It usually starts the same way. A turnaround crew finishes a UT thickness survey on a 12-inch API 570 piping circuit in crude service, and three grid points come back below t-min. The unit engineer needs an answer inside a week: run to the next turnaround, install a clamp, or pull the spool for repair. The plant has good API 510/570/653 inspectors and a capable Level II bench, but nobody on staff has run more than a handful of API 579-1/ASME FFS-1 Level 2 assessments, and this one is trending toward Level 3. That's the moment most owner-operators pick up the phone and look for an outside ASNT Level III with fitness-for-service experience — not because their inspection program failed, but because the assessment in front of them has moved past what a routine mechanical integrity program is built to handle day to day.
\nFitness-for-service assessment is an engineering discipline performed or approved by qualified engineers; the NDT side of it (flaw detection, sizing and data quality) is where an ASNT Level III contributes. Atlantis NDT does not perform FFS assessments; our ASNT Level III consulting covers NDE procedures and data quality only. The work sits at the intersection of NDT data quality, fracture mechanics, and code interpretation, and getting any one of those three wrong produces a defensible-looking report with the wrong answer. This post walks through what API 579-1/ASME FFS-1 actually requires, where the NDT data feeds directly into the calculation, and the specific situations where bringing in outside expertise stops being optional.
\n\nAPI 579-1/ASME FFS-1, jointly published by API and ASME, is not a single procedure — it's a library of part-specific assessment methods, each built around a distinct damage mechanism. Part 4 covers general metal loss, the straightforward case of broad, relatively uniform thinning across a shell or head. Part 5 covers local metal loss (LTAs — local thin areas), which requires mapping the actual footprint of the thin region rather than averaging it away. Part 6 addresses pitting corrosion, using pit-couple and pit-chart methods that are unforgiving of sloppy pit-depth data. Part 7 covers blistering and HIC/SOHIC damage from wet H2S service. Part 9 handles crack-like flaws, where the assessment is a fracture mechanics calculation built on flaw size, orientation, and material toughness. Part 12 covers dents, gouges, and dent-gouge combinations, common in pipeline and piping mechanical damage cases.
\nEach part uses the same three-tier structure. Level 1 assessments use conservative screening curves that a trained inspector can apply directly from field data — minimum thickness, remaining wall, pit density from a chart. Level 2 requires calculating a Remaining Strength Factor (RSF) and comparing it against a code-allowable RSFa, which demands more precise input data and a working understanding of the underlying equations. Level 3 is reserved for cases where Level 1 and Level 2 don't clear the component, or where the owner-user wants a more accurate, less conservative remaining life instead of walking away from a serviceable asset. Level 3 routinely calls for numerical analysis — elastic-plastic FEA for local metal loss, detailed fracture mechanics for crack-like flaws — and a documented engineering evaluation that a jurisdictional authority or insurer can audit.
\nAPI 579-1 Part 2 is explicit that Level 3 assessments require an engineer experienced in the relevant analysis methods, and most owner-user FFS programs pair that engineering sign-off with an independent NDT review, because the entire calculation rests on data an inspector collected in the field. A plant's in-house Level II staff can be excellent at running the scans and just as good at Level 1 screening, and still have never built the flaw characterization package a Level 3 crack-like flaw case needs: precise through-wall depth profiling, tip location, and an uncertainty budget on the sizing technique itself. That gap — not a deficiency in the base inspection program — is what a fitness-for-service consulting engagement is built to close.
\n\nEvery part of API 579-1 is only as good as the flaw characterization feeding it, and this is where NDT technique selection stops being a formality and starts directly changing the calculated answer.
\nOwner-operators who keep a clean, trended history of thickness and flaw data, rather than a folder of disconnected turnaround reports, make every future FFS case faster and cheaper to run, because half the Level 2/3 workload is reconstructing corrosion rate and prior flaw geometry from old records. This is one of the more overlooked reasons plants standardize inspection data in a system like Atlantis NDT ERP rather than in scattered PDFs: when the next FFS case comes up, the corrosion rate trend and prior UT grid data are already in one place instead of buried in five years of turnaround binders.
