It Looked Right. It Fit. But It Was the Wrong Alloy- Positive Material Identification (PMI) Case study
Updated: Jul 27
What a routine PMI demonstration at a process plant revealed about replacement parts, documents, and the limits of visual inspection
The components on the table
During a recent Positive Material Identification demonstration at a process manufacturing facility, the maintenance and engineering team brought a set of components into the meeting room: a pipe elbow, a tee, a bolt, and a small instrument valve assembly still attached to its manifold.
The situation behind the demonstration was ordinary. The team was preparing to replace a number of components during maintenance. The original installed parts had been specified appropriately for their service — some as Type 304, others as Type 316L. Replacement parts had been sourced through the normal purchasing route. They were the right size, the right dimensions, the right fit, and the right finish.
Nothing about them looked wrong, because nothing about a wrong alloy ever does.
Establishing the baseline
We began with the original components, to establish what “correct” looked like on this equipment.
The elbow — specified as 316L — returned a chemistry consistent with 316-series stainless steel: Fe 66.6%, Cr 17.5%, Ni 12.5%, and Mo 2.18%.
That molybdenum figure is the one carrying information. Molybdenum is a key compositional distinction between 304- and 316-series stainless steels, and contributes significantly to the improved localised-corrosion resistance of 316-series material, particularly in chloride-containing environments. A 316-type stainless typically contains roughly 2–3% Mo; a 304-type contains essentially none.
What XRF cannot tell you is whether a material is 316 or 316L. The L designation is defined by low carbon content, and carbon is too light an element for X-ray fluorescence to measure. So the accurate statement is this: the component was specified as 316L, and its measurable XRF chemistry was consistent with the 316/316L family. That is a useful result — it is simply not the same statement as “confirmed as 316L,” and the difference matters when writing an inspection record.
The valve assembly returned chemistry consistent with the 304-series — chromium and nickel in the expected balance, and molybdenum at effectively zero.
The originals were consistent with what had been specified. The specification had been followed when the plant was built.
Then we tested the replacements
We configured the analyser into PMI pass/fail mode, loaded acceptance limits corresponding to the specified grades for each component, and measured the replacement parts.
They failed.
Not marginally. The measured chemistry was inconsistent with the specified grade and fell outside the configured acceptance limits for the grade the parts had been bought to replace — a part ordered as one thing, delivered as something else, accepted into stores, and queued for installation into a corrosive service line.
Nobody in that room had done anything careless. The parts fit. The paperwork existed.
And none of that is a measurement.
Right size ✓ Right dimensions ✓ Right documents ✓ Right marking ✓ RIGHT ALLOY ? Only measurement can answer that.
Why replacement parts are the dangerous case
There is a version of this problem that plants already worry about: old, unmarked spares of uncertain origin, sitting in a storeroom for decades. That version is real, but it is at least known to be uncertain. People treat it with suspicion.
The version encountered here is more dangerous, because it arrives wearing a clean face. A newly purchased replacement part carries an implicit assumption of correctness. It was ordered against a specification. It came with documentation. It is new. Every instinct in a maintenance workflow says: install it.
Substitution in the replacement-parts supply chain happens for mundane reasons:
A distributor is out of stock in the specified grade and supplies a “commercially equivalent” alternative — equivalent in dimensions and pressure rating, not necessarily in corrosion performance.
The order passes through several intermediaries, and the grade callout degrades at each hop from a precise designation into a general description.
The certificate is generic, belongs to a different heat, or is never checked against the physical item in front of the storekeeper.
The part is marked, but marked incorrectly — and a stamp is a claim, not evidence.
The common thread is that every control in that chain is a document control. Documents can be wrong, mismatched, or simply absent. Only the metal knows what the metal is.
What PMI (Positive Material Identification) actually is
Positive Material Identification is the practice of measuring a material’s chemistry to confirm it is consistent with the alloy it is supposed to be — before it goes into service, rather than after it causes a problem.
In a chemical or fibre production facility this is not a paperwork exercise. These sites run caustic, acids, sulphur compounds, chlorides and elevated temperatures through hundreds of kilometres of piping. The engineer who specified 316L for a particular service did so because material selection formed part of the corrosion and integrity strategy for that service. Substituting a 304-type material where 316L was specified can materially change the expected corrosion performance — and installing the wrong grade quietly removes an engineering decision from the plant.
The removal will not announce itself for months or years, and that delay is what makes it expensive. The wrong alloy usually performs acceptably at first. It holds pressure, it seals, it passes the hydrotest. The consequence appears later, in service, as an unexplained leak, a stress corrosion crack, or a shutdown nobody can immediately account for.
The pass/fail function: how the mismatch was actually found
It is worth pausing on how the problem surfaced, because it was not found by anyone interpreting chemistry at the table.
A handheld XRF analyser can be operated in two ways. In analysis mode it returns a full measurable elemental composition and a best-match grade from its library — useful for engineers, and useful when the material is unknown. In PMI pass/fail mode, you tell the instrument in advance what the material is supposed to be. You configure acceptance limits — from a standard grade, or from your own internal specification — and the instrument compares the measured elemental chemistry against those limits and returns a single verdict.
