Positive Material Identification (PMI) for Valve Bodies and Trim

Why PMI Matters More Than the Certificate Says

Every alloy valve ships with paperwork. The mill test certificate states the grade, the heat number, the chemistry. And yet every year, plants discover carbon steel fittings inside stainless systems, 316L trim that is actually 316 with too much carbon, or duplex steel with a ferrite content so far off spec that chloride cracking follows within months. The paperwork was genuine. The metal was not.

Positive Material Identification exists because material mix-ups happen in real shops: castings get sandblasted and re-stacked, bars from two heats share a rack, a welder picks the wrong filler from an unmarked tin. PMI is the physical verification step that catches these errors before the valve enters service. For alloy services — chloride-containing brines, sour gas, hydrogen, seawater — it is less an inspection nicety and more a survival requirement.

XRF vs OES: Two Methods, Two Jobs

The two dominant PMI techniques look similar from the outside — point the instrument, pull the trigger, read the alloy — but they work differently and suit different checks.

X-ray fluorescence (XRF)bombards the surface with X-rays and reads the secondary fluorescence emitted by each element. It is fast, fully non-destructive, and handheld instruments cover most common grades in seconds.

• Best for: sorting, spot checks on finished machined parts, verification of installed valves
• Limitation: XRF cannot measure carbon. It reads the metallic matrix, so it separates 304 from 316 by molybdenum, but it cannot confirm the L in 316L, and it cannot distinguish carbon steel from low-alloy steel reliably by carbon content
• Surface condition matters: paint, scale, and heavy oxide layers distort readings and must be removed first

Optical emission spectroscopy (OES)strikes the surface with a high-energy spark and reads the light emitted by vaporized material. The spark burns a small crater a millimeter or so across — technically a minor localized repair, not a true non-destructive test.

• Best for: full chemistry including carbon, verification of carbon equivalence in carbon steels, confirming 316L versus 316, checking sulfur in free-machining grades
• Limitation: requires argon, a flat contact area, surface preparation, and a slightly more disciplined procedure; leaves a small burn mark that may need blending on pressure-boundary surfaces

The practical rule many inspectors use: XRF first for fast screening and sorting, OES when carbon or sulfur content is the deciding factor. On a critical order, both may appear in the same inspection plan at different stages.

What Gets Tested on a Valve

A valve is not one material. A single ball valve in alloy service may combine a cast body, forged stem, wrought seat rings, hardfaced trim, and low-alloy bolting — each with its own grade and each with its own opportunity for error. A serious PMI plan addresses:

• Body and bonnet (cast or forged pressure-boundary material)
• Stem, ball, disc, and gate (trim material per API 600 trim numbering or the purchase spec)
• Seat rings and internal fasteners
• Weld deposited overlays (Inconel 625 cladding, Stellite hardfacing) — where XRF verifies the deposited layer, not the substrate
• Welds and heat-affected zones on welded-end valves, both the weld metal and the parent pipe-side material
• Stud bolts, nuts, and gasket retainers — the most commonly mixed items in the entire supply chain, because they arrive loose and identical-looking

Skipping bolting is a classic oversight. A B7 stud torqued into a valve specified for 316/316H bolting will not announce itself until the first chloride-rich washdown corrodes it.

When a Small Chemistry Deviation Becomes a Big Problem

Not every off-spec reading is dramatic, and that is precisely the danger. Three examples from field experience:

Carbon steel posing as stainless.A 304 casting sandblasted to look like every other casting in the shop. XRF catches it instantly — no chromium response. This is the simplest fraud and the most common.

316 sold as 316L.The molybdenum reads fine, so XRF passes it. But carbon above 0.03% means the valve will sensitize in the heat-affected zones of field welds, and intergranular corrosion follows in hot organic or chloride acids. Only OES, or a laboratory combustion carbon analysis, closes this gap.

Duplex with the wrong phase balance.Duplex 2205 and super duplex depend on an austenite-ferrite balance near 50/50. Nitrogen controls the balance, and XRF does not read nitrogen well. A casting solidified too fast, or poured with the wrong nitrogen argon purge, can end up ferrite-heavy with reduced toughness and poor corrosion performance. Feritscope checks and OES for nitrogen are the verification tools here.

Acceptance Criteria and Witnessing

PMI without acceptance criteria invites arguments. A good specification states, for each component category:

• Which method is acceptable (XRF alone, or OES required for carbon-bearing decisions)
• Which elements are checked and against what grade limits (typically Cr, Ni, Mo, Cu, Ti, Nb, plus C when OES is used)
• Sampling extent — common industry practice screens 100% of pressure-boundary alloy components and trim items, with documented instrument calibration blocks of known chemistry
• Reading tolerance — how close a field reading must sit to the grade window before it triggers a laboratory confirmation
• What happens on a fail — quarantine, re-identification of the entire lot from the same heat, and root-cause reporting

Witnessing matters. An inspector who watches the calibration check on a known standard, then watches readings taken on your actual castings, has far more confidence than an inspector reading a report later. When you review our manufacturing and testing capability, ask specifically whether PMI records include instrument ID, calibration standard ID, and raw readings — a report that only says “OK” is not much of a report.

Making PMI and EN 10204 Work Together

EN 10204 3.1 and 3.2 certificates certify the heat chemistry of the mill product. They are necessary and they are not sufficient. The certificate follows the heat; PMI follows the part. Between the mill and your pipeline sit casting, machining, welding, assembly, and storage — and the part’s identity can drift at any of those steps.

The strong combination looks like this:

• 3.1/3.2 certificates for every pressure-retaining heat, reviewed against the purchase spec
• PMI on the finished components, heat-number matched where markings survive machining
• Traceability stamps or laser marks linking each major part to its certificate

One reinforces the other. A certificate with no PMI trusts every hand that touched the metal afterward. PMI with no certificate tells you what the surface is made of but not the heat’s full mechanical test history.

The Fraud Patterns Worth Knowing

Nobody enjoys writing this section, but procurement engineers should know the recurring patterns. Deliberate substitution is rare but not fictional; far more frequent is benign mix-up that looks identical on paper. Both are caught the same way:

• Counterfeit or re-marked bar: grinding off an old heat number and re-stamping. Depth-etching and stamp verification plus PMI close this down.
• “Twinned” certificates: a genuine certificate photocopied with a new heat number. Physical PMI fails to match the paper, and the discrepancy surfaces immediately.
• Trim substitution: WCB body with genuine certificate, cheap 410 stem where 17-4PH was specified. Trim rarely gets witnessed — insist it does.
• Bolting downgrades, as noted above, remain the most common find.

Writing PMI Clauses That Hold Up

A purchase order line saying “PMI required” accomplishes little. Effective clauses specify the method and the evidence:

• State the method per component: XRF screening for all alloy pressure parts and trim; OES mandatory for carbon-content decisions and duplex grades
• Require calibrated instruments with valid calibration certificates traceable to recognized reference blocks
• Define the sampling plan and the witness points you intend to attend
• Require records listing component, heat number, instrument, reading, and operator — retained and handed over with the final documentation package
• State the consequence of a failed reading, including lot-wide re-verification