Reducing CO emissions, improving combustion efficiency and ensuring heat is released where it is useful

Improving steel melting efficiency

As a leading European steel producer specializing in recycling scrap metal using the electric arc furnace (EAF) for steelmaking, the GMH Gruppe (GMH) has a long-standing commitment to sustainability and has been a pioneer of green steel since the 1990s, operating with electric furnaces in steel melting. Their goal is to become climate neutral by 2039 through three key levers – renewable electricity, biogenic coal and green hydrogen/electrification. Unlike traditional steelmaking, which relies on conventional carbon sources, GMH operates leveraging a second route that primarily uses electricity. The core equipment in this process is the EAF, which serves as the main aggregate for steel production. This large furnace operates partly using electricity, delivered through a central electrode. The remaining share of heat comes from burners installed along the furnace walls, providing additional energy via natural gas combustion. This represents a significant amount of energy consumption given the plant’s overall demand of over 100 megawatts (MW).

Overarching goals remain consistent year after year across industries: improving energy efficiency reduces costs while enhancing sustainability metrics. However, GMH must carefully assess any intervention, whether they can implement it without halting production, or if it requires planned shutdowns during scheduled summer or winter stoppages. This ongoing effort focuses on pinpointing process hotspots where energy consumption is higher than necessary and targeting them to increase overall efficiency.

To help achieve these goals, GMH leveraged Siemens Digital Industries Software’s Simcenter™ STAR-CCM+™ software, which is part of the Siemens Xcelerator business platform of software, hardware and services.

Targeting burner design for optimization

GMH identified the burner performance as a critical area for improvement. The conventional burner presented several challenges in understanding the complex combustion dynamics. This made it difficult to predict and control flame behavior, affecting efficiency and consistency. The existing burner could not fully meet evolving production demands. Reducing energy consumption was a key objective for sustainability and cost efficiency.

Simultaneously, the company expected the new design to operate with natural gas and hydrogen fuels; however, operating on 100 percent hydrogen remains under development. To address these challenges, GMH required an innovative approach to burner design that could optimize theoretical performance while remaining feasible for real-world implementation.

Applying CFD simulation

At GMH, process modeling has been a core approach since the early 2010s. Given the extreme operating conditions inside an electric arc furnace, with temperatures above 1,600 degrees Celsius (°C) and over 140 tons of molten steel contained within a sealed environment, it is physically impossible to observe or measure conditions directly. No materials, sensors or cameras can withstand such heat, and even sampling molten steel only provides limited, localized chemical data. The GMH team had previously never simulated turbulent combustion and complex chemistry of this scale.

To gain a broader understanding of what happens inside the furnace, GMH relies on physics and computational modeling. By simulating the entire system, including chemical reactions, slag distribution and gas-phase behavior, engineers can identify sources of inefficiency, such as incomplete combustion or excess carbon oxide (CO) formation, and target improvements at their origin.

For this project, GMH applied Simcenter STAR-CCM+ computational fluid dynamics (CFD) software to the combustion of natural gas in the burners, aiming to optimize the burner design. The simulation focused specifically on the final segment of the burner, which is the most complex and critical for flame formation. This section determines how oxygen and natural gas mix and react, affecting the flame’s dynamic behavior.

Simulating reacting flows presented a major challenge due to the complex nature of mixing and underlying chemistry, with various chemical compositions interacting to form a flame. To address this, the team applied a stepwise approach, starting with simpler equilibrium-based models that were less computationally demanding and could run on smaller systems within hours. These initial models provided useful insights into chemical reactions but did not fully capture the observed flame dynamics. To achieve higher accuracy, the team introduced progressively advanced flamelet models, incorporating turbulence effects at various detail levels.

The simulation studies enabled precise modeling of reacting flows, heat transfer and turbulence inside the furnace, allowing for iterative testing of various burner geometries and flame characteristics.

Ultimately, the project employed a large eddy simulation (LES) method for turbulence modeling, combined with a high-fidelity flamelet generated manifold (FGM) model for combustion. This approach required significant computational power and time but delivered results that closely matched real-world burner behavior. By comparing simulated flames with physical burner performance, the team validated the model confidently and reached a stage where the burner design was ready for implementation.