\n\nA handful of damage mechanisms account for most outside FFS consulting calls in the industry:
\nA well-scoped outside FFS engagement isn't a rubber stamp on whatever the plant's Level II already found. It typically includes an independent review of the raw NDT data and the technique used to collect it — is the sizing method appropriate for the flaw type and material thickness in the first place — a check of which API 579-1 part and level actually applies given the damage mechanism, an independent RSF or remaining-life calculation run in parallel with, not copied from, the plant's internal number, and a documented peer-review sign-off suitable for a jurisdictional inspector, insurer, or corporate mechanical integrity audit. For crack-like flaw cases specifically, that includes confirming the flaw sizing uncertainty was accounted for in the fracture mechanics input, since API 579-1 Part 9 explicitly calls for the assessment to account for NDE sizing error.
\nThe output is a report an owner-user can defend in front of an API inspector or their own corporate engineering review, not just a spreadsheet with a passing number in it. That documentation and review discipline is the same reason plants pair FFS work with organized NDT reporting software — the underlying UT grids, TOFD sizing data, and prior turnaround history all need to be retrievable and auditable years later when the same component comes up for its next assessment.
\n\nWhen an outside Level III opens a plant's FFS file for independent review, a handful of issues show up often enough to be worth naming. The first is a Level 1 screening curve applied to a flaw geometry the curve wasn't built for — using a general metal loss curve on what is actually a local thin area, which understates the true severity because the averaging effect of general metal loss doesn't apply to a localized flaw. The second is UT thickness grids sized for a routine CML survey rather than for the specific LTA under evaluation — five-point condition monitoring location data is fine for trending general corrosion, but it's the wrong resolution for defining an LTA's exact footprint for a Part 5 calculation. The third is a corrosion rate calculated from only two data points years apart, which produces an average rate that can mask a recent acceleration from a process change, a new corrosion inhibitor program, or a shift in feedstock. The fourth, specific to crack-like flaw cases, is a flaw sizing report that states a single through-wall depth without any stated sizing uncertainty or technique validation — API 579-1 Part 9 expects that uncertainty to be part of the input, and its absence is one of the first things an independent reviewer will flag.
\nNone of these findings mean the original inspection was done poorly. Most come from a Level II inspector correctly executing a standard turnaround scope that was never designed to produce Level 3 FFS-grade input data in the first place. Recognizing that gap early, ideally before the turnaround window closes, is usually cheaper than discovering it during an independent review months later.
\n\nNot every below-t-min reading needs a consultant. A straightforward Level 1 general metal loss case that clears the screening curve with margin is well within most in-house programs' capability. Bring in outside expertise when any of the following apply:
\nAn outside ASNT Level III consulting engagement on an FFS case is typically measured in days of engineering time, not weeks — the value is in getting a defensible, code-correct answer fast enough to make the run/repair decision on the turnaround's actual schedule, not months after the unit is already back on stream.
\n\nThe most effective FFS consulting engagements don't replace a plant's internal API 510/570/653 program — they plug into it. An outside Level III who understands the plant's existing procedures, prior turnaround history, and corrosion rate trends can scope exactly the incremental NDT work an FFS case needs — an extra TOFD scan here, a tighter UT grid there — instead of re-inspecting everything from zero. That's also where building NDT capability in-house pays off over time: a Level II staff that's been through NDT training specifically oriented around FFS-relevant sizing techniques — TOFD, phased array corrosion mapping, encoded scanning — produces field data that's already close to what a Level 3 case needs, which shortens every future engagement and reduces how often an emergency below-t-min reading turns into a scramble.
\nFor plants weighing whether a specific reading, pit survey, or crack indication needs to go to Level 2 or Level 3, or whether an in-house Level II sign-off is sufficient, the fastest path to a clear answer is usually a short scoping call with the FFS engineer and an NDT Level III before committing engineering hours to a full assessment, so the data needed for the assessment level is collected once.
\n\n\n\nAtlantis NDT pairs field expertise with software: NDT inspection management software — Atlantis ERP, a digital twin platform for asset integrity, and NDT reporting software. Build your team with NDT training & certification (ASNT SNT-TC-1A) and ASNT certification pathways, or bring in ASNT Level III consulting. Affordable, accessible, fully customizable — book a free consultation.
\n