{"slug":"/standards/api-578","title":"API RP 578: Material Verification and PMI for Alloy Piping","description":"API RP 578 sets out a material verification programme for alloy piping, including silicon content in carbon steel below 0.10% for sulfidation service.","h1":"Material verification programmes under API RP 578: scope, extent and records","answer":"API RP 578 describes a material verification programme (MVP) for alloy components in new construction and existing piping systems. It defines when positive material identification is performed, who is qualified to perform it, how much of a system is examined, and how results are recorded. It sets no design or acceptance limits itself; the owner's specification and the referencing code do.","expansion":"The practice exists because alloy mix-ups are silent. A 316L flange in a 6Mo line, a carbon steel elbow in a 1¼Cr–½Mo hot line, or a 9Cr spool substituted with 5Cr will pass every dimensional and NDE check and then fail in service, sometimes catastrophically. Material verification is therefore a supply-chain and records discipline dressed as an examination technique. RP 578 asks the owner to write a programme that names the systems in scope, the extent of examination for each, the analytical method and its detection limits, the qualification of the technician, and the disposition route for a non-conforming component. The commonest technical trap is the instrument itself: a handheld XRF cannot see carbon, so it cannot distinguish 304 from 304L, nor confirm carbon steel grade. Where the carbon or low-atomic-number content governs, optical emission spectroscopy or laboratory analysis is the only defensible method.","source":"Sources: API RP 578 Guidelines for a Material Verification Program (MVP) for New and Existing Alloy Piping Systems; API 570 Piping Inspection Code; API RP 939-C Guidelines for Avoiding Sulfidation Corrosion Failures in Oil Refineries; ASTM A106, A312 and A335 material specifications; ASTM E1476 metals identification and sorting; ASTM E1621 wavelength dispersive XRF; OSHA 29 CFR 1910.119 mechanical integrity.","table":{"caption":"Alloy substitutions a material verification programme is written to catch","columns":["Intended material","Substitution found in practice","Why other checks miss it","Service consequence","Method that detects it"],"rows":[["Type 316L stainless","Type 316, higher carbon","Identical appearance and identical XRF spectrum","Sensitisation and intergranular attack after welding","OES or laboratory analysis"],["1¼Cr–½Mo piping","Carbon steel elbow or fitting","Passes dimensional and NDE; markings lost at cutting","Accelerated high-temperature oxidation and creep damage","Handheld XRF"],["9Cr–1Mo (P91)","5Cr–½Mo (P5) spool","Similar mill markings and surface appearance","Loss of creep strength at design temperature","Handheld XRF with a current alloy library"],["Carbon steel above 0.10% Si","Low-silicon carbon steel","Both fully meet the same ASTM specification","Sulfidation rate far above the predicted curve","OES quantifying silicon"],["Matching alloy weld deposit","Mismatched filler from an unlabelled quiver","Base metal PMI passes; the weld is never tested","Preferential corrosion or cracking at the joint","XRF or OES on a cleaned weld cap"],["Nickel alloy 625 valve trim","Type 316 trim","Assembled component; no visible external difference","Rapid attack in the intended corrosive service","XRF on the disassembled component"]],"note":"Every row is a component that passes visual, dimensional and volumetric NDE. None of them is found by inspection of workmanship."},"facets":[{"q":"Does API RP 578 apply to new construction or existing piping?","a":"Both. For new construction the programme targets receipt inspection, identity transfer at cutting, fabricated spools and weld deposit, where a substitution can still be replaced cheaply. For existing systems it targets components already installed in high-consequence service, usually examined during turnarounds against a risk-based extent. The two halves need different scope statements, different sampling logic and different record destinations, and a single generic programme usually serves neither well."},{"q":"Can handheld XRF verify that stainless steel is an L grade?","a":"No. X-ray fluorescence cannot detect carbon, so it cannot distinguish 316 from 316L or 304 from 304L. Where the carbon grade governs — sensitisation resistance, weldability, or a specification that explicitly names an L grade — the method must be optical emission spectroscopy or laboratory analysis. Programmes that specify OES and then execute with XRF for convenience are a recurring and easily proven audit finding."},{"q":"Why does silicon content matter in carbon steel piping?","a":"Because high-temperature sulfidic corrosion rates depend strongly on it. Carbon steel below roughly 0.10% silicon corrodes considerably faster than the higher-silicon material the prediction curves assume, and silicon is not a controlled element in many common product forms, so two fully conforming components can behave completely differently in the same line. The result is localised thinning at one spool or elbow that fixed-point thickness monitoring can miss entirely."},{"q":"Does a material verification programme have to include weld deposit?","a":"Yes, and omitting it is the commonest structural gap in an otherwise credible programme. Correct base metal joined with the wrong filler is still a wrong-material line, and the weld is where failure initiates. Field welds, shop spools and repair welds are the highest-risk populations because that is where an unlabelled electrode gets used. Readings must be taken on a cleaned weld cap with dilution understood, and dissimilar joints need a stated expected chemistry."