In short
Accurate slag analysis is critical in the steel industry for process control, efficient use of raw materials, and consistent product quality. The two main methods are X-Ray Fluorescence (XRF) and Optical Emission Spectroscopy (OES) — the same optical-emission principle steel mills already rely on for metal analysis, here applied to slag in the QLX9 analyzer. XRF is the established laboratory standard, valued for CRM-based calibration and comparable, standardized results. OES is the faster, process-near alternative, delivering results in seconds with little or no sample preparation. For real-time process control, OES increasingly complements — and in many workflows replaces — laboratory XRF.
X-Ray Fluorescence (XRF): the established backbone
XRF has long served as the backbone of slag analysis in the steel industry. Its use of certified reference materials and decades of operational experience have made it a reliable, well-understood technique in many steel mills. For routine laboratory analysis and standardized reporting, XRF continues to deliver consistent and comparable results. However, its longer analysis times, multi-step sample preparation, and limited suitability for real-time process control increasingly highlight the gap between traditional laboratory methods and the demands of modern steel production.
Advantages of XRF
Industry standard. XRF has long been the preferred technique for slag analysis in steel mills, widely trusted for its reliability, consistency, and operator familiarity.
CRM-based calibration. XRF uses Certified Reference Materials (CRMs) for calibration, providing highly accurate, standardized results. Its accuracy has been validated by international round-robin tests, ensuring comparability across facilities.
Disadvantages of XRF
Slow analysis time. Including sample preparation, XRF typically needs about 10–15 minutes to deliver a result — significantly slower than OES. This delay increases labor costs and limits time-sensitive decision-making.
Post-mortem results. Because results arrive late, adjustments can usually only be made after a process step is complete, making XRF unsuitable for true in-situ monitoring or dynamic operations.
Higher maintenance. XRF systems require regular maintenance and replacement of wear parts, raising operational costs over time.
Sample-preparation risk. The multi-step preparation process (crushing, grinding, de-metallization, pressing) introduces several error sources:
- Insufficient or inconsistent de-metallization, leading to fluctuations.
- Grinding issues such as oversized grains or agglomeration, reducing accuracy.
- Segregation during pressing, introducing inconsistencies.
Complex equipment footprint. XRF workflows require several devices — grinders, presses, crushers, de-metallizers — which are costly to buy, maintain, and operate, and which take up floor space that is often scarce in EAF steel mills.
Light-element precision depends on preparation. This is the key nuance often misstated. XRF can measure light elements such as silicon and magnesium — silicon (as SiO₂) is in fact a major, routinely measured slag component. The challenge is precision versus effort: with fast pressed-pellet preparation, the precision for light elements like Mg (and, to a lesser degree, Si) is limited; achieving high precision generally requires fused-bead preparation, which is accurate but slow and labor-intensive. In other words, XRF forces a trade-off between speed and light-element accuracy.
Slag-analysis images: QLX9 in laboratory operation · slag sample input into the QLX9 · EAF slag sample ready for analysis without further preparation · technician operating the QLX9 for a rapid result.
Optical Emission Spectroscopy (OES): the modern contender
OES is emerging as a modern alternative for slag analysis, built for the speed and flexibility of today's steelmaking. Based on the same optical-emission principle already trusted in metal analysis, it now brings rapid elemental data to slag. With minimal sample preparation and analysis times of seconds, it enables near real-time insight directly at the process. Strong light-element performance and the ability to average many measurements across a larger, more representative sample make it especially suitable for dynamic production environments. As steel mills move toward tighter process control, lower operating costs, and faster decisions, OES is increasingly seen as a forward-looking complement — or successor — to laboratory-based methods.
Advantages of OES
Minimal sample preparation. OES needs little to no preparation — no grinding, pressing, or de-metallization. At most, the sample is crushed, a step also required upstream of XRF for vacuum-pipe transport or grinding anyway.
Speed. Results in as little as ~20 seconds (a full cycle typically under one minute), giving near-instant, in-situ feedback — ideal for real-time decisions and reduced labor.
In-situ analysis. Results directly on-site enable immediate adjustments to slag composition in electric arc furnaces (EAF) or ladle furnaces (LF), reducing process disruptions.
Larger, more representative samples. OES can analyze tens to hundreds of grams at once and collects many spectra per measurement, averaging out inhomogeneity and reducing variability.
Direct analysis, no cooling. OES analyzes the sample directly — including warm or non-conductive material such as slag — with no cooling step required, well suited to fast-paced production.
Consistency with minimal upkeep. OES delivers stable, precise results over time, requiring only occasional standardization rather than constant recalibration.
