A UT thickness reading by itself tells you almost nothing. A crude unit piping spool reading 0.312 in. against a nominal 0.375 in. wall could be in excellent shape or six months from a leak — the number alone can't tell you which. What tells you which is the remaining life calculation: how fast the wall is thinning, how much wall you're required to keep, and how long you have before you cross that line. Get this calculation right and your inspection interval is defensible in an audit, in a management-of-change review, and in front of a jurisdictional inspector. Get it wrong — usually by using the wrong corrosion rate, the wrong retirement thickness, or comparing readings from different corrosion monitoring locations as if they were the same point — and you either inspect assets that don't need it or, worse, miss the ones that do.
\nThis is the calculation that sits underneath every risk-based inspection (RBI) program built to API 580/581, every API 570 piping circuit reassessment, every API 653 tank internal/external interval, and every API 510 pressure vessel fitness-for-service review. It's also one of the most commonly miscalculated numbers in the industry, not because the math is hard — it's arithmetic — but because the inputs are so easy to get sloppy about.
\n\nThe remaining life (RL) equation is simple on paper:
\nRL = (t_actual − t_min) / CR
\nwhere t_actual is the current measured wall thickness, t_min is the minimum required (retirement) thickness set by the governing design code or a fitness-for-service assessment, and CR is the corrosion rate in inches (or mils) per year. The retirement thickness itself comes from a pressure design calculation — for piping, typically the ASME B31.3 or B31.1 minimum wall equation based on design pressure, temperature, allowable stress, and pipe outside diameter; for pressure vessels, the ASME Section VIII, Division 1 formula for the governing component (shell, head, nozzle); for tanks, the API 653 minimum shell thickness table by course and diameter, or the calculated bottom plate minimum.
\nWhere this goes wrong in practice: engineers pull t_min from a generic \"corrosion allowance\" figure on an old datasheet instead of recalculating it against current code minimums, or they use nominal thickness instead of actual as-built thickness as the numerator's starting point. Both errors compound. A retirement thickness that's 20 mils too conservative can force an unnecessary shutdown; one that's 20 mils too thin can push an asset past its safe operating life without anyone noticing, because the RL calculation will report a rosier number than reality supports.
\n\nCR is where most bad remaining-life numbers actually come from. You have two candidate rates: the short-term rate, calculated from the two most recent inspections, and the long-term rate, calculated from the original (or earliest reliable) thickness to the most recent reading, averaged over the full service life. API 570 and API 653 both direct you toward the more conservative — that is, the higher — of the two, unless there's a documented, understood reason the short-term rate reflects a real and permanent change in service conditions: a feedstock change, a new corrosion inhibitor program, a process unit revamp.
\nIn practice this means: if a piping circuit ran at 2 mils per year for fifteen years and then the last two inspections show 8 mils per year, you don't get to average that into a comfortable 3 mils/year long-term rate. You use 8 mils/year until you understand why it changed — new sulfur content in the crude slate, a lost corrosion inhibitor injection point, a change in flow regime that's now causing erosion-corrosion at an elbow. Conversely, a single anomalously high short-term reading driven by measurement error or a localized, non-representative CML shouldn't be allowed to drive the whole circuit's interval down without checking it against a second UT technique or a grid of surrounding points.
\nEach governing code translates remaining life into an inspection interval slightly differently, and mixing them up is a common finding in third-party integrity audits.
\nFor in-service piping, API 570 sets the reinspection interval at the lesser of half the remaining life or a fixed code maximum — tighter for Class 1 (highly hazardous, high-consequence) circuits than for Class 2 or 3 piping. A circuit calculated at 9.0 mils/year with 0.088 in. of wall left above minimum has roughly 9.8 years of remaining life; the reinspection interval is capped well inside that window, not run out to the full RL figure. This half-life convention exists specifically to catch an acceleration in corrosion rate before the circuit gets anywhere near its retirement thickness.
\nTank floor and shell remaining life is calculated separately by component, because the corrosion mechanisms differ: floor plates typically corrode from the soil side (external) or product side (internal, often concentrated at the water-oil interface), while shell courses corrode more uniformly from atmospheric or vapor-space exposure. Internal inspection intervals are tied to floor remaining life, capped at a code maximum regardless of how long the RL math says you have — because internal floor corrosion under coating disbondment or microbiologically influenced corrosion (MIC) can be highly localized and easy to underestimate from a sparse UT grid alone. This is why 100% floor scanning with magnetic flux leakage (MFL), followed by UT verification of flagged indications, has become standard practice rather than an optional extra on tanks past a certain age.
\nVessel remaining life follows the same core equation, but t_min is calculated per component — shell, head, nozzle neck, reinforcing pad — using the applicable ASME Section VIII design formulas, and the governing component is whichever has the shortest remaining life. That's frequently a nozzle neck at a process connection rather than the shell, because nozzles see turbulence-driven erosion-corrosion and are, in a lot of programs, under-monitored relative to the shell courses that get most of the UT grid attention.