},{"q":"What extent of PMI does API RP 578 require?","a":"It does not fix a number. The extent is a risk decision the owner makes and documents, informed by consequence of failure, the alloy family, the service and confidence in the supply chain, ranging from spot sampling to 100% of a population. What the practice does expect is that the extent is written down per system, and that the programme states how the sample expands when a discrepancy is found."},{"q":"What happens when a component fails verification?","a":"The programme should define it in advance: quarantine and physically mark the component so it cannot be reinstalled, expand the sample to the batch, heat or supplier, investigate back through the material test report, and obtain an engineering disposition. 'Use as is' is legitimate only when an engineer has evaluated the actual material against the actual service and signed that evaluation. Schedule pressure is not a disposition."}],"sections":[{"heading":"Scope: what API RP 578 covers and what it leaves to the owner","paragraphs":["API RP 578 describes a material verification programme for alloy components in the process industries — piping systems, and by common extension valves, fittings, flanges, gaskets and weld deposits — covering both new construction and existing installed systems. Its subject is the risk that a component is not the alloy the drawing says it is, and its content is a framework: what to include in scope, how to select the extent of examination, which analytical techniques are suitable, how to qualify the person operating the instrument, and what records must be kept and where.","It does not set design limits, corrosion allowances or acceptance criteria for material condition. It does not replace the material specification; the acceptable chemistry is whatever the applicable ASTM or ASME SA specification and the owner's engineering specification say it is. And it does not tell you how many components to test — the extent is a risk decision the owner makes and documents, informed by consequence of failure, the alloy family involved, the service and the confidence the owner has in its supply chain.","RP 578 is also not limited to stainless steel, which is the assumption behind most under-scoped programmes. Low-alloy chrome-moly piping in hot hydrogen service, nickel alloys in HF or caustic service, copper-nickel in seawater systems, and plain carbon steel with a specified minimum silicon content are all legitimately inside a well-written programme, and each brings its own preferred analytical method."]},{"heading":"Writing the programme: scope, extent, method, qualification, records","paragraphs":["A material verification programme that will survive an audit answers five questions in writing. Which systems and which component types are in scope, listed by line number or service class rather than by a general statement of intent. What extent of examination applies to each — 100%, a stated percentage, or a defined sampling rule. Which analytical method is approved for each alloy family, with its known limitations stated. How the operator is qualified and by whom. What record is produced, where it is retained and for how long.","The expansion rule is the clause most often missing entirely. If a sample of ten components returns one wrong alloy, the programme has to say what happens next — typically expansion to a larger sample, or to 100% of that batch, heat or supplier — because a single discrepancy is evidence about the population rather than an isolated curiosity. Programmes written without an expansion rule discover mixed material and then spend a week arguing about what the discovery obliges anyone to do.","Building this document, aligning it with the owner's written practice and with the inspection procedures that reference it, and then defending it in an audit is [ASNT Level III consulting](/consulting) work. The programme is a technical authority document with named methods and named limits, not a form to be filled in by whoever is available on the day."]},{"heading":"New construction versus existing piping systems","paragraphs":["In new construction the leverage sits at receipt and at fabrication. Verifying alloy at material receipt, again after cutting and identity transfer, and again on the completed weld catches the substitution while it is still cheap to replace. The commonest failure is a programme that tests at receipt only and then allows an unmarked offcut, an unlabelled quiver and a shop-fabricated spool to reintroduce exactly the risk that the receipt inspection was written to remove in the first place.","For existing systems the question is different. The piping has been in service for decades, has been repaired by contractors whose records no longer exist, and the isometric is a statement of original intent rather than of present fact. Here the programme is driven by consequence — hydrogen service, high-temperature sulfidic service, HF alkylation, amine, and any line where a wrong component fails suddenly rather than by measurable thinning. Turnaround windows are when this work is realistically done, so scope has to be defined months in advance.","Results are only useful if they attach to the component permanently. Recording each PMI result against an isometric and a component identifier inside an [inspection data management system](/inspection-data-management-system), rather than in a standalone report filed by campaign, is what stops the same line being re-tested three turnarounds later because nobody can find the earlier result or trust that it referred to this elbow."]