Have PMI testing limitations changed with the advances in technology?

Q: SEVERAL YEARS AGO THIS COLUMN DISCUSSED LIMITATIONS OF PMI TESTING. HAS THE SITUATION CHANGED WITH ADVANCES IN PMI TECHNOLOGY?

A: The short answer is that the situation has changed in some ways, but that PMI still has limitations. The long answer is a bit more complicated. First, let’s clarify what is meant by the term PMI (which stands

API RP 578, Material Verification Program for New and Existing Alloy Piping Systems; ASTM E1476, Metals Identification, Grade Verification, and Sorting; and the ASME B&PV Code Section I, Rules for Construction of Power Boilers, Nonmandatory Appendix B, Positive Material Identification Practice all cover what is commonly known as PMI testing. The definitions differ slightly in these documents, but in general they correspond with the standard concept that most people have in mind regarding PMI—it is a process used to provide the nominal composition of the alloy being evaluated to provide some assurance that it is indeed the specified material.

A number of devices are sold for the purpose of alloy or grade identification. The two most common use either x-ray fluorescence (XRF) spectrometers or optical emission spectrometers (OES).

The XRF devices are by far the most commonly used of the two. Newer XRF devices are generally gun-shaped, operate on rechargeable batteries (much like a portable drill), and use either a radioactive source or x-ray tube to excite the unknown material to generate x-rays, which are then read by the spectrometer. The onboard computer then uses this information to determine the chemistry of the unknown material and can compare that result with compositions stored in an alloy library to identify the alloy.

The previous article (which appeared in VALVE Magazine, fall 2004) discussed the accuracy limitations of the XRF devices that were in use at that time. The newer devices are much more accurate. In fact, there was one fairly recent instance where a newer XRF device caught a composition problem in some castings. The identified discrepancy was somewhat small and several years ago would have been attributed to inaccuracy of the XRF unit. In this case, the results from the XRF unit were verified to be accurate by third-party laboratory chemical analysis. The problem was ultimately determined to have been caused by improper calibration of an OES lab unit at the foundry. This is testimony to the much-increased accuracy of XRF devices in recent years.

On the other hand, the XRF units are still not capable of analyzing all elements of consequence in the alloys used in the process industries. For example, they do not provide information on carbon, nitrogen, phosphorus, sulfur or silicon. This means they still cannot verify the carbon content in carbon steels, alloy steels, stainless steels or other alloys. They cannot distinguish between standard and low-carbon grades of stainless steel. They cannot verify the nitrogen content in many of the newer stainless steels, which are nitrogen-alloyed for increased strength and resistance to chloride pitting. They cannot verify that silicon, sulfur and phosphorus content has been met in any alloys.

More recently, developed portable OES devices, which analyze optical light wavelengths and intensities generated by a spark discharge, can provide quantitative measurement of the elements that cannot be measured by the XRF units. These devices are generally more expensive than the XRF devices, and the spark discharge approach removes a small amount of material, leaving a small burn mark on the surface. Although these devices are “portable” (i.e., not confined to a laboratory), they are much less portable than the XRF devices. In addition, according to ASTM E1476, they are unproven when separation is based on carbon, sulfur or phosphorus. In other words, their ability to discern a low-carbon grade of stainless steel (0.03% maximum carbon) from a standard carbon grade (carbon greater than 0.03%, but no greater than 0.08%) is questionable. Because of their higher cost and reduced portability, these devices are not nearly as popular as the XRF devices. However, where there is a need for analysis of elements that cannot be measured using the XRF devices, the portable OES devices are useful.

As stated in the previous article, if a customer wants components in a valve fully certified to a specified grade chemistry, this can only come from the certified material test reports analysis, which in turn requires implementation of controls to ensure heat number traceability. The proper use of the portable ­analyzers is to verify nominal alloy identification to provide confidence that the material being examined is actually the proper alloy. These devices should not be used to verify full compliance with the compositional requirements of the material specification.