XRF vs OES vs Lab Analysis: Choosing a Test for Nickel Alloy Scrap

What Is the Difference Between XRF and OES?

XRF measures X-rays given off by the surface of the metal when it is irradiated, while OES measures light given off by a small amount of metal vaporised by an electric spark. That difference in physics explains almost everything else.

Because XRF only excites atoms and removes nothing, it is non-destructive, needs little preparation and fits in one hand. But light elements give off X-rays too weak to measure well, and carbon gives nothing usable at all. Spark OES does burn the metal: a ground spot is sparked under argon, and the light from the plasma is split into wavelengths, each element having its own lines. Carbon, phosphorus, sulfur and boron emit in the far ultraviolet, which air absorbs, so OES optics run under vacuum or argon. That is how OES reads the elements XRF cannot.

For the details of how a handheld analyser reads a nickel alloy, and its specific blind spots, see XRF nickel alloy identification.

XRF vs OES vs Lab Analysis Side by Side

Handheld XRF wins on speed and convenience, spark OES on light elements and carbon, and the laboratory on precision, trace elements and lot-level evidence. The table sets out the practical differences for scrap work.

The burn mark and grinding are irrelevant for scrap headed to a furnace. They matter only for material that might be resold for use, such as surplus pipe or bar.

Which Elements Can Each Method Measure?

The element that decides the grade decides the method, so it helps to see coverage element by element.

Two points stand out. First, nothing in the main-metal rows separates the methods, which is why XRF handles most wrought grades on its own. Second, no single method covers everything: the laboratory pairs ICP-OES for metals with combustion for carbon and sulfur.

Spark OES or ICP-OES: What Is the Difference?

Both measure emitted light, but spark OES vaporises solid metal directly with a spark, while ICP-OES dissolves a sample in acid and sprays it into a hot argon plasma. The shared "OES" causes real confusion in enquiries and contracts.

• Spark OES works on solid metal, on site or at a yard, and gives a result in about a minute. It needs a flat, ground surface and a calibration suited to nickel alloys.
• ICP-OES is a laboratory method. Drillings or chips are weighed, dissolved and analysed against prepared standards. It is slower, but it handles turnings and composites that cannot be sparked, and it reaches lower levels for many elements.

Laboratories also run benchtop wavelength-dispersive XRF, which is far more precise than a handheld unit and is covered for nickel-base alloys by ASTM E2465. So "lab analysis" is not one method: it is whichever combination of ICP-OES, laboratory XRF, wet chemistry and combustion the laboratory uses to report every element the buyer asked for.

When a buyer writes "OES report required", ask which one they mean. A spark OES printout from the loading point and an ICP-OES certificate of analysis from a laboratory are different documents with different weight.

Why Carbon and Sulfur Need Combustion Analysis

Carbon and sulfur are measured in the laboratory by burning a weighed sample in oxygen and measuring the gases produced. ASTM E1019 covers the determination of carbon, sulfur, nitrogen and oxygen in steel, iron, nickel and cobalt alloys by combustion and inert gas fusion techniques.

Why bother on scrap? Carbon separates low-carbon grades from their standard versions, such as 316L from 316, and confirms that Ni-Resist scrap is nickel cast iron rather than a wrought alloy. Sulfur is a concern for melters because it weakens nickel alloys at grain boundaries, and cutting oils and residues on turnings can add it. For the wider family of harmful trace elements, see tramp elements in remelting.

Which Method Should You Use for Each Lot?

Start with XRF for anything wrought and nickel-based, and escalate only when a light element, carbon or a trace element decides the grade or the value.

What About Handheld LIBS?

Handheld LIBS (laser-induced breakdown spectroscopy) sits between XRF and spark OES: a laser vaporises a tiny spot and the instrument reads the emitted light. Depending on the model, it can read light elements, including carbon in some units, in a handheld format. The third edition of API RP 578, published in 2018, added LIBS to the methods it discusses for material verification.

LIBS still needs a clean, ground spot and leaves a small mark. In scrap trading it is used much like portable OES: to settle the light-element questions that XRF leaves open.

Why XRF, OES and Lab Results Sometimes Disagree

Different methods test different amounts of metal from different places, so small disagreements are normal and large ones usually point to the sample, not the instrument.

• Surface against bulk. XRF reads the top fraction of a millimetre. Scale, coatings and chromium-depleted layers shift its result; OES grinding and lab drilling go below them.
• Spot against composite. A spot reading describes one piece. A laboratory composite of turnings describes the average of the drum. If the drum mixes 625 and 718 chips, both are "right".
• Segregation in castings. Cast revert can vary from the root of a blade to the tip, so readings at different points differ.
• Calibration. Each instrument is only accurate for alloy types it was calibrated on. A spark OES set up for steels can misreport a cobalt-rich superalloy.

When results differ, look at the separating element first. If both methods agree on it, the grade is settled even if minor elements wander.

What Should a Sales Contract Say About Testing?

A contract should name the method, the sample, the place of testing and whose result governs if the two sides disagree. Most testing disputes on scrap come from a vague line such as "material subject to analysis", agreed before anyone asked which analysis.

• Method. Handheld XRF, spark OES or a named laboratory method, and the elements to be reported. If carbon or aluminium matters, say so.
• Sample. Every piece, a stated number of pieces per bundle, or a composite built a stated way.
• Place and witness. At the seller's premises, at loading, or on arrival, and whether an independent inspector witnesses it.
• Laboratory. For referee analysis, a laboratory both sides accept, often one accredited to ISO/IEC 17025.
• Tolerance and remedy. What happens if the result falls outside the declared grade: a price adjustment, regrading or rejection.

Agreeing these points before shipment costs nothing. Arguing about them after a container has landed costs freight, time and goodwill.

How Buyers and Sellers Combine Methods

Many professional lots use XRF on every piece and OES or laboratory analysis on a representative sample. The XRF record shows each piece is the declared grade; the OES or lab result proves the light elements, carbon and trace elements for the lot.

For laboratory work, the sample matters as much as the instrument:

The retained split is what settles a dispute if the buyer's numbers differ from yours. For export lots, buyers often appoint an independent inspection company to witness sampling and testing at loading; buying nickel alloy scrap from India covers how that is usually arranged. The records that tie readings to pieces are covered in positive material identification for scrap lots, and the relevant ASTM methods are listed on the testing and identification standards page.

When you send us an enquiry, include whichever readings or reports you have, XRF, OES or laboratory, and say which method produced them. We'll review your enquiry on that evidence.

FAQ