Pass. Or fail.
This is what makes PMI deployable in a working maintenance department rather than only in a laboratory. The technician at goods-in does not need to know that molybdenum separates the 316-series from the 304-series, or what an acceptable nickel range looks like. They need to know that the screen went red. The metallurgical knowledge lives in the configured limits, set once by someone qualified to set them; the operator’s task becomes a short check with an unambiguous outcome.
Three points on using it correctly:
A pass means the measured chemistry met the configured acceptance criteria — no more than that. It is not a certification against every requirement in the material specification. Carbon content, heat treatment condition, mechanical properties and manufacturing class are not established by a chemistry measurement, and where those are critical they need documentation or a complementary method.
Set the limits from your own specification, not only the generic grade range. If your material specification is tighter than the standard, configure your range. The instrument enforces whatever it is told to enforce.
Treat a fail as a result, not as a bad reading. Confirm it on a suitably prepared surface at a second location, then follow the plant’s quarantine and escalation procedure — rather than re-shooting until a preferred number appears.
What handheld XRF measures — and where its limits are
A handheld XRF analyser excites the atoms in a material with X-rays and reads the characteristic radiation they emit. Within seconds it returns an elemental composition, a grade match, or a pass/fail verdict. It is non-destructive, leaves no mark, requires no consumables, and works on installed equipment as readily as on a part in a box.
What it measures well: chromium, nickel, molybdenum, manganese, copper, niobium, titanium, vanadium, tungsten, cobalt — the elements that differentiate stainless steels, duplex grades, nickel alloys and low-alloy steels. Catching an unspecified alloy substituted for a 304- or 316-series material, as in the case above, is well within its capability.
Where it stops: carbon, as described earlier, and the light elements generally. Handheld XRF covers a large proportion of routine PMI work on stainless steels, nickel alloys and many other engineering alloys — but it is not a universal answer. LIBS, optical emission spectrometry or laboratory analysis may be required where carbon or other analytically critical light elements cannot be adequately measured by XRF. A sensible programme uses handheld XRF for the broad, fast, high-volume verification it does well, and routes the questions it cannot answer to a method that can.
There is also a boundary worth stating plainly, because it applies to every method here: PMI verifies chemistry; it does not automatically verify every property implied by a material specification. Mechanical properties, heat treatment condition and strength class are not determined by elemental composition alone. Chemistry is one part of conformity, not the whole of it.
Surface condition matters. XRF measures a thin surface layer, so paint, coatings, plating, heavy oxide, corrosion products and contamination can distort a result. Where surface condition is likely to affect the measurement, clean or grind to bright metal first — a judgement applied where necessary, not a step required before every measurement.
Reading results that sit near a specification boundary
Pass/fail mode is the right tool at the counter, but engineers should still look at the underlying chemistry, particularly when a result sits close to a limit.
Every XRF measurement carries a reported uncertainty. It is important not to treat that displayed ± figure as automatically defining formal conformity or non-conformity against a specification. What a near-boundary result does indicate is that the measurement warrants additional attention — a longer measurement time, improved surface preparation, measurement at multiple locations, and application of the plant’s defined acceptance and escalation procedure.
The grade name on the screen is a conclusion. The chemistry is the evidence behind it.
The instrument used in this case
The analyser used in the demonstration described here was a ProSpector handheld XRF analyser. What mattered in that case was not a headline specification but the workflow: acceptance limits could be configured for the specific grades the components were specified to, and the replacement parts could then be checked by anyone in the room against a clear pass or fail outcome — which is what turned an abstract concern about material grade into a decision that could be made on the spot.
The bottom line
The original components in that facility were consistent with what had been specified. Somebody had specified them properly, and somebody had installed them properly. The problem did not enter through the design, the engineering, or the maintenance team’s judgement. It entered through a replacement part that arrived looking exactly like what had been ordered.
It took seconds and a red screen to catch it.
PMI is not about distrusting suppliers or colleagues. It is about recognising that dimensions, documents and markings all describe what a part is supposed to be — and putting a measurement where the assumption used to sit.
Which leaves two questions this article deliberately does not answer. Where else in a plant does the same reasoning apply? And what analytical capability would a programme built on it actually require?
Continue reading:
Part 2 — Where PMI Applies in Petrochemical and Process Plants: which components and which points in the maintenance workflow warrant verification, and how to turn testing into a documented programme.
Part 3 — Choosing a Handheld XRF for Plant PMI: which alloys each analyser handles, the detector’s role, and the specification questions that actually change your day.
Looking for a handheld XRF analyzer for PMI, alloy grade identification, or material verification?
ITechSphere supplies and supports the full Elvatech ProSpector handheld XRF range in Thailand — for field and industrial PMI, alloy identification and material verification — backed by local calibration, training, and service. Whatever you test, we help you set the instrument up correctly for your material, so the numbers you rely on are numbers you can trust.
ITechSphere Co., Ltd. — authorized Elvatech distributor in Thailand.
.webp)
Comments