},{"heading":"PMI methods and the limits that decide which one is defensible","paragraphs":["Handheld X-ray fluorescence is the workhorse of field material verification. It is fast, non-destructive, needs minimal surface preparation beyond removal of coating, scale and paint, and identifies most alloying elements from roughly magnesium or aluminium upward depending on the instrument and its detector. Its hard limitation is that it cannot detect carbon at all, and detects silicon, phosphorus and sulfur poorly. It therefore cannot separate 304 from 304L, 316 from 316L, or confirm carbon steel by chemistry rather than by exclusion.","Optical emission spectroscopy burns a small spot and reads the emitted spectrum, and it can quantify carbon and silicon. That makes OES the required method wherever the carbon grade matters — L-grade verification for corrosion service or weldability — and wherever silicon content in carbon steel governs sulfidation rate. It is slower, needs a clean prepared surface, leaves a visible burn mark, and requires an argon supply, which is why programmes that specify OES and execute with XRF because the argon ran out are a familiar audit finding.","Laboratory analysis of a removed sample or a drilled chip remains the reference method, and it is the correct route for arbitration, for trace elements, and wherever a field result is contested by a supplier with a commercial interest in the outcome. The programme should state in advance which method resolves a dispute, so that the question is settled before there is money attached to the answer."]},{"heading":"The silicon problem: carbon steel in high-temperature sulfidic service","paragraphs":["High-temperature sulfidic corrosion rates in carbon steel are strongly influenced by silicon content. Low-silicon carbon steel — below roughly 0.10% Si — corrodes markedly faster than the higher-silicon material that the standard corrosion prediction curves assume, and the difference is large enough to consume a design corrosion allowance well ahead of schedule. Because silicon is not a specification-controlled element in many common carbon steel product forms, two components that both fully meet their ASTM specification can behave completely differently in the same line.","This is why API RP 578 and API RP 939-C treat carbon steel in sulfidation service as a material verification problem rather than merely an alloy problem, and why the chosen method has to be capable of quantifying silicon rather than simply confirming that a component is not stainless. It also explains a specific failure pattern that recurs in refinery experience: a line that thins at one spool, one elbow or a single old repair section while adjacent, identical-looking components remain sound.","That pattern is exactly the one fixed-point thickness monitoring is worst at catching, because a low-silicon component sitting between two condition monitoring locations is invisible to the survey. Owners subject to OSHA process safety management have to demonstrate that piping remains fit for service and that deficiencies are corrected, so a documented material verification programme belongs in the same record as the rest of the [mechanical integrity](/mechanical-integrity-software) evidence."]},{"heading":"Verifying weld deposit, not just base metal","paragraphs":["A pipe of the correct alloy joined with the wrong filler is still a wrong-material line, and the weld is where the failure will initiate. RP 578 addresses weld deposit explicitly, and any programme that inspects only base metal has a hole in it large enough to fail through. Field welds, shop-fabricated spools and repair welds made under schedule pressure are the highest-risk populations, because those are precisely the circumstances in which an unlabelled electrode from an open quiver gets used without anyone recording it.","Testing the deposit has practical complications that the procedure must address. The reading must be taken on the weld cap, cleaned and free of slag and oxide, and the operator must understand dilution: a cap reading on a thin deposit can be influenced by the base metal beneath it and report a chemistry that is neither the filler nor the pipe. For dissimilar-metal joints the expected deposit chemistry is neither of the two base metals, so the acceptance range has to be stated in the procedure or the technician will raise a false non-conformance.","Repair welding after the original PMI campaign is the classic reintroduction path, and it defeats programmes that were otherwise well written. Any programme whose scope statement ends at construction handover, with no requirement to re-verify after a repair, a tie-in or a component replacement, will drift steadily back to the condition it was written to fix, usually within a single maintenance cycle."]},{"heading":"Non-conformance, disposition and traceability back to the MTR","paragraphs":["When a component returns the wrong chemistry, the programme has to define what happens next: quarantine and physical marking so the component cannot be quietly reinstalled, expansion of the sample, investigation back through the material test report and the supplier, and an engineering disposition. 'Use as is' is a legitimate disposition only where an engineer has evaluated the actual material against the actual service conditions and signed that evaluation. It is not a mechanism for avoiding a replacement when the crane is booked.","Traceability is the part of the record that decays fastest. A PMI result recording 'line 12-P-4501, elbow' and nothing further cannot be relocated by anyone. A result that records the isometric number, the spool, the component tag, the instrument serial number, the calibration verification, the operator and the date can be. Instruments should be verified against certified reference standards at the start and end of each shift, and those verification records retained, because a day's results taken on an instrument that failed its end-of-shift check are all suspect.","Where PMI has been subcontracted, the deliverable is a set of reports whose quality varies considerably between vendors. Independent [review of those reports](/report-validation) against the programme — correct method for the alloy, calibration checks present and passed, weld deposit included, expansion rule applied where a discrepancy appeared — is usually where the gaps become visible for the first time."]