\n\nA remaining life number is only as good as the CML it came from, and CML placement is where a lot of programs quietly fail. A grid of five thickness points on a straight run of 20-inch pipe tells you almost nothing about the elbow, tee, or reducer in that same circuit, because those fittings see different flow dynamics and corrode at different rates — often two to three times faster at an elbow extrados or immediately downstream of a control valve, where flashing or erosion-corrosion concentrates wall loss.
\nA defensible CML program places points at:
\nGrid consistency matters as much as grid placement. If successive inspections don't hit the exact same CML — same location, same orientation, same measurement technique — the resulting \"corrosion rate\" is really just noise between two unrelated data points. This is one of the areas where scanning UT or phased array corrosion mapping earns its keep over single-point spot readings: a C-scan or grid map at a CML captures the true minimum thickness in the footprint rather than whatever single point the technician happened to hit, and it re-locates reliably from one turnaround to the next using photographic and dimensional references.
\n\nConsider a 12-inch, Class 1 piping circuit carrying atmospheric tower bottoms at 650°F in a crude unit processing a sour, high-TAN feedstock. Sulfidic corrosion is expected and evaluated against the McConomy curves relative to sulfur content and operating temperature. Original nominal wall is 0.500 in. (Schedule 40); t_min per the current B31.3 calculation is 0.291 in.
\nTurnaround inspection history: 0.437 in. eight years ago, 0.402 in. four years ago, and 0.379 in. this turnaround. Long-term rate from original nominal, over 22 years in service: (0.500 − 0.379)/22 = 5.5 mils/year. Short-term rate from the last two readings: (0.402 − 0.379)/4 = 5.75 mils/year. The two are close enough here that the governing rate is straightforward — the higher figure, 5.75 mils/year.
\nRemaining life: (0.379 − 0.291)/0.00575 = 15.3 years. Under API 570's half-life convention for a Class 1 circuit, the reinspection interval is capped well below that full figure — in practice, most programs would set this circuit for reinspection inside 5 years, both because of the Class 1 designation and because sulfidic corrosion mechanisms are known to accelerate nonlinearly if feedstock sulfur content or TAN increases. The RBI program should also flag this circuit for closer monitoring the moment the crude slate changes, independent of the calendar-based interval, because corrosion rate assumptions in an RBI model are only valid as long as the process conditions they were built on stay the same.
\n\nThe output that actually matters to a plant manager, an insurance underwriter, or a jurisdictional inspector isn't the remaining life number itself — it's the inspection interval derived from it, along with the documented basis for that interval. A defensible package includes the governing corrosion rate and which of short-term/long-term was used and why, the t_min calculation and its code basis, the CML history showing consistent measurement locations, and the code-mandated interval cap applied on top of the raw RL/2 math. Auditors reviewing a mechanical integrity program under OSHA's Process Safety Management standard (29 CFR 1910.119) or a comparable state program will ask for exactly this chain, not just a spreadsheet with a final \"years remaining\" column.
\nThis is also where the practical difference between a Level II technician taking a UT reading and an ASNT Level III reviewing the resulting remaining-life package shows up: the Level III role is to catch exactly the errors above — the mismatched CML, the stale t_min, the averaged corrosion rate — before they turn into an inspection interval that looks defensible on paper but wouldn't survive a closer look.
\n\nA single-circuit remaining life calculation is a five-minute spreadsheet exercise. A refinery with several thousand active CMLs across hundreds of piping circuits, dozens of vessels, and a tank farm is a different problem entirely. Spreadsheet-based tracking tends to fail in predictable ways: CML naming conventions drift between turnarounds, the corrosion-rate formula gets copy-pasted incorrectly into a new row, and nobody notices a circuit's interval quietly slipping past its due date because there's no automated flag tied to the calculation itself.
\nThis is the gap that purpose-built NDT inspection management software and NDT reporting software are built to close — CML history that persists automatically across turnarounds, corrosion rate and remaining life calculated the same way every time with the correct governing-rate logic built in, and inspection intervals that flag themselves as they approach the code-mandated cap rather than relying on someone remembering to check a spreadsheet. For facilities managing integrity data across a whole asset base, a digital twin platform takes this further by mapping CML history and remaining-life status directly onto a 3D model of the unit, so a turnaround planner can see at a glance which circuits, vessels, and tank components are approaching their reinspection window without hunting through separate reports for each equipment type.
\n\nRemaining life calculation is a core competency that belongs inside an owner-operator's mechanical integrity team, not solely with a third-party inspection contractor, because the people setting inspection priorities need to understand exactly what's driving each interval. NDT training that covers UT thickness gauging alongside the API 570/653/510 remaining-life framework gives inspection and reliability staff the ability to sanity-check a contractor's numbers rather than simply accepting a report at face value. For programs that need an outside check on methodology — CML grid design, RBI corrosion rate assumptions, or a specific fitness-for-service question — bringing in ASNT Level III consulting for a program review is typically far cheaper than discovering the gap during an incident investigation.
\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