Skip to content

Bye-bye Bottleneck - Real time slag analysis with QLX9 at Kardemir Karabük

Redefining Slag Analysis

Short Summary:
At KARDEMİR, high-throughput integrated steel production requires reliable and fast process data. In the ladle furnace, conventional slag analysis created delays and limited process control. With the implementation of QLX9, slag analysis became simple, fast, and always available. This enables real-time decisions without additional personnel or automation.

Initial Situation / Plant Context:
KARDEMİR is one of Turkey’s leading integrated steel producers, located in Karabük. The company operates a full production route with blast furnace (BF), basic oxygen furnace (BOF), and ladle furnace (LF).

The plant produces approximately 2.5–3.5 million tons of crude steel per year. The focus is on long products such as rails, heavy profiles, and structural steels. KARDEMİR is known as Turkey’s only rail producer and plays an important role in national infrastructure and industrial development. The company focuses on quality, efficiency, and continuous process improvement.

In this environment, the ladle furnace is a key step for final steel quality. Slag composition directly influences desulfurization, alloy yield, energy consumption, and refractory wear. Fast and reliable slag data is therefore essential.

Challenge (Pain Points):
In integrated steel plants, slag analysis is often the limiting factor in process control. While metal analysis with spark-OES is fast and well integrated, slag analysis has traditionally been too slow and too complex.

The main limitation is sample preparation. Conventional methods such as fusion or pressed pellets combined with XRF can only process one sample at a time. Even automated systems do not remove this limitation, as the pellet press remains a sequential bottleneck.

In daily operation, this leads to delays. When multiple samples arrive at the same time, for example from parallel ladle furnace operations and BOF, they must be processed one after another. This can result in waiting times of 20 minutes or more for individual samples.

Because of this, slag data is often available too late. It is used for documentation rather than for active process control.

QLX9 at Kardemir in Karabük
QLX9 at Kardemir in Karabük
Laboratory Team with the QLX9 at Kardemir in Karabük
Laboratory Team with the QLX9 at Kardemir in Karabük

Solution (Implementation of QuantoLux Technology)
At the beginning of 2025, KARDEMİR implemented the QuantoLux QLX9 system for slag analysis in the laboratory, directly next to existing spark-OES analyzers. QLX9 is based on Laser OES (LIBS) technology and follows a fundamentally different approach. Sample preparation is reduced to simple crushing. No fusion or pelletizing is required. The analysis itself takes only a few seconds and can be performed by trained laboratory or furnace personnel.

The complete workflow requires about 30 seconds of manual effort plus approximately 15 seconds of analysis time. There is no need for automation or additional personnel. Each sample can be analyzed immediately, without waiting time. This removes the typical bottleneck of slag analysis and enables parallel workflows.

Results (Measured Impact)
With QLX9, slag analysis becomes a real-time tool for daily operation. Analysis time is reduced from several minutes to seconds. The number of samples per shift increases significantly, and preparation-related delays are eliminated. In addition to speed, the system delivers stable and reproducible results. The table below shows typical reproducibility values from selected slag samples under real operating conditions.

Element Sample 1 SD   Sample 2 SD   Sample 3 SD
Fe 2–3 % 0.045   5–6 % 0.188   1–2 % 0.0145
Si 20–25 % 0.19   20–21 % 0.154   21–22 % 0.111
Mn 6–7 % 0.133   8–9 % 0.249   1–2 % 0.0389
Al 9–10 % 0.066   3–4 % 0.057   8–9 % 0.067
Cr 0.03–0.04 % 0.0017   0.06–0.07 % 0.0016   0.07–0.08 % 0.002
Mg 5–6 % 0.044   3–4 % 0.066   6–7 % 0.0433
Ca 50–55 % 0.278   54–55 % 0.51   58–59 % 0.193
Ti 0.5–0.6 % 0.0103   0.5–0.6 % 0.0111   0.5–0.6 % 0.0128
V 1–1.1 % 0.0079   0.9–1.0 % 0.0116   0.2–0.3 % 0.0104

These results show that Laser OES provides reliable precision across a wide range of elements, even without complex sample preparation.

Operational Impact (Daily Work)
Daily work in the laboratory becomes significantly simpler. Operators can analyze slag samples at any time without planning or queuing. Metal and slag analysis can run in parallel, creating a more complete and consistent process picture. This changes the way the ladle furnace is operated. Instead of relying on limited and delayed data, operators receive continuous feedback. This supports faster decisions and more stable process control.

Strategic Context
Integrated steel plants operate at very high throughput. In this environment, even small inefficiencies have a large impact. At the same time, there is a clear trend towards more data-driven production and simplified workflows. QLX9 supports both requirements. It enables fast and simple slag analysis without adding complexity. Laser OES complements existing XRF methods, especially where fast feedback and minimal preparation are required.

Economic Relevance
In high-throughput steel production, even small improvements have a large financial impact:

  • Better alloy yield and optimized desulfurization
  • Longer refractory lifetime
  • More stable process conditions

At the production scale of KARDEMİR, even sub-percentage improvements result in significant annual cost savings.

Conclusion
At KARDEMİR, the implementation of QLX9 shows that removing analytical bottlenecks can fundamentally change process control. By making slag analysis simple and immediately available, slag chemistry becomes an active control parameter—supporting more efficient, stable, and data-driven steel production.