},{"heading":"Audit findings and procurement traps in material verification","paragraphs":["The findings repeat across owners with depressing regularity. A programme that names 'critical piping' without defining it anywhere. XRF used to verify L-grade stainless. No expansion rule. No weld deposit verification. Operator qualification asserted as 'trained by the instrument vendor' with no record of what that involved. Calibration verification performed once at the start of a campaign rather than per shift. Results held in a spreadsheet with no link to an isometric. Scope that stops at mechanical completion. None of these needs a new instrument to fix.","The procurement trap is buying PMI as a headcount rather than as a programme. A day rate for a technician with an XRF gun delivers readings; it does not deliver a defensible material verification programme, and at audit the owner is asked for the programme rather than for the readings. Specify the deliverable in the order: the scope list, the extent and expansion rules, the approved method per alloy family, the calibration regime, the record format and identifiers, and the non-conformance route.","Atlantis writes and executes material verification programmes for owners and EPC contractors, including turnaround-scope PMI campaigns, weld deposit verification and independent review of existing programmes, with technicians deployed for the duration of the campaign so the same team carries the traceability through a scope that changes daily. To scope a programme, [request a consultation](/contact)."]},{"heading":"Positive material identification is not a pass/fail test on its own","paragraphs":["API RP 578 is frequently implemented as though a PMI reading either confirms the specified alloy or does not, with the result filed as pass or fail. That understates what the practice actually asks for. A PMI reading returns an elemental composition, and matching that composition against the correct specification and grade requires knowing what was actually ordered — the nominal composition range for the intended UNS designation — not simply confirming that the reading looks like a plausible stainless steel.","This distinction matters most in the cases material verification exists to catch. A 316 stainless component substituted with an unmarked 304 stainless part will both read as austenitic stainless on a basic PMI scan; distinguishing them requires the analyser to resolve molybdenum content specifically, and a PMI programme that records only a generic alloy family rather than a full elemental match will pass a substitution that a specification-literate review would catch. The instrument did its job; the interpretation against the wrong reference did not.","A defensible PMI programme therefore records the full elemental result against the specified UNS or ASTM grade for every reading, not merely a pass/fail against alloy family, and retains that record with the component's traceability documentation — because the audit question that follows a field failure is rarely whether PMI was performed, it is whether the recorded result actually supported the disposition made at the time."]}],"faq":[{"q":"How does API RP 578 relate to API 570?","a":"API 570 governs in-service inspection, repair and alteration of piping systems and requires that materials used in repairs and alterations are verified as correct. API RP 578 supplies the method: how a material verification programme is scoped, executed, recorded and expanded. In practice the API 570 inspector owns the piping decision, and the RP 578 programme provides the evidence that the material actually in the line is what the drawing claims it to be."},{"q":"How often should PMI be repeated on an existing system?","a":"It is event-driven rather than interval-driven. Re-verification is triggered by repair, component replacement, tie-ins, any work performed by a contractor whose material control is unknown, and by the discovery of a discrepancy elsewhere in the same population, batch or supplier. A programme whose scope ends at mechanical completion, with no re-verification requirement after maintenance, will drift back to the condition it was originally written to correct."},{"q":"Does PMI damage the component?","a":"Handheld XRF is non-destructive and needs only local removal of coating, scale or paint at the reading spot. Optical emission spectroscopy burns a small area and leaves a visible mark, which must be dressed afterwards and, on corrosion-critical or thin-wall surfaces, considered before testing rather than after. Drilled chip sampling and coupon removal are destructive and reserved for arbitration. The programme should state which method is approved for each situation."},{"q":"Can a contractor's existing PMI reports be accepted without review?","a":"They should not be. Review checks that the method suited the alloy in question, that instrument calibration verification was performed and recorded per shift, that weld deposit was included and not only base metal, that the expansion rule was applied wherever a discrepancy appeared, and that each result links to a component identifier on an isometric. Reports that record only a line number and an alloy name cannot be relocated later and carry limited evidential value."},{"q":"Does Atlantis execute material verification programmes as well as write them?","a":"Yes. Atlantis writes MVP documents, executes turnaround and new-construction PMI campaigns including weld deposit verification, and independently reviews existing programmes and their records against the practice. Technicians and Level III technical authority mobilise from Houston and Hyderabad and remain deployed for the campaign duration, which matters on turnarounds where scope changes daily and continuity of the same team is what preserves traceability."}]}