Low maintenance. With no wear parts to replace regularly, OES instruments are largely maintenance-free and cost-effective over their lifespan. Routine operation needs only a check sample and a standardization sample before the system is ready.
Disadvantages of OES
Calibration approach. OES calibration for slag typically relies on secondary/internal reference materials rather than CRMs, which currently makes it best suited to internal process monitoring rather than formal, standardized quality reporting.
Adoption still growing. OES for slag is proven in a growing number of steel mills but has not yet reached the universal adoption of XRF — though this is changing as more facilities recognize the benefits.
OES vs. XRF at a glance
| Criterion | XRF (laboratory) | OES (QLX9) |
|---|---|---|
| Analysis time | ~10–15 min (incl. prep) | ~20 s; full cycle < 1 min |
| Sample preparation | Crush, grind, de-metallize, press | Little to none (crush only) |
| Process suitability | Post-process / lab | Real-time, in-situ at EAF/LF |
| Calibration | CRM-based (standardized) | Secondary/internal references |
| Light elements (Si, Mg) | Good with fused beads; limited with fast pellets | Strong, no weight constraints |
| Sample size / representativeness | Small pressed pellet | Tens–hundreds of g, many spectra averaged |
| Maintenance / wear parts | Regular | Minimal, largely wear-free |
| Footprint | Multiple prep devices | Compact, single instrument |
| Industry adoption | Established standard | Growing |
Precision: a real-world comparison
Published QuantoLux comparison data for slag shows OES precision on par with high-quality lab preparation, and far better than fast pressed-pellet XRF:
| Component | XRF — pressed tablet (SD) | XRF — fused bead (SD) | OES / QLX (SD) |
|---|---|---|---|
| Al₂O₃ | 0.65 % | 0.26 % | 0.32 % |
| SiO₂ | 0.93 % | 0.14 % | 0.20 % |
| MgO | higher than fused bead | reference-grade | 0.10 % |
The takeaway: the fast XRF route (pressed pellets) trades away light-element precision, while the precise XRF route (fused beads) trades away speed. OES delivers fused-bead-class precision at process speed, by averaging many measurements across broken material instead of relying on labor-intensive homogenization.
Why OES is well positioned for modern slag analysis
XRF remains the standard, thanks to its reliability, CRM-based calibration, and widespread adoption, and many mills already run legacy XRF systems that work well for routine reporting. But its slow turnaround, ongoing maintenance, and the speed/accuracy trade-off for light elements make it less suited to fast, dynamic production.
OES addresses these limitations directly. Its speed, minimal preparation, and strong light-element performance — including silicon and magnesium, which are critical for slag chemistry — make it well matched to modern steelmaking. Real-time monitoring helps optimize slag composition to improve foaming, reduce refractory wear, and increase yield. For mills handling volatile input streams such as secondary resources, OES provides the speed and flexibility to respond dynamically, while low maintenance and long-term reliability help reduce operating costs.
Frequently asked questions
Is OES better than XRF for slag analysis? It depends on the goal. For standardized, CRM-based reporting, XRF remains the reference. For fast, in-situ process control, OES is generally the better fit because it delivers results in seconds with minimal preparation. Many mills use them together — OES for real-time control, XRF for formal QC.
How much faster is OES than XRF? OES delivers a result in roughly 20 seconds, with a full cycle under a minute. XRF typically takes 10–15 minutes once sample preparation is included — often reducing analysis time from about 15 minutes to around 1 minute in practice.
Can XRF detect light elements like silicon and magnesium in slag? Yes — silicon especially is a routine XRF analyte. The real issue is precision versus preparation effort: fast pressed-pellet XRF has limited precision for light elements, while fused-bead preparation restores precision but is slow. OES achieves strong light-element precision without that trade-off.
Does OES require certified reference materials (CRMs)? Typically no. OES slag calibration usually uses secondary/internal reference materials, which suits internal process monitoring well. CRM-based XRF still has the edge for formal, standardized reporting.
Can OES replace XRF entirely? For real-time process control, increasingly yes. For standardized quality certification, many facilities still keep XRF. The practical trend is complementary use, with OES taking over the time-critical, in-situ role.
Conclusion: OES as the future slag analyzer of choice
While XRF is still the backbone of slag analysis in many steel mills, OES is emerging as the stronger choice for modern, fast-paced operations. The industry's need for faster, more accurate, and more cost-effective analysis makes OES a natural evolution. For facilities prioritizing real-time in-situ control, reliable light-element analysis, and lower labor and maintenance costs, OES is well placed to become the default solution. Investing in OES today means meeting current demands — and preparing for the demands of tomorrow's steelmaking.
More from the author: https://www.linkedin.com/in/alexander-schlemminger-44768187/