This allowed the team to predict heat transfer, optimize combustion and improve efficiency while maintaining steel quality.

“Our commitment to Simcenter STAR-CCM+ goes back to our very first project at the rolling bench, where we could have precise control of the rolling temperature and set the material properties of the cooling bed without the need for additional heat treatment at the cooling bed,” says Eike Runschke, head of simulation at GMH.

Using CFD for design variation and burner optimization

With a validated CFD framework in place, the focus shifted from understanding existing burner behavior to actively improving performance using design variation and optimization. The GMH team used the simulation model as a virtual test bench to explore modifications to the burner’s internal geometry, particularly in the final segment governing fuel-air mixing and flame stabilization.

The team varied key geometric parameters iteratively and evaluated them against performance criteria, including flame length, peak temperature, CO formation and heat transfer. CFD results showed designs promoting faster, more homogeneous mixing produced shorter, more stable flames with significantly lower CO emissions. This allowed the team to converge efficiently on an optimized geometry before any physical modification.

Going from simulation to manufacturing

After simulation, the challenge shifted to manufacturing. The team needed to design the burner segment so they could construct it precisely, integrate it correctly into the furnace wall and maintain the optimized fluid dynamics the model identified. The optimized burner geometry emerging from the CFD-driven design iterations featured complex internal flow paths that could not be manufactured using conventional techniques. Hence, GMH chose 3D printing as the preferred manufacturing method, ensuring accurate realization of the simulated design.

The project culminated in developing a new 3D-printed burner, the first of its kind at GMH, which has been working flawlessly. This achievement demonstrated that CFD modeling could go beyond theoretical analysis to drive tangible improvements in operational efficiency. By replacing one of the five burners, together with accounting for roughly 10 percent of the furnace’s total energy input, the team achieved a significant technical milestone, supported by careful communication and credibility built from previous CFD projects.

“Thanks to using Simcenter STAR-CCM+, we have gained understanding into mechanisms that would otherwise remain hidden from us,” says Bryan Manuel, a work student at GMH.

Collaborating as a key success factor

A key factor enabling successful collaboration across departments, including operations, research and development (R&D) and simulation, was effective communication. The team emphasized early and ongoing dialogue with personnel involved in the main process to build acceptance and convey potential efficiency gains. Moreover, the CFD simulations played a key role in communicating design concepts to stakeholders. Visualizations helped intuitively convey complex physical phenomena, including temperature distribution and heat transfer.

Achieving a measured impact

Today, one of the five burners in the furnace has already been successfully upgraded, and initial results are highly promising. Consistent with CFD predictions, the optimized burner operates on natural gas and is designed to accommodate up to 50 percent hydrogen. Additionally, experimental tests have confirmed a shorter flame, with higher flame temperature and a significant reduction in CO production. Early indications also point to significant natural gas savings, underscoring the efficiency benefits of the new design.

“Without Simcenter STAR-CCM+ simulation, this project would not have been possible,” says Dr. -Ing. Riadh Omri, CFD expert at GMH. “CFD gave us the confidence to go from concept to implementation and prove real efficiency gains.”

Looking towards the future

Looking ahead, the team plans to scale up burner optimization gradually. GMH will monitor the current 3D-printed burner in production for approximately six months to evaluate durability and performance. During this period, the company may use simulations to refine the design, potentially simplifying components to reduce manufacturing complexity and cost. Once they can validate the performance of the first burner, GMH may install additional burners incrementally and apply similar optimization approaches to other production sites within GMH Gruppe.

Although the burner optimization process has been largely iterative due to the limited space available for modifications and the constraints of staying within a specific burner design category, the team plans to apply automated optimization to the entire burner geometry in the future, including all conduits, allowing for a more integrated and efficient design process.

“The highlight is that we now have a new, 3D-printed burner – the first of its kind at GMH,” says Runschke. “That’s a major step forward.

“Our next project with Simcenter STAR-CCM+ is already lined up. Together with industry partners and universities, we want to develop burners that can burn green ammonia cleanly and efficiently, as an alternative to hydrogen, because without sufficient hydrogen, green steel will need a viable backup.”