs before moving on: remaining life, the half-life of the remaining allowance, and the resulting interval check. When a practice question offers several interval choices, eliminate any that exceed the controlling limit first, then compare the survivors against the classification and documentation conditions. This ordering makes numeric answer choices self-checking instead of guesswork.
Class 1, 2, and 3 services: matching consequences to coverage
API 570 sorts piping into consequence classes, with Class 1 covering services whose leakage carries the most severe consequences. Class drives the base inspection interval and coverage expectations, and adjustments require documented corrosion data.
Classification hinges on the service description in the code's definitions — factors such as flammability and toxicity at specified thresholds — not on the fluid's everyday name. A scenario describing a 'process water line' may still qualify for a stricter class if carryover or mix-point conditions fit a higher-consequence definition. When reviewing, read each class definition fully and list the service descriptors that trigger it; do not shortcut by fluid category.
Second decision layer: the code permits owner-users to adjust intervals when supported by documented inspection history and corrosion data. Practice items test whether an extension is justified or merely desired. Train the conditional: no documented rate data, no defensible extension. Pair each classification question with the interval question it feeds, because the two concepts are designed to be answered as one chain.
Scenario drill: what to decide when a reading falls below required thickness
When current measured thickness reaches or falls below the minimum required thickness, remaining life is effectively exhausted and the decision shifts from scheduling to action: repair, replacement, re-rating, or removal from service under the code's conditions.
Worked practice scenario: a reading comes in at 0.175 in against a required 0.180 in. A plausible mistake is computing remaining life with a long-term rate that averages twelve mild years, getting a small positive number, and concluding the line can keep running on a shortened monitoring interval. The better decision recognizes that a sub-thickness reading is an action trigger — the evaluation path is repair, replacement, re-rating to a lower pressure, or shutdown — not an interval adjustment. Getting this sequence wrong matters because it substitutes a paperwork answer for a fitness-for-service decision.
Practice the adjacent case too: a reading above required thickness but at the retirement criterion leaves a defined window to plan and execute a repair before the limit is reached. Train yourself to name the action category first — continue with monitoring, plan a repair, or act immediately — and only then compute supporting numbers. This ordering keeps the math honest and mirrors how scenario questions are constructed.
A four-week preparation sequence with self-check rubric
Sequence study in four passes: code structure and definitions, calculation drills, circuit and classification scenarios, then timed mixed practice. Finish each week with a written self-check so readiness is observed, not assumed.
Week one: read the code for structure, then build flashcards for definitions — circuit, TML, deadleg, injection point, class boundaries. Week two: drill thickness, rate, and remaining-life calculations until you can set up any of them from raw data without notes. Week three: work scenario sets that chain classification to TML placement to interval decisions. Week four: run timed mixed sessions where you must locate clauses in the code while solving, because lookup speed is part of the skill.
Capstone exercise: draw a fictional unit with one injection point, one deadleg, and two circuits of different service severity. Place TMLs, assign classes with your stated reasons, invent a reading history, and compute a rate, remaining life, and interval for each circuit. Use the rubric below to grade yourself. Milestone scores are learning checkpoints only — they measure your drill performance, not a predicted exam result.
One short administrative note: confirm application steps, exam windows, and eligibility directly with API's Individual Certification Programs rather than relying on secondary summaries, since administrative details change.
- Rubric — TML map: an injection point, deadleg, and low point each carry at least one TML (0–2 points).
- Rubric — classification: each class assignment cites a specific service descriptor from the code definitions (0–2 points).
- Rubric — calculation: the rate uses the correct reading pair, remaining life uses current minus required thickness, and the interval check names the controlling limit (0–3 points).
- Rubric — action: each circuit ends in a named action category — monitor, plan repair, or immediate action — justified by the numbers (0–3 points).
- Readiness check: you can complete one full chain (classify, place TML, compute, decide) in under fifteen minutes using the code only for